ࡱ> y bjbj {{]'(f)f)f)D)))**+)Ar- /(/002 55t$[ *̧f)b622b6b6̧f)f)00H$@@@b6f)0f)0@b6@@…0@p=7<<A777xJ7f)ؐb6b6@b6b6b6b6b6̧̧<b6b6b6Ab6b6b6b67b6b6b6b6b6b6b6b6b6 &:  Broadband performance monitoring and reporting in the Australian context Consultation paper 14 August 2013 Contents  TOC \o "1-2" \f \h \z  HYPERLINK \l "_Toc364245657" 1 Overview  PAGEREF _Toc364245657 \h 1  HYPERLINK \l "_Toc364245658" 1.1 Why is the ԭ considering this program?  PAGEREF _Toc364245658 \h 1  HYPERLINK \l "_Toc364245659" 1.2 What are the likely benefits for stakeholders?  PAGEREF _Toc364245659 \h 2  HYPERLINK \l "_Toc364245660" 2 Consultation process  PAGEREF _Toc364245660 \h 5  HYPERLINK \l "_Toc364245661" 3 International broadband performance monitoring and reporting programs  PAGEREF _Toc364245661 \h 6  HYPERLINK \l "_Toc364245662" 4 Broadband performance monitoring and reporting in Australia  PAGEREF _Toc364245662 \h 8  HYPERLINK \l "_Toc364245663" 4.1 Testing methodology  PAGEREF _Toc364245663 \h 8  HYPERLINK \l "_Toc364245664" 4.2 Scope  PAGEREF _Toc364245664 \h 10  HYPERLINK \l "_Toc364245665" 4.3 Metrics  PAGEREF _Toc364245665 \h 20  HYPERLINK \l "_Toc364245666" 4.4 Reporting  PAGEREF _Toc364245666 \h 22  HYPERLINK \l "_Toc364245667" 5 Summary of questions  PAGEREF _Toc364245667 \h 25  HYPERLINK \l "_Toc364245668" Attachment A  PAGEREF _Toc364245668 \h 28  Overview The ԭ is considering the implementation of a monitoring and reporting program in relation to the performance of broadband internet services in Australia, subject to securing financial support. A broadband performance monitoring and reporting program would involve testing data transfer rates and other quality of service metrics for broadband services. The ԭ would then compile the resulting data and publicly release reports discussing the results. Similar programs have been implemented in a number of international jurisdictions including the UK, US, Singapore and New Zealand. These programs are outlined in further detail in section  REF _Ref361132010 \r \h 3 and throughout this paper. The purpose of this consultation process is to enable the ԭ to develop a plan for a robust monitoring program which would provide meaningful and insightful reporting for stakeholders on the performance of broadband services in Australia. The ԭ is seeking to engage closely with industry and consumer groups (and any other relevant stakeholders) on the appropriate testing methodology, scope, quality of service metrics and reporting framework for the proposed program. These are discussed in section  REF _Ref361663954 \r \h 4. The ԭ considers that a broadband monitoring and reporting program would encourage competition and efficient investment in infrastructure in the Australian broadband market as well as providing significant benefits for consumers. However, the scope and nature of any program, if implemented, will depend upon technical and budgetary considerations. Following this consultation process, the ԭ intends to release a position paper in 2014 outlining how it would implement a broadband performance monitoring and reporting program if the necessary funding was obtained. Why is the ԭ considering this program? In the last 10 years there has been a significant increase in the demand for high quality broadband services in Australia, which has been fuelled by the development of new devices and attractive service offerings by internet service providers (ISPs). There were 12.2 million internet subscribers in Australia at the end of December 2012, with almost 98 per cent of these subscribers acquiring a broadband service. Broadband services are high speed, always-on internet services provided over a wide range of technologies. In Australia, broadband services are provided over copper (ADSL and VDSL), hybrid fibre-coaxial cable (HFC or cable), optical fibre, wireless (mobile and fixed) and satellite networks. In 2009, NBN Co was created to roll out the National Broadband Network (NBN). The aim of the NBN is to improve the availability of broadband services to the Australian public regardless of where they live. This infrastructure project recognises the increasing significance of broadband services in the lives of consumers. As broadband services become more widely available through the NBN rollout, and as demand for higher data transfer rates (commonly referred to as speeds) increases, the ԭ expects that the quality of broadband services will become more and more important for consumers. The ԭ has also previously raised concerns regarding the marketing of broadband services, particularly around the use of potentially misleading headline claims by ISPs. The ԭ has investigated a range of claims about the performance of both fixed-line and mobile broadband services. In addition, the ԭ published industry information papers setting out the ԭs position on the marketing of ADSL, mobile broadband and HFC/optical fibre services in 2007, 2009 and 2011 respectively. What are the likely benefits for stakeholders? Consumers A broadband performance monitoring and reporting program could provide transparency to consumers about the performance of broadband services available in Australia. To date, ISPs have competed largely on the basis of price, headline speeds and download quotas. While some information on service quality (e.g. average data transfer rates, reliability) is available for mobile broadband services, this information has not been as freely available for consumers of fixed-line broadband services. Information on the real-world performance of broadband services is important to enable consumers to choose the right service for their needswhether acquiring a new service or considering renewing an existing service. Service quality between different networks can vary significantly. For example, two ISPs may offer a plan at the same price point with the same amount of data but if one provider has invested heavily in network capacity while the other has not, the end-users experience is likely to be significantly different with each ISP. The ԭ considers that competition works best when consumers are informed and can make the right choices for their needs. Comparison data would not only serve to help consumers choose the fastest or most reliable service but would also enable consumers to weigh up their needs and budget to make an appropriate cost/performance decision. In addition, a testing program would provide protection to consumers by holding ISPs accountable for any performance claims, including headline speed claims. This could be achieved by making consumers aware of the real-world performance of a particular ISP which they could then compare against the claims being made, as well as providing the ԭ with data which would enable it to assess whether certain claims may be misleading and/or deceptive under the Australian Consumer Law. The ability to differentiate services based on quality may also lead to an increase in overall investment in capacity by ISPs and improved services for consumers as a result. Experiences in other jurisdictions suggest that service performance for individual ISPs with poor testing results in one reporting period has improved significantly the next reporting period. Internet service providers A broadband performance monitoring and reporting program would give ISPs access to information on the real-world performance of their services and their competitors services, enabling them to better differentiate their services in a way which is open and transparent. Specifically, the proposed program would allow ISPs to differentiate their products and services from their competitors based upon the results of the ԭs testing results. For example, an ISP could advertise that its service offerings represent greater value than those of an ISP with similar network performance. This would draw out an extra competitive dimension in addition to price and download quotas. The ԭ recognises that many ISPs and network operators have been or are beginning to roll out additional fixed and mobile broadband infrastructure and increase capacity in order to address capacity constraints, increase functionality and improve consumer experiences. In this regard, the ԭ considers a broadband performance monitoring and reporting program would reward ISPs for these investments by drawing consumer attention to improvements in service quality and allow them to more readily substantiate any performance claims across their service offerings. The proposed program would also allow ISPs to identify potential service quality issues which they could in turn raise with their network operator. Further, in some cases the ԭ could use test results to triangulate on an issue within an access network. For example, if the ԭ observed poor performance from all ISPs supplying NBN-based services in a specific area, this would likely indicate an issue at the wholesale access level rather than the ISP level. While ISPs are ultimately responsible for ensuring they deliver performance outcomes in line with their retail product marketing, the ԭ would take into account any failures to deliver promised performance by network operators when analysing data and preparing reports. In addition, if there was evidence of a network operator over-promising and under-delivering the ԭ could consider enforcement action for misleading or deceptive conduct and/or for failure to comply with any applicable regulatory determinations. Testing companies Information technology companies (referred to throughout this paper as potential testing companies) could also benefit from being involved in an Australian broadband performance monitoring and reporting program. There would be potential for the company administering any program implemented by the ԭ to offer additional services to ISPs. For example, the testing company could provide individual ISP testing, based on the ԭs testing methodology, on the performance of certain services so as to allow the relevant ISP to substantiate particular performance claims. End-user volunteers Under the ԭs proposed testing approach (discussed in section  REF _Ref361735769 \r \h 4.1 below), the ԭ would be relying on consumers across Australia to volunteer their broadband service for testing. In addition to assisting the ԭ with administering the proposed program, volunteers would likely be provided with software which allows them to check the performance of their own connection. This would enable volunteers to evaluate the performance of their service on a regular basis and determine the adequacy of the service for their needs. The testing program would be designed to ensure that volunteers privacy is maintained and would not record private internet traffic. The identity of volunteers would not be disclosed. Similarly, the program would be designed so that volunteers are not inconvenienced by the running of tests across their service. The testing equipment would require minimal electricity and would likely run tests only when the user is not using their connection (testing units deployed internationally can conduct tests in very small connection inactive windows such as between clicks when a user is browsing the internet). Consultation process The ԭ is seeking submissions from a range of stakeholders including industry and consumer groups on the specific questions raised in this consultation paper, but invites comments on any other matters relating to the introduction of a broadband performance monitoring and reporting program in Australia. All submissions to this consultation paper will be considered public and posted on the ԭs website. If you wish to provide commercial-in-confidence material to the ԭ, a public version should also be provided. The public version should clearly identify the commercial-in-confidence material by replacing the confidential material with [c-i-c]. The ԭ and AER general policy on the collection, use and disclosure of information is set out in the ԭ-AER information policy: the collection, use and disclosure of information. A copy of the guide is available on the ԭ website. The ԭ prefers to receive submissions in electronic form, either in Adobe PDF or Microsoft Word format which allows the text to be searched. Submissions will be accepted until 5:00 pm on Friday, 13 September 2013. Submissions received after this time may not be considered. Please forward submissions to:  HYPERLINK "mailto:BroadbandPerformance@accc.gov.au" BroadbandPerformance@accc.gov.au For further information, please contact: Rebecca Larsen (03) 9290 1939. International broadband performance monitoring and reporting programs Broadband performance monitoring and reporting programs have been implemented by a number of international regulators including in the UK, US, Singapore and New Zealand. In 2008, Ofcom commenced a research program into the performance of fixed-line broadband services delivered to UK residential consumers. Ofcom currently monitors ADSL, HFC and fibre-to-the-cabinet (FTTC) services. The purpose of the program was to identify average actual download throughput speeds, along with a number of other metrics, to determine the consumer experience with broadband services. Ofcoms Chief Executive, Ed Richards, stated in 2011 that the UK broadband market had been transformed since Ofcom published its first report in early 2009. One of the results of the research is that consumers are increasingly moving towards higher rated services and enjoying faster speeds as a result of greater access to broadband performance information. In 2012, Mr Richards reiterated that Internet providers are working to meet consumer demand through network upgrades and the launch of superfast packages, giving consumers faster speeds and greater choice. In 2011, the Federal Communications Commission (FCC) released the first Measuring Broadband America report which outlines the results of a nationwide study of actual residential fixed-line broadband performance in the US. This study arose as a result of the National Broadband Plan recommendation that the FCC obtain and publicly release detailed and accurate measurements of consumer broadband performance. The FCC monitors ADSL, HFC, and fibre as well as satellite services. The FCC has indicated an intention to report on fixed wireless services in the future. The FCCs Measuring Broadband America program focusses on throughput speeds experienced by consumers in peak or busy periods. Since it began, the FCC has observed that ISPs have improved in both their ability to deliver what they promise to customers and in the overall speeds that they can and are delivering. The FCC has stated that it has seen significant investment by ISPs in their networks and the gap between advertised and actual speeds has narrowed. Singapores Infocomm Development Authority (IDA) commenced its Consumer Broadband program in 2011. The Consumer Broadband program is designed to provide consumers with information to enable them to compare different ISPs broadband services. The IDA provides information to consumers on the performance of services as well as pricing in the form of interactive charts on its website. The New Zealand Commerce Commission (NZCC) commenced an annual monitoring and reporting program in 2012. The NZCC monitors ADSL, HFC, FTTH and FTTN/VDSL. The NZCC has not yet produced a report on the results of the program, though its contractor TrueNet releases a brief broadband performance report each month. The NZCC has indicated the study will be more oriented towards measuring the kind of performance that is delivered to consumers compared to research undertaken for previous reports on broadband performance quality (which appear to have been highly aggregated). The ԭ considers that these international experiences provide some insight into how a broadband performance monitoring and reporting program could be implemented in the Australian context. Broadband performance monitoring and reporting in Australia An Australian broadband performance monitoring program which is credible and robust would require careful design and implementation. If the relevant stakeholders, including ISPs and consumers did not have faith in the accuracy of the program, they would be unlikely to constructively engage with the results it produced. The ԭ is, therefore, seeking views on the following aspects of the proposed program: the testing methodology the scope of testing which performance metrics should be tested how the results of the testing should be reported. There are some aspects of the consumer broadband experience which would fall outside the scope of this program. For example, the ԭ does not propose to monitor or provide information on broadband pricing or individual consumer broadband service performance. Rather, the focus would be on providing high level information to consumers on the performance of particular ISPs and broadband networks. It is also important to note that there are other factors which may affect individual consumer broadband experiences which would not be captured by the program. For example, consumers may experience slow data transfer rates for certain websites or when trying to access certain content servers. The use of several devices over a connection, poor configuration of residential wireless gateways and slow devices (e.g. computers) or networking equipment (e.g. modem/routers) may also have an effect on an individual consumers broadband experience. Testing methodology There are a number of ways to test broadband performance. These include probe-based, laboratory and software-based testing. The ԭ considers that a probe-based testing methodology would be most appropriate in the Australian context. Probe-based testing has been adopted in the UK, US, Singapore and New Zealand. This approach requires the recruitment of volunteers acquiring fixed-line broadband services. A sample is then chosen from the pool of volunteers to ensure that the results are representative of the market segments being tested (e.g. region, service technology). Each market segment requires a threshold number of volunteers to be monitored to ensure that the data is reliable. Volunteers are provided with hardware monitoring units (HMUs) which contain software which is programmed to run periodic tests on various aspects relating to the performance of the service. HMUs may be installed at various points on the end-users connection. However, the most likely position is between the end-users router or modem and their residential network. This allows the HMU to determine when the network is free to run teststhereby avoiding disrupting the end-users personal use. Below is a diagram outlining where the HMU may be located in the broadband service supply chain.  The HMUs being used in other jurisdictions do not log private internet traffic and the identity of volunteers is not disclosed to ISPs. This ensures the privacy of volunteers and avoids the manipulation of testing results by ISPs. Furthermore, the HMUs require only a small amount of the volunteers data quotas.  Probe-based testing generates large amounts of data. In the UK, for example, HMUs are scheduled to run over 14,000 tests over the course of a day. The probe is configured to behave like an end-user performing a range of normal tasks using their internet connection (e.g. browsing websites, sending/receiving emails, watching online videos, making internet phone calls). In doing so, the probe measures a range of variables which directly impact upon the real-world end-user service experience. As a result, the probe-based approach can produce a quality and quantity of data that supports detailed analysis and reporting. Software-based tests can be conducted in a number of ways and may involve downloading and installing testing software on a volunteers device (PC, laptop or smartphone) and using this software to run periodic tests. Alternatively, software may be used to run passive network tests to monitor traffic over the network. There are limitations with software-based testing due to factors unrelated to the quality of the service provided by the ISP, which may adversely affect the results. For example, the capabilities and configuration of the volunteers device or the number of devices the volunteer has connected to the network. Laboratory testing involves setting up laboratories at sites across the country to conduct tests. However, this approach involves high set-up costs and an increased risk of ISPs and network providers manipulating the results by providing services of a higher standard to testing areas. The probe-based approach has generally been supported by ISPs involved in the abovementioned international programs because it produces more reliable and accurate data than other options. However, the ability to compare the performance of services and particular ISPs is impacted by the ability to acquire a sufficient number of volunteers to allow for meaningful statistical analysis. This is discussed further in section  REF _Ref361131994 \r \h 4.2. Ofcom, the FCC, the IDA and the NZCC have all engaged third party companies to recruit volunteers and conduct the testing. The ԭ would likely also engage an independent third party to conduct the testing and, in the event that the ԭ decided to implement the program, would publish a formal request for tender. In that regard, the ԭ would appreciate views from potential testing companies on a number of the issues discussed in this paper which would affect the technical and cost viability of the program. Questions on which the ԭ is seeking views: Do you agree that a probe-based testing methodology would be the most reliable and accurate approach for the Australian context? If you consider an alternative approach preferable, what approach do you prefer and why?Scope Should the ԭ secure the financial support required to implement a monitoring program, a key issue to resolve would be the breadth of data which should be collected by the proposed program and the sample configuration which would allow for accurate results while meeting budgetary constraints. As the primary aim of any broadband performance monitoring and reporting program would be to provide information to industry and consumers on the real-world performance of broadband services in Australia, the ԭ would need to ensure that the sample is representative of the Australian broadband market. The ԭ is currently seeking views on four parameters which would help determine the scope of any program in Australia: the types of services which should be tested the cities or regions from which data should be collected the number of ISPs whose services should be tested the number and type of speed tiers which should be tested. If the ԭ was to adopt a probe-based approach, the number of probes for each service, ISP, region and speed tier would need to be determined. In this regard, it is important to note that the accuracy and quality of the testing data increases with the number of distributed probes. However, the number of distributed probes (and therefore, the cost of the program) increases rapidly with each additional ISP, region or speed tier. The ԭ would, therefore, need to weigh the benefits of obtaining particular information with the associated costs for obtaining it. Increases in costs which do not proportionately increase the benefits provided by the program would be unlikely to be adopted. The ԭ requests that stakeholders keep this in mind when responding to the questions outlined below. The ԭ is committed to engaging closely with ISPs to ensure that the results of any testing would accurately represent the current state of the broadband services market, particularly as it changes over time, and any other issues which may affect the reliability of reporting. For example, the ԭ would seek to ensure that ISPs with a particular customer base would not be disadvantaged by decisions made on the appropriate sample set. Services Broadband services in Australia may be supplied to premises using ADSL, HFC, optical fibre, fixed wireless or satellite technologies. Alternatively, they may also be supplied over 3G or long term evolution (LTE or 4G) mobile networks. The majority of broadband services in Australia are currently provided using ADSL, HFC and mobile (3G/4G) technology. However, as the rollout of the NBN progresses, the ԭ expects the number of optical fibre, fixed wireless and satellite services to increase. ADSL services are provided over Telstras ubiquitous copper network and constitute the majority of fixed-line broadband services in operation in Australiaapproximately 80 per cent in December 2012. There are also two major HFC networks operated by Telstra and Optus. These networks are located in Adelaide, Brisbane, the Gold Coast, Melbourne, Perth and Sydney. iiNet operates smaller HFC networks in regional Victoria as well as FTTP and FTTN networks in the ACT and Queanbeyan. Whilst there are a number of additional optical fibre networks operated around Australia, these are mainly in greenfield estates (that is, new housing estates which are being or have recently been developed on vacant land) and do not constitute a large proportion of the Australian fixed-line broadband services market. For example, as at December 2012 HFC services represented around 16 per cent of fixed-line broadband services compared to 2 per cent for optical fibre. Many consumers will continue to use ADSL and HFC services throughout the rollout of the NBN. Therefore, the ԭ considers that any broadband performance monitoring and reporting program should include ADSL and HFC services as it would provide useful information to consumers in the short to medium term. As the rollout accelerates and consumers migrate to the NBN, these services would become less relevant and the testing of services provided over the NBN would increase. Testing probes could be redeployed to NBN-based services and/or additional volunteers could be sought over time so that a representative sample is maintained. Below is a table of the largest fixed-line broadband networks in Australia based on the number of services in operation: Network technologyMajor network operatorsCoveragePercentage of total fixed-line broadband servicesADSLTelstra Optus, TPG, iiNet (via ULLS)National~80%HFCTelstra, Optus, iiNetSelected metropolitan cities and regional VIC~16%FTTPiiNet, OptiCommSelected areas across Australia (incl. ACT and greenfield estates)~2%FTTN (VDSL)iiNetACT/QueanbeyanNBNNBN Co*Roll out in progress*Until recently, the majority of international regulators have chosen not to include mobile broadband services in their monitoring and reporting programs. The FCC indicated in its July 2012 report that mobile broadband services were not included in the study due to special challenges inherent in measuring the actual performance of mobile networks but that it was working to address those challenges. The FCC has now indicated that it will be releasing a separate report on mobile broadband performance in 2013. Ofcom undertook a one-off mobile broadband performance study in 2010. Ofcoms study of mobile broadband performance focused on services delivered via dongles and data cards, as opposed to those delivered to smartphones and other mobile devices (which form the majority of mobile broadband services in Australia). This study was not continued. However, Ofcom has indicated that it will be carrying out research around December 2013 to assess the average mobile broadband speeds received by 3G and 4G customers. Ofcom expects to publish the results in spring 2014. Australia has very high mobile broadband penetration rates, with mobile technology accounting for 49 per cent of all broadband connections. Consumer reliance on mobile broadband continues to increase with the growing popularity of smart phones and tablets. However, the testing methodology outlined in this paper would not be suitable for testing the performance of mobile services and their inclusion would significantly increase the costs of implementing the program. Accordingly, the ԭ is proposing to limit the scope of any program to fixed-line broadband services in the short term, but would consider options for the inclusion of mobile broadband services in the future. Another issue to consider is whether the ԭ would monitor residential broadband services alone, or whether it would include business-oriented services in the program. Of the international regulators mentioned above, the NZCC is currently the only one which monitors both business and residential services. Small businesses are increasingly relying on the internet for their day-to-day operations with some being completely internet-based. Similar to residential consumers, small businesses are unlikely to have visibility over the performance characteristics of their broadband services and would benefit from having such information available when assessing whether particular services are suited to their needs. From a technical perspective, small businesses will often obtain the same kinds of services as residential customers even if they are offered faster fault rectification times and/or better support in terms of customer service availability. On the other hand, larger businesses are likely to be in a better position to ascertain whether a particular service is appropriate for their needs, and in many cases will use business-grade services which have specific performance commitments. For those reasons, and given the additional costs associated with obtaining this data, the ԭ would not currently propose to monitor business-grade services. In summary, the ԭ considers that an appropriate broadband performance and monitoring program in Australia would monitor residential and small business fixed-line broadband services provided over ADSL and HFC networks as well as the NBN (including optical fibre, fixed wireless and satellite services). Questions on which the ԭ is seeking views: What services should be included in the ԭs proposed performance monitoring and reporting program? In particular: Do you agree that the ԭ should monitor ADSL, HFC and NBN-based broadband services? Do you agree that the ԭ should monitor small business broadband services? Are there any other services which you consider should be included in the proposed program? In your response, please outline reasons.Regions Another key question is the geographic scope of the proposed program. The ԭ is currently considering whether data should be aggregated at a national level or whether it should involve a representative sample from specific cities or regions to provide more localised analysis. The FCC has stated in its report that it analyses performance data at the national level as the results are not collected in a way that permits meaningful conclusions at the local level. At this stage, the ԭ has a preference to monitor broadband performance at a more disaggregated level which could assist ISPs and network operators to identify areas where further investment in capacity should be made and to provide consumers with a better idea of the broadband performance likely to be experienced in their area. One approach may be to monitor major metropolitan areas onlyfor example, the 10 largest Australian cities by population. Whilst this would be the simplest approach, the results may not reflect the broadband experiences of consumers in rural and regional areas due to the different network characteristics in these areas (for example, longer backhaul routes). The ԭ could, however, in addition to metropolitan cities, obtain data on a state-by-state basis for rural and regional customers. For example, each State/Territory could be separated to include regions consisting of the largest metropolitan cities (e.g. capital cities and large metropolitan areas such as the Gold Coast and Newcastle) and a single region consisting of all rural and regional areas outside the relevant metropolitan cities. Ofcoms participants are distributed in three defined geographical markets ranging from the least populated Market 1 (approximately 12% of premises in the UK) to the most populated Market 3 (approximately 78% of premises). Ofcom also separates services into urban (Markets2/3) and rural (Market 1) markets to compare differences in service quality arising from the different geographical characteristics. Whilst Ofcom does report by region, it has stated that the sample sizes in certain regions of the UK result in a larger error margin for some regions as compared to others. This may be an issue for the ԭ to consider, particularly for rural and regional areas in Australia, where it may be difficult to obtain sufficient sample sizes for each service, ISP and speed tier. It is important to note that the ԭs proposed monitoring and reporting program would not provide information on individual consumer broadband performance. However, it would show trends on service performance between different technologies, ISPs and regions. For example, it would likely be able to show if a particular ISP or network operator has not invested in sufficient capacity (such as backhaul transmission) in a particular region. The ԭ is seeking views from stakeholders on the appropriate geographical regions for the proposed program. Questions on which the ԭ is seeking views: How should the ԭ determine which regions to monitor as part of any program? In particular: How many Australian cities do you consider should be monitored as part of the proposed program? How could these be determined by the ԭ? Would you consider State or Territory regions which encompass rural and regional areas outside of each major city would be sufficient to provide information to consumers living in these areas on the performance of broadband services? For example, a Victorian rural/regional delineation which encompasses services outside of metropolitan Melbourne.Internet service providers There are a large number of ISPs supplying ADSL services in Australia. Similarly, there are currently over 40 retail ISPs supplying services over the NBN. This number is likely to increase as the rollout progresses. In regards to HFC, there are only three providers in AustraliaTelstra, Optus and iiNet. Telstra and Optus provide the majority of HFC services. The ԭ does not propose to monitor each ISP providing broadband services in Australia. In addition to the high costs which would be incurred in doing so, it may be difficult to establish a representative sample for smaller ISPs. Both the FCC and Ofcom currently monitor the largest ISPs by market share. In Ofcoms report from November 2012, it measured eight ISPs accounting for over 60 per cent of the UK residential broadband market. Similarly, the most recent FCC report presents performance results for 14 ISPs (including one satellite provider) which serve over 80 per cent of the US broadband market. In Australia, the four largest ISPs accounted for around 90 per cent of subscriber share in the provision of fixed-line broadband services as at December 2012. However, the ԭ recognises the benefits to competition, and ultimately to consumers, in testing the services of ISPs beyond the very largest. The ԭ is therefore minded to consider the inclusion of any ISPs for which a sufficient sample of services can be obtained. One approach could be to monitor ISPs with the largest number of services in operation for each technology; for example, the five largest NBN-based satellite ISPs in Australia. Similarly, ISPs could be chosen by region; for example, the five largest ISPs in rural and regional Victoria. Having regard to the above, the ԭ is seeking views from stakeholders on the appropriate approach to determine which ISPs should be monitored as part of the proposed program. Questions on which the ԭ is seeking views: How should the ԭ determine which ISPs to monitor for ADSL and NBN-based services? For example: Should the ԭ monitor the largest ISPs by total market share in the Australian fixed-line broadband market? Should the ԭ monitor the largest ISPs by market share for each technology? Should the ԭ monitor the largest ISPs by market share for each region? If you consider that another approach to determining which ISPs to monitor is preferable, what is it and why do you prefer that approach? Should the ԭ monitor all providers of HFC in Australia, or limit testing to the two major networks operated by Telstra and Optus?Speed tiers There are a number of speed tiers for each of the different services provided by ISPs over the NBN. NBN fibre customers currently have access to maximum download and upload speeds ranging between 12/1Mbps and 100/40Mbps. NBN Co is planning to offer two speed tiers for fixed wireless services (12/1Mbps and 25/5Mbps) and while satellite customers on the interim satellite service currently have access to speeds up to 6/1Mbps, this will increase to 12/1Mbps on the long term satellite service. Both Telstras and Optus HFC services are capable of delivering speeds of around 100Mbps. However, the ԭ understands that 30Mbps services may still be available on some plans and in particular areas. On the other hand, iiNet offers four HFC speed tiers ranging between 12/1Mbps and 100/8Mbps. At this stage, the ԭ is unlikely to monitor every NBN or HFC speed tier. This would increase the overall costs of the project and may not be particularly useful in the early stages of the NBN rollout if there are insufficient customers acquiring a particular service to be able to obtain a representative sample. The ԭ is therefore considering which speed tiers would provide consumers and industry with the most useful guidance on the overall performance of an ISP or particular technology. One option may be to monitor the most popular or entry-level speed tiers with a view to adding more speed tiers as the rollout progresses. However, monitoring the highest available speed tier may give the best indication of the performance of the broader networkas it will be more visible, particularly during peak times, if an ISPs network is congested. There are two ADSL service offerings in Australiacommonly referred to as ADSL1 and ADSL2+. ADSL1 and ADSL2+ services are capable of delivering broadband services of up to 8Mbps and 24Mbps respectively (depending on factors such as copper length and quality). However, these peak speeds are rarely achieved by end-users in the ordinary day-to-day use of their service. The majority of services provided in metropolitan areas today use ADSL2+ technology. However, monitoring these services alone may fail to capture the experiences of consumers in rural and regional areas where only 64 per cent of exchanges are ADSL2+ enabled (with the remainder of services provided over ADSL1). Accordingly, the ԭ considers the testing program should also monitor ADSL1 services. The below table summarises the speed tiers currently available over broadband networks in Australia (as outlined above): Network technologyAdvertised speed tiersADSLUp to 8Mbps for ADSL1 and up to 24Mbps for ADSL2+HFCBetween 12/1Mbps and 100/8MbpsNBN FibreBetween 12/1Mbps and 100/40MbpsNBN Fixed wirelessUp to 12/1Mbps and 25/5MbpsNBN Satellite Up to 12/1Mbps (6/1Mbps for interim service)It should be noted that, whilst ADSL services are capable of high access line speeds, the data transfer rates experienced by end-users are significantly affected by distances from the exchange to the end-users premises. Ofcom normalises data in relation to ADSL customers based on distance from the exchange to ensure that there are no biases between ISPs with different distance profiles. Ofcom has, however, used the raw data to provide information to the public on the variation of speeds depending on the distance from the exchange to the end-users premises. Ofcom currently separates services into three speed tiers: services which have a headline speed ranging between 2 and 10Mbps, services which have a headline speed ranging between 10 and 30Mbps and services with headline speeds above 30Mbps. Ofcom includes ADSL2+ services in the 10 to 30Mbps speed range. These groupings are useful in that they allow different service technologies which have similar headline speeds to be compared on the basis of actual performance. The FCC has adopted a similar approach but with lower speed ranges (up to 3Mbps, 3 to 10Mbps, and above 10Mbps). Another benefit of this approach is that it would allow the ԭ to monitor ISPs which have chosen to offer different headline speeds over their network(s) as compared to other providers (for example, in the case of iiNets HFC network). The ԭ has not come to a view on which approach would be more appropriate in the Australian context and is seeking views on this issue. Specifically, whether to measure specific speed tiers or to identify several speed ranges (e.g. up to 12Mbps, between 13 and 30Mbps, between 31 and 100Mbps) as has been adopted by the FCC and Ofcom. Questions on which the ԭ is seeking views: Do you agree the ԭ should test both ADSL 1 and ADSL2+ services? Should the ԭ test specific speed tiers for HFC and NBN-based services or should it test services falling within particular speed ranges? Please explain if and why you prefer a particular approach.Sample size Having regard to the discussion regarding each of the parameters above, and the need to balance the benefits of obtaining particular information with the associated costs for obtaining it, the ԭ is seeking views on the importance of each of the identified parameters when determining the scope of the program. For example, is it more important to include additional regions in the study or to have more probes per region? Is it more important to test a large number of ISPs or to have more probes per ISP? An example of how each of the different parameters can have an effect on the number of distributed probes and, accordingly, the cost (and viability) of implementing the program is outlined in the table below. For this example, the ԭ has chosen between 10-15 probes per sample set. However, the ԭ is seeking views, particularly from industry, on this issue. For example, how many probes would be required to represent an average consumers experience when acquiring a specific ISPs NBN-based fibre package with a headline speed of 12/1Mbps in the city of Melbourne? The total number of probes outlined in the table below represents the minimum number of participants which would be required by the ԭ to acquire the specified sample example. Similarly, the number of services in the table includes both the relevant technology and speed tier (e.g. one service could be a 100Mbps HFC service). No. of ISPsNo. of cities and/or regionsNo. of servicesNo. of probes per sample setTotal no. of probes5102101,000 probes7154104,200 probes102041512,000 probesBy way of contrast, Ofcoms report from November 2012 provides the results for just over 2,000 participants, which were drawn from a pool of over 40,000 volunteers. The participants were chosen in order to ensure that the results are representative of the UK broadband population. Sample quotas were set by geographic market definition, market shares (by ISP and product), and distance from the exchange. The FCC recruited approximately 10,000 participants from over 75,000 volunteers in the US with results from around 6,700 participants reported in its February 2013 report. Samples were organised with a goal of covering the major ISPs, each of the relevant broadband technologies, the four Census regions (Northeast, Midwest, South and West) and the three different speed ranges. The ԭ is also seeking views from potential testing companies on the importance of particular parameters for obtaining a representative sample and for the viability (both technically and financially) of the proposed program. Questions on which the ԭ is seeking views: What is the minimum number of probes which would be required to provide robust results on the broadband performance likely to be experienced by consumers acquiring a particular ISP package or offering in a particular region (i.e. per sample set)? Which of the variables (ISP, geographic region, speed tier or size of each sample set) is most important and why?Metrics Data transfer rates or throughput speeds The key metric for any broadband performance monitoring and reporting program in Australia would be downstream and upstream data transfer rates or speeds. That is, the rate at which data is transferred (i.e. downloaded or uploaded) over a particular connection. When testing throughput, the FCC considers both sustained and burst speeds. Sustained speeds refer to the data transfer rates averaged over a period of several seconds and burst speeds refer to the effect of techniques used by cable-based ISPs to temporarily increase data transfer rates. The FCC monitors and reports on average sustained peak throughput between 7pm and 11pm on weeknights. Ofcom reports on average speeds over a 24 hour period, maximum or off-peak speeds (which are measured between the hours of 12am and 6am) and peak speeds (measured between 8pm and 10pm). Both the FCC and Ofcom compare ISP performance for each of the above against the relevant ISPs advertised headline speed (usually as a percentage). The ԭ would likely seek information on performance of services during peak periods as this is when most consumers use their services. However, it may also seek information on off-peak performance to ascertain the level of service degradation during peak periods. Questions on which the ԭ is seeking views: What information regarding download and upload data transfer rates (or speeds) would be most useful for ISPs and for consumers? In particular: Do you agree that the ԭ should monitor both peak and off-peak data transfer rates? What is the daily peak or busy period for demand on broadband bandwidth in Australia? To what extent are burst speeds available for consumers in Australia and should they be accounted for in the ԭs proposed testing program?Additional quality of service metrics While data transfer rates are an important factor, there are additional factors which affect the overall performance of broadband services and the end-users overall broadband experience. The ԭ considers that a robust monitoring program could include testing of (but need not be limited to) the following additional metrics: Packet loss: The proportion of data packets lost during the transmission of data across a network. Latency: The time it takes for a packet of data to travel to a designated point in the network and back (often referred to as round-trip latency). Jitter: The rate of change in latency. This metric relates to the stability of a connection. Factors which affect end-user web browsing experiences: Webpage browsing speed: The time taken to load a webpage (i.e. to fetch the main HTML and assets such as text, basic code and content). Domain name service (DNS) resolution: The time taken for an ISP to translate website names (such as website.com) into IP addresses (e.g. 123.255.255.00). DNS failure: The proportion of requests for which the DNS service cannot translate a domain name to an IP address. Ofcom monitors each of the above metrics and compares the results on an average and on-peak basis. In addition to data transfer rates, the NZCC monitors latency and webpage browsing speed. The FCC also monitors latency. The IDA obtains information relating to on-net peak throughput speeds, webpage download speeds for local and international (US) websites, latency and packet loss. The ԭ notes that some over-the-top content service providers (such as internet protocol television providers) may be affected by additional factors relating to service quality which have not been listed above. In this regard, the ԭ is interested in what these factors are and whether they can and should be tested by any program implemented by the ԭ. The ԭ is seeking views from all relevant stakeholders on the quality of service parameters which should be monitored by any testing of broadband performance in addition to data transfer rates. Questions on which the ԭ is seeking views: What additional quality of service parameters should the ԭ monitor so as to obtain rich and meaningful information regarding the performance of broadband services in Australia? In your response, please state each factor which you consider should be tested and why.Reporting The ԭ considers that any reporting on broadband performance should focus on empowering consumers to make informed decisions. It should also provide ISPs with transparency over competitor offerings to be able to compete on quality and make decisions on investment in capacity. As previously stated, the ԭ is committed to engaging closely with industry on any issues which may affect the reliability of reporting. For example, the ԭ recognises the need to ensure that there are procedures in place to take into account any non-ongoing network issues (such as natural disasters) which may affect performance of services at a particular time and/or place. In addition, the ԭ may request additional information from providers on a confidential basis to clarify certain performance aspects and would likely provide a copy of the report to relevant ISPs prior to its release. Ofcom and the FCC publish the results of their testing on a six monthly and yearly basis respectively. The reports are in narrative form, containing observations and commentary regarding trends in the performance of ISPs and across the industry. They are also accompanied by illustrative charts. For example, the following graph was provided in Ofcoms recent report from November 2012 to demonstrate the differences in performance between particular ADSL ISP packages during peak, off-peak (maximum) and as an average over a 24 hour period.  Both the FCC and Ofcom choose a particular month in the period and report extensively on those results, rather than averaging across the entire year. However, these results are often compared to results from previous testing periods. On the other hand, the IDA publishes data on a monthly basis through interactive graphs on its website. The IDA does not appear to provide any analysis or observations on the results, though it may choose to do so in the future. The NZCC has not published any reports to date. However, its contractor TrueNet releases a report each month on broadband performance using graphs and providing high-level commentary (including any noteworthy improvements or declines in performance of particular ISPs). There is necessarily a trade-off between the level of detail reported and the frequency with which data can be made public. The ԭ could, for example, report monthly, quarterly, bi-annually or annually. The level of detail in the reports would, however, decrease the more regularly that the results were released. The ԭ has preference for a reporting framework which allows it to provide a medium to high level of detailed commentary such as those provided in the bi-annual Ofcom and annual FCC reports. However, a summary of certain performance results could also be released more often (e.g. monthly or quarterly). The ԭ is seeking views on what would be the best approach in the Australian context. The ԭ also recognises that there is a trade-off between the currency of the data that can be reported on and the level of detail that can go into any reported analysis. Consequently, consumers may be accessing out-of-date information when performance results are not reported in real time. Furthermore, reporting on the metrics discussed in this consultation paper will not capture every aspect of the end-user broadband experience. However, the information provided by the proposed program would substantially improve the current position of consumers and industry who currently have little visibility over the real-world performance of broadband services. The ԭ is also considering how it will treat the raw data collected by the program. For example, in addition to releasing the public reports, the FCC provides raw data from testing directly to the public for analysis. Another option would be to make the raw data available solely to participating ISPs. The ԭ is therefore seeking views from industry, as well as any other interested stakeholders, on whether the ԭ should make the raw data publicly available and to whom. Any third party testing company which is ultimately contracted to conduct the testing may also have an interest in retaining access and control over the raw data, and may therefore wish to submit views on how the ԭs treatment of this data could affect the viability of the program. Questions on which the ԭ is seeking views: What do you consider is the best approach to reporting on broadband performance in Australia? In particular: How often should the ԭ report on the results of its broadband performance testing? Do you agree that the ԭ should provide detailed observations, commentary or analysis on the results of testing? To what extent would industry (e.g. ISPs) value access to the raw data collected by any testing program and want to have access to it? Summary of questions Questions on which the ԭ is seeking views: Testing methodology Do you agree that a probe-based testing methodology would be the most reliable and accurate approach for the Australian context? If you consider an alternative approach preferable, what approach do you prefer and why? Services What services should be included in the ԭs proposed performance monitoring and reporting program? In particular: Do you agree that the ԭ should monitor ADSL, HFC and NBN-based broadband services? Do you agree that the ԭ should monitor small business broadband services? Are there any other services which you consider should be included in the proposed program? In your response, please outline reasons. Regions How should the ԭ determine which regions to monitor as part of any program? In particular: How many Australian cities do you consider should be monitored as part of the proposed program? How could these be determined by the ԭ? Would you consider State or Territory regions which encompass rural and regional areas outside of each major city would be sufficient to provide information to consumers living in these areas on the performance of broadband services? For example, a Victorian rural/regional delineation which encompasses services outside of metropolitan Melbourne. Internet service providers How should the ԭ determine which ISPs to monitor for ADSL and NBN-based services? For example: Should the ԭ monitor the largest ISPs by total market share in the Australian fixed-line broadband market? Should the ԭ monitor the largest ISPs by market share for each technology? Should the ԭ monitor the largest ISPs by market share for each region? If you consider that another approach to determining which ISPs to monitor is preferable, what is it and why do you prefer that approach? Should the ԭ monitor all providers of HFC in Australia, or limit testing to the two major networks operated by Telstra and Optus? Speed tiers Do you agree the ԭ should test both ADSL 1 and ADSL2+ services? Should the ԭ test specific speed tiers for HFC and NBN-based services or should it test services falling within particular speed ranges? Please explain if and why you prefer a particular approach. Sample size What is the minimum number of probes which would be required to provide robust results on the broadband performance likely to be experienced by consumers acquiring a particular ISP package or offering in a particular region (i.e. per sample set)? Which of the variables (ISP, geographic region, speed tier or size of each sample set) is most important and why? Metrics What information regarding download and upload data transfer rates (or speeds) would be most useful for ISPs and for consumers? In particular: Do you agree that the ԭ should monitor both peak and off-peak data transfer rates? What is the daily peak or busy period for demand on broadband bandwidth in Australia? To what extent are burst speeds available for consumers in Australia and should they be accounted for in the ԭs proposed testing program? What additional quality of service parameters should the ԭ monitor so as to obtain rich and meaningful information regarding the performance of broadband services in Australia? In your response, please state each factor which you consider should be tested and why. Reporting What do you consider is the best approach to reporting on broadband performance in Australia? In particular: How often should the ԭ report on the results of its broadband performance testing? Do you agree that the ԭ should provide detailed observations, commentary or analysis on the results of testing? To what extent would industry (e.g. ISPs) value access to the raw data collected by any testing program and want to have access to it? Attachment A United StatesUnited KingdomSingaporeNew ZealandServices testedADSL, HFC, fibre (FTTH) and satelliteADSL, HFC and fibre (FTTC)ADSL, HFC and fibre (FTTH)ADSL, VDSL, cable and fibre (FTTH)Regions testedNational level (the four Census Regions - all states across the US)Region level (three geographic markets across the UK)National levelNational levelNumber of ISPs tested14 largest ISPs in the market, representing over 80% of the US residential broadband market8 largest ISPs in the market, representing over 60% of the UK residential broadband market3 ISPs20 ISPsSpeed tiers testedUp to 3Mbps Between 3 and 10Mbps Above 10MbpsBetween 2 and 10Mbps Between 10 and 30Mbps Above 30MbpsAll of the published plans for each of the 3 ISPsInformation not availableSample size~6,700 probes~2,000 probes~900 probes~400 probes  Australian Bureau of Statistics, 8153.0 Internet activity, Australia, December 2012.  DBCDE, Statement of Expectations, 17 December 2010 ( HYPERLINK "http://www.dbcde.gov.au/__data/assets/pdf_file/0003/132069/Statement_of_Expectations.pdf" http://www.dbcde.gov.au/__data/assets/pdf_file/0003/132069/Statement_of_Expectations.pdf). See also:  HYPERLINK "http://www.nbnco.com.au/about-us/our-purpose.html" http://www.nbnco.com.au/about-us/our-purpose.html.  Schedule 2 of the Competition and Consumer Act 2010.  The specific technology used for providing these services is generally VDSL.  Ofcom does not monitor fibre-to-the-home (FTTH) or satellite services as these services represent only a small proportion of the total broadband market in the UK ( HYPERLINK "http://stakeholders.ofcom.org.uk/market-data-research/other/telecoms-research/broadband-speeds/" http://stakeholders.ofcom.org.uk/market-data-research/other/telecoms-research/broadband-speeds/).   HYPERLINK "http://consumers.ofcom.org.uk/2011/07/consumers-benefit-from-uk-broadband-speed-surge/" http://consumers.ofcom.org.uk/2011/07/consumers-benefit-from-uk-broadband-speed-surge/   HYPERLINK "http://media.ofcom.org.uk/2013/03/14/average-uk-broadband-speeds-hit-double-figures/" http://media.ofcom.org.uk/2013/03/14/average-uk-broadband-speeds-hit-double-figures/   HYPERLINK "http://www.fcc.gov/measuring-broadband-america" http://www.fcc.gov/measuring-broadband-america  The FCC obtains the data for ADSL, HFC, fixed wireless, satellite and fibre services. To date, the FCC has not reported on fixed wireless due to a lack of volunteers acquiring these services.  FCC, 2012 Measuring Broadband America: July Report A report on consumer wireline broadband performance in the US, July 2012, page 49.   HYPERLINK "http://www.ida.gov.sg/applications/rbs/chart.html" http://www.ida.gov.sg/applications/rbs/chart.html   HYPERLINK "http://www.comcom.govt.nz/media-releases/detail/2012/commerce-commission-selects-truenet-to-test-broadband-quality" http://www.comcom.govt.nz/media-releases/detail/2012/commerce-commission-selects-truenet-to-test-broadband-quality  In the illustrated example the HMU includes wireless router functionality. This is not mandatoryif an end-user wanted to continue using their own equipment the HMU could sit in between the modem and the end-users existing wireless router.  In the US, program participants with data quota restrictions can have their tests set to run less frequently.  Ofcom, UK fixed-line broadband performance, November 2012: Research Report, page 51 (available at  HYPERLINK "http://stakeholders.ofcom.org.uk/market-data-research/other/telecoms-research/broadband-speeds/broadband-speeds-nov2012/" http://stakeholders.ofcom.org.uk/market-data-research/other/telecoms-research/broadband-speeds/broadband-speeds-nov2012/).  In the US, volunteers are allowed to keep their testing device after the program has completed as a means of encouraging volunteers to participate. The volunteers are also provided with performance monitoring software for their own connection. Similar schemes are used in the UK and New Zealand to attract participants.  Approximately 98 per cent in December 2012 (see: Australian Bureau of Statistics, 8153.0 Internet activity, Australia, December 2012).  Australian Bureau of Statistics, 8153.0 Internet activity, Australia, December 2012.  Australian Bureau of Statistics, 8153.0 Internet activity, Australia, December 2012.  Australian Bureau of Statistics, 8153.0 Internet activity, Australia, December 2012.  FCC, 2012 Measuring Broadband America: July Report A report on consumer wireline broadband performance in the US, July 2012, page 9.  FCC, 2013 Measuring Broadband America: February Report A report on consumer wireline broadband performance in the US, February 2013, page 33 and  HYPERLINK "http://www.fcc.gov/measuring-broadband-america/2013/February" http://www.fcc.gov/measuring-broadband-america/2013/February. Consumers who volunteer for this program will download an application onto their smartphones, which will serve the same function as fixed-line HMUs.   HYPERLINK "http://media.ofcom.org.uk/2013/03/14/average-uk-broadband-speeds-hit-double-figures/" http://media.ofcom.org.uk/2013/03/14/average-uk-broadband-speeds-hit-double-figures/  Australian Bureau of Statistics, 8153.0 Internet activity, Australia, December 2012.  Ofcom, UK fixed-line broadband performance, November 2012: Research Report, page 54.  Source: publicly reported subscriber numbers, data obtained under the ԭs CAN and Infrastructure Record Keeping Rules. The iiNet Group is considered a single ISP for this purpose.  NBN Co, Product and Pricing Overview for Service Providers, December 2011, page 19.  Telstra Wholesale list of internet grade ADSL enabled exchanges (available at  HYPERLINK "http://telstrawholesale.com.au/products/broadband/adsl/adsl-reports-plans/index.htm" http://telstrawholesale.com.au/products/broadband/adsl/adsl-reports-plans/index.htm)  Ofcom, UK fixed-line broadband performance, November 2012: Research Report, page 13.  Ofcom, UK fixed-line broadband performance, November 2012: Research Report, Figure 1.1, page 6.  Ofcom, UK fixed-line broadband performance, November 2012: Research Report, page 53.  FCC, 2013 Measuring Broadband America: February 2013 Technical Appendix, July 2012, page 5.  Ofcom, UK fixed-line broadband performance, November 2012: Research Report, page 13.   HYPERLINK "http://www.ida.gov.sg/applications/rbs/chart.html" http://www.ida.gov.sg/applications/rbs/chart.html.   HYPERLINK "https://www.truenet.co.nz/news/Performance" https://www.truenet.co.nz/news/Performance  The NZCC only reports on ISPs where there are 5 or more probes working during any particular month (this amounts to approximately 7-8 ISPs according to the monthly TrueNet reports available at  HYPERLINK "https://www.truenet.co.nz/news/Performance" https://www.truenet.co.nz/news/Performance)     Broadband performance monitoring and reporting in the Australian contextAugust 2013 ii  PAGE \* MERGEFORMAT 2 Broadband performance monitoring and reporting in the Australian contextAugust 2013  PAGE \* MERGEFORMAT 1 Broadband performance monitoring and reporting in the Australian contextAugust 2013 4KLNO`akopqrz{|ž|ukfkYPGPhImHnHuh hI0Jjh hI0JU hO25jhO25U hp ;h. hRdh.hkqh&5#h>m5C$ϴ$#khrq h>hr hrqh>h16h>h 6h> ? @ A B   # $ % & ' ( ) * + G H 鯠~鯠l#jXhIUmHnHujXh hI0JU#jWhIUmHnHujhIUmHnHuhImHnHu*hIhI5CJOJQJaJmHnHujWh hI0JUhImHnHuh hI0Jjh hI0JU(H I J M N a b c } ~  ٺِـٺnِ#jZhIUmHnHujZh hI0JUhImHnHu#jYhIUmHnHujhIUmHnHuhImHnHu'hIhICJOJQJaJmHnHuh hI0Jjh hI0JUjYh hI0JU$        ! = > ? @ C D M N O i j k m n o p q r ٺِـٺnِ#js\hIUmHnHuj[h hI0JUhImHnHu#jy[hIUmHnHujhIUmHnHuhImHnHu'hIhICJOJQJaJmHnHuh hI0Jjh hI0JUjZh hI0JU$           ٸَ~ٸlb^ZhU hJ(jhO25U#jg^hIUmHnHuj]h hI0JUhImHnHu#jm]hIUmHnHujhIUmHnHuhImHnHu*hIhI5CJOJQJaJmHnHuh hI0Jjh hI0JUj\h hI0JU!  * b=g*mjHl !"gd.4gdgd#gdggd5 gd#gd%gdgd7i"gdS[gdU  ! ) * 5 6 < A B W QTvw'5@!#FGh%hVpj^hUjhU h6Mh hhhK6thh3hIqI hzhohg hzh=hz hzhz hzh7i" h h hhxhO26?b;Y]gY`bc*/139:IRVWྺܚhZhIh.Ah=hahzNSjh7i"h7i"0J&U h7i"h7i"h7i"ht30hgh\hBAy hzh hzh3hVphh%hh{ja_h3Uh3jh3U4m .LPVW~z{ijorw~  0FH3?FKL_cp¾֢ h|h.4h.4hIqIhh3hYK)h%hhK^hgh5 hhtThhK6thlMhI=hBWh.Ajha0J&UhZhIhzNS=pvw| + !!!(!)!!!!!!!u"""""""k########($)$\$$$%%%\%%%%%&&'''+(C(ʹֵ饭ơh0nhUz9hv*h3='hKh@{UhP+jhK^hU0J&UhphUhP!<h!h8,hhHv hVHhhhh.4hK6t h|h.4>"\$w$%'T)--?-.}//e01 464h567D88B999gdj=gdigddgd: 7^7`gd gd8,gd8,gd#gdC(P((())Q)R)S))V*f*********#+$+%++++3,5,,,,,---.-4-?-_-`------.%.1.p........// //-/=/M/{/|/}//Žhphfih~^- hQThQThQThhZdhP+h@{Uhn~h h]@h4Ih hh}hcJ h0nh0nh0nhihh+@//////////00001$1+181B1F1N1h1u1|1111111122#2F2_2k2v2222233,393@3]3}34 4!45464[4f4r4444445f5¾ºh:h h:h: h@ʵCFFHHH3J< & F gd=fB & F ^`gdzgd8{gdS[gdWgd{gd@S 7^7`gdlgd>>>>>>>>>>???)?*?+?OLOXOaOOOhPoPrPsPvPPPhK hhMkhMk h :h hhhYhph~;hUhK6t hh hhhh$ huih`h43h` h2.h`hROhz h2.hBWhI h2.h3 h2.hMk1JLLMOQQ:SUXhIg hh jhzU0J&UhhK6thhzUhbbhaFchy'jh0J&Uh(hURh<%l0l1l2lhs5 hp5hH>R5s0&խ6sR]<h-h!hThYyh hhzUh{|mjhbb0J&UhaFchbbh59oooooooo8pGppppppp,qhR5h3='hR51ooo( $IfgdiWkd$$Ifl\ l!vvvv  t(0644 lagp(ytXIJooooooo:kdo$$Ifl\ l!vvvv t0644 lagp(ytXIJ $Ifgdoo-p2p3p8p:1 $IfgdiWkd$$Ifl\ l!vvvv t0644 lagp(ytXIJ $Ifgd8pHppppp90 $IfgdiWkd$$Ifl4\ l!vvv`v t0644 lagp(ytXIJ $Ifgdpppppp90 $IfgdiWkdu$$Ifl4\ l!vvv v t0644 lagp(ytXIJ $Ifgdpppppr:3gdYK)kd'$$Ifl\ l!vvvv t0644 lagp(ytXIJ $Ifgdruwxs{ }<~j~~5 & F $IfgdiW & F $Ifgd=fB & F$Ifgd5 $IfgdmgdYK)gdylgd~R uuuuuuvv0v1vHvLvNvRv}vvvvvvvvvvww!w-w2w3w8w^wbwgwpwqw}wwwwwwwwwwwx'x(x,x>xFx`xxxx%yEyJyUyhyiy溶޺hjhh& hC>hXohXoh~BhIhURh}mh_hK hK6th2hy'hRhaFch]nh~Rh%jh_0J&Uh)%h(>yy"'2hXo hC>haFch}mhaFcho^Fh2hj h?h?h&hXoh?hLh(hR\epvͅ΅ۅ2>ST{Ȇ123:HIQRV!"PYZaչյչ h2.h:h~ h{7hkhq h{7h h6jhc&0J&U hjh h h{7h4O0h#h*}ih h2 hPhPhPhMk@ v{̊ϊMNgorwۋcltŌƌ̌͌݌ čύüҩ hFuh h*h:hMPh|h oh*}ih2 hh/ hO4Ph/hO4Ph:5 h{7h:h4O0h1rh6h{7hP h{7hPhMkhL h2.h:h:8NIDgd gdmkkdw$$Ifll!!  t 0644 l` ap yt: & F$Ifgd| & F$Ifgd5 & F$Ifgd5 $Ifgd:5DHQ^q͎юÏُ̏ۏHhij-AERWXg!ŽŽŽŹŽůūh-V*hK h hMPhhjhrL80J&UhK6thVhrL8hJYho h6hoh:h4bh)0h1rhkhh3='hzUhz}hFuhG=jec & F$Ifgd  & F$Ifgd5 & F$Ifgd5 $Ifgd:gdYK)gdogd  Փ !'2DENOPfnopqє$/CcdeƕוݕFQ¾ܶ𧠧غؘh h__ hh__ hh:hc$ h:5hGh2h6hoh4b h6h: h6hz}hEoh1rh!hz}hdWbh~h:hFuhkhJY: Tw|آ̘Ԙ`™ș $.HNbњҚӚ^_e}ӛԛۛSkx亶Ⱥ̺hNVhS)h)0h#fhYhh 3h1rjh%0J&UhlMhh0hjhMkhFuh*h:h4bhYhK6hkhEohq> 0.~ytoojjaa $Ifgdsgd>4gdgd:gdFukd$$Ifll!!  t 0644 l` ap yt: ʜٜ "# +>EfjlĞӞ4:>?Eaqɟ̟П12cƿܿƿܴܰܰܰܰhYK) hEoh* hEohxJ hEohEo hEohhS)h L6 hEoh>4 hEoho hEohh>4 h>4hoh4O0h#h hkhhheI5h1rhNVh8cqrx|Ġ012IR-.ɢϢ#BYkpy!<LNWXԼԼԸh|<jh. 0J&Uh)0h2.h jhC0J&U hhCh L6 h>4hChCh hH>hl5h3='hl5hWhlhEoh!hjh4O00J&Uh#hYK) hEohEo hEoh3./4fd[[ $Ifgdskd$$Ifl0l!  t0644 lagpytXIJfgkxoo $Ifgdskdw$$Ifl0l! t0644 lapytH>xof $Ifgds $IfgdiWkd$$Ifl0l! t0644 lapytH>ɢxof $Ifgds $IfgdiWkd$$Ifl0l! t0644 lapytH>"xof $Ifgds $IfgdiWkd$$Ifl0l! t0644 lapytH>"#Yߩ Pxsng^QD & F$Ifgd/5 & F$Ifgd5 $Ifgd|<gdYK)gd2.gdskd$$Ifl0l! t0644 lapytH>178QT\drީߩ :fkqv%ӭ֭"$'a HT_yGHžɺɺɳɳɳɫɫɫɫثɫɺh-Eh h*-hKhLw hOhBh) hhChChB hjh|<h/5h! h|<5hc$ h|<5 hEohhjh2.h|<h4O0>%#_Ӱ~yttmdd[dd $Ifgd0n $IfgdBgd{gdBgdBkd3$$Ifll!!  t 0644 l` ap yt|< Ht˰Ӱڰ -./12346789:CDEGHJKLMOPV]:>ӲO./TZr/:окжЮкЪh*-h)hc&h hLwh)0jhc&0J&UhKhBhC hhB h)5 hB5hhB5 hOh)h-Eh)C kd$$IflrX l!  t20644 lap2ytB, $IfgdB,-/24.%%% $IfgdBkd$$IflrX l! t0644 lap2ytB47DEH%kd$$IflrX l! t0644 lap2ytB $IfgdBHKMP^ $IfgdK $IfgdB^_r.'"gdBgd{kd`$$IflrX l! t0644 lap2ytBrUz LGBgd) gd`kd$$Ifll!!  t 0644 l` ap ytB & F$If^`gdj & F$Ifgd{ $IfgdBgdYK):Uݵ'bxyz /2ȷ '(8Koͺκ=>̼μ#ȼиидХhh@ h) 5hc$ h) 5 huih) h)0hh<4)h!hyhohVhtTho>h) hNhh=h{h4O0h)hBhKhc$ hB5h*-8 (>ϼ%} & F$Ifgdv> & F$Ifgd5 & F$Ifgd5 $IfgdogdYK)gd) gdo>#*-17CD|}ý 9:Mmq<>BGXY|߿Frt{ !"%)*+,8RS亲ũŢܮܮŢhF hoho> hi:+6hi:+h"hi:+6 hXhXh"h) ho>6 heI5ho>h"ho>6h-Eh)hXho>h) hW6hv>h/h@= 4|߿t ,~ytllllddd & F gd=fB & F gd=fBgdtTgd) kd$$Ifll!!  t 0644 l` ap yto $[{dgy8@EHIIJKxy~,%&ؼسԳhE hROh*-hx-huih ht30h:h)htT htT5hc$ htT5h-hW6hghThhBh) h! hW6hW6 hW6hi:+ hW6h) hoho>hj8Ky\WRRgdt30gd:kdi$$Ifll!!  t 0644 l` ap yto> & F$IfgdYK) $Ifgdo>gdYK)gd) * !"# tu-Dm05>9>@줬֤hYh-Ehyhx-hA"h"jhx-0J&U h,h~.h) jh[ hmUmHnHuh,jh~.0J&Uh~.ht30h\hyhMhghT<!#@`@^PJ & F$Ifgd5 & F$Ifgd5 $IfgdpgdYK)gdE gdt30@[q+w67_`;<S@_]^OPakrzh7nhA"h"hh*hc$ hp5hT h,hE hE hK6th0ZLhQhjhh'Nh jh|hIh hhROhgh,hyhY<*HIJKLMab$&}~QS$&.RYps|rsALRSÿÿÿÿDZǃhgh  hhUhUhfshU6hMPhdWbhkP hhkP hO4PhkP h.hV:*h.hV:*hfshfs6hm2hp5 hm2ho hm2hdlhMhS[hp h$QIhA"h8uhA"2JKLb&|wnaTT & F$Ifgd. h$If^hgdfs $Ifgdm2gdS[gdS[~kdd$$Ifll!!  t 0644 l` ap ytc$ S&.s^T & F"$IfgdUh$If^hgdU & F!$IfgdMP & F!$Ifgd. & F$Ifgd. & F$Ifgd.h$If^hgdfsST(+,jmuvBC<=>@M !hAzPJaJh PJaJhiWhAzPJaJhZhAz5PJaJhiWht aJhiWhAz5PJaJ hiWhAzhAz hm2hp hm2h.hYK)h7hhdihfshfs6h.hg h5h.4T(,jvC= & F $Ifgd. & F$Ifgd!1 & F$Ifgd.h$If^hgdfs & F$Ifgd.=>?MN\ku~yi]TTTTOFf $IfgdZ $$Ifa$gdAz & FP^`PgdZ/gd8kd$$Ifll!!  t 0644 l` ap ytc$  $IfgdiW $IfgdZ $Ifgdc& $Ifgd $IfgdAz!0e6-$$ $IfgdiW $IfgdAzkd$$Iflrl }l!. / 3   t 0644 lagp ytXIJer-kd$$Iflrl }l!. / 3   t 0644 lagp ytXIJ $IfgdiW!]^jknpx~-.:찦 hahjjhjUhjjhj0J&UhZheCJaJh PJaJhiWhAzaJ ht aJjh0J&PJUaJhZhAz5PJaJhiWht aJhiWhAzPJaJ7, $IfgdiW $IfgdAz6- & F$IfgdiW $IfgdAzkds$$Iflrl }l!. / 3   t 0644 lagp ytXIJC] $IfgdiW & F$IfgdiW]^jx6-$ $Ifgd $IfgdAzkdI$$Iflrl }l!. / 3   t 0644 lagp ytXIJx-kd$$Iflrl }l!. / 3   t 0644 lagp ytXIJ $IfgdiWpm-Yx$ <^`hj0JmH nH u5&()WXYZ[ , xyzlht1bhj0JmH nH u hhjh=hhj0JmH nH u hi)hjjhj0J&Uhbc_hj0J&CJH*hbc_0J&CJH*hbc_hbc_0J&CJH*hbc_hbc_hbc_0J&CJjhbc_hj0J&CJUhhj0JmH nH ujhjU hshhjhj*cd"#$QTU[\;<BGdvXY[w{ºɯp#hphj0JCJaJmH nH uhphjCJaJjhphjCJUaJhO DhjCJaJhjCJaJhjhjCJaJh,hj6 hjhjjhj0J&U h"|hj hxJhjjhxJhj0J&U hj0JjhjUhj,#[;[js,:r^`rgdYK)$ J^`JgdYK)m$Irx[$\$^`rgdYK)$gdYK)m$$r^`rgdYK)m$SUV[\jkpqst,-:;<=Iz|}׼ק靗h"hjCJ hjCJjhjCJUhhj0JmH nH u h*.hj hj0JheQhj0JmH nH u h7nhjjhjUjhj0J&UhjhjCJaJhjhjCJaJ8CDFGHIKLNOQRTUVWXYZ[\]^_`bcdȿے{shI0JCJ hkqhjh$Vh3jh{9>Uh{9>hPhj0JPJmH nH uhhjPJjhjPJU hjPJhiWhjPJhZhjPJjhj0J&Uhj hjCJht1bhj0JCJmH nH ujhjCJU-GHJKMNPQSTUVWXYZ[\]^_`abcdgdkqgd$gdYK)m$"#./012IJKLʽʽٹhZheCJaJh{9>hQ0JCJmHnHujhc@hI0JCJUhj hXIJhjhc@hICJhI0JCJhc@hI0JCJd/01/gd8gdXIJ !$dNgdI 6&P 1F:pt . A!"#$.% F 00P&P 1F:pt . A!"#$.% DpF00P&P 1F:pt . A!"#$.% Dp< 00&P 1F:pK6t. A!"#$.% ,SDdn  S <A?ACCC-logo-BW "`RjROEi\:%xRFRDF>ROEi\:%xRJFIFddDuckyAdobed   #%'%#//33//@@@@@@@@@@@@@@@&&0##0+.'''.+550055@@?@@@@@@@@@@@@" !1AQaq"2B#Rbr3ᢲCS$ғTs4et%56 ?*R fH2ǏX\R_4]putnTjl?tH(6Wc\4D[쏧ԦچimXwai{ϰ.=͵KvnC3: l)rS48zǭEȰD)}r~uoplL"-ˋ? m˸=xVŵ+:۬0mV;WYnij|BF#wG18(6Un?*,<焍n}xvU{imUlO7>9+7-R8;@8+:rc_r ?k:r6ح$rd rHWʴiRim:kFuPR-`l`AuBf\g[Z]퍐Ws8Lh|<|PM-k:G%%qk>҃3kf(~8Dv|0rV}ǫR,]5 K8tBnylFF5!>R D@DDD@DDD@DDD@DDD@DDD@DDE[-Wa%OlO]߇G-g|OܩkԒ>Y,.B\WG|R6X[$d9!9.xȭ ~vF.<]Tσ_>ID@DDD@DDD@DDD@DDD@DDD@DDD@DDD@DDD@DDD@DDD@DDD@E3UddE#mE+&FyG#C[\2 h wOV8;22}Nk~!&mY50GbHdt8GPY;GocSH@ȸKRijaůw7C8\>f+ͰDђ8Ua5<|Opܹmj+p |ɿlG;aإ<~O\#vO/ppidWD@DDD@DDD@DDD@DD dĮQݻøٸӯAka[{{4fBzƫAs̸c\9Aii9d}ZzUJ)GMAYc'+"=8 hIIe4 0\5%s:bc4o>;]ym(2zg=g(4EUS_fV%R?6g}etm@*G`cyK,. G͗ >\ݵI݉jPh `&}.L(tʖ6ri9|b/.|~1AKd6vWME8;,}p4$C`o+ڹnm;R8gAٽҽ[/sπ9!p#uN61 (O_x%kzȟzc##sA 3#%N? 2ddqjw^?D_SIk@A",B<{Z+;S;\=[=˞ ~vvcgD^Q@DDD@DDDAۏl0a\쎪ͽ;CmcZz}+G63nͭR .Wx18k9X;oͫZ؍9 80;ρ+崡6崥4F;;9Q־Shd̎ZC|~,pUG>(_H$3ǣ|5z6vn a՜׸y [ݙE &Üqz ;]X|7 &fyiù n$L^̏Ȝ.}CQ6q5e 8.ke+&ֿhiNKΖ Opf-iwsp։}CfhM汼 ÇQξ0K:pH\\'TxôuM@\"9A1ԂѿKuɸ1/t[Tɯ߆1oO_e[r}Nj{vuoCl@Hş.2#a\p?ηda3## \}X'V֒Գ'_r[>}F&Y̷ܹ韡TwO6ױj{$w}Wo>miF[⹔N׾9DYވco^cQFSvsKvA|`!+EavnRb/`xM:{yPDDW),n,9lo` t6cgaCǛxr!ɨu,i+ߍiD^W]i$Fl0ڎw#>zrq־W%MԯD.MİLzZ7lyt~\c,s n߸tΒ裖dsu'ׂ6w\ŭ~r<ꊡuyܭͣkEQt<=JU|s֗8֌|";v64k=%q%>$v򴟕1oG=T:" "" >mQt`}v*uohc=VI:󰙸pxJԢ Cvv#Y3uqy$#<o eDy.N$䃖)Dŏ6>Uմ?бr?ϯ˲T+ <3*CikGU%o᷏XlH^N$yø=GՉm+k^D Yvjl:TX/>ռ4<`]N*Ŷ29l̲8sCҖ {/%Pص&:!I/ihs]T+I͙c{ tc >V{N1|m],ak\^ַ᎕- {tf7Ԕ'ӌ;}l 3aV^G&vqFMl20e /D!qɘ=]cU,ثt$>Lq'uOa٣MF k#k[Š:FdK=I5zR^%܂7$td&28ȹynz6${YZCD>ȴ۩61,)Ռ>.<2ITګ4ժp_V)>v%D3;>J :˳rRZ r5i盾tLdNh-tQ6:"_mүeF]z-Csp~ >] eu >wkC['l8B7r^ __7MOZbJ/p`b6Se*k# HQW{ (:xd8azqd=pǼpS ym\Ekaq@Ѹc:93d@^kgd* Hcȱ޼ e{P䎣DI.0~k[l9m-;j7k~cA_WZS6}ԍբ s9zxu*nljBƽ? Z~Pn"" """ """ *~~VWwo{՛e[Jk#ǥWv}nX9gP4{]eZũD5|һiqĊ׮l?w?uJK5fcu̮!8(^' 9`Lbhݱ?X $C̏T.%@4O}%M.sKנ]0w~w4y.y6Nv~Z/î)`}zGZa]tanjMpwqy)ۈfX$K{ i;=Z?eݓS?1Koi:"(=gxh8cfyßcjmD@DDEw,Fq$0K#8~;Jѣݴp~#%ܰ{6  wGԴ.)A#ߐlqԴ8}*M}n5K[3H%or[ okvoRn_4JqF="73_Wl1XFז=5#CQm팻  ՠ$rIn7חc$43si![j65J`wR\{Y9(/hlw$M+\2rEnsp֩OBnZq~N!AEF=큳OJ 6o.__`7γZ!{3ZkcqA 5F6.CĭO:I]lcd%ϴgܰ>3l9 ̟ yI6+#喌G҃s[{ͤ+0mo';74Ib?*9 Er$EzĶe~yy4yg.d_b6)k7?AgV%|@t)uE{W: .yaٻU%v2L2<k ¹#*^W88ឯz *'obkY[7!z!}nEwFN p9>]~ $U`ϵ3yĆ$;fѶI'%2 XaqѠI[!RF^@v9> CsĞQcG{O*bnVAiPfDDD@DD cG<֭fk,M^>qGČSjk$ԗ4+fk͓97 Ջ盏247pB6ݙI^2Mjgt|Y!OǭlweIֽN-(-wWVzGYXƵϔn3\v1$ZOoi;+ 5eH~a1OciBۭR蹲W@o^n-=I=JsVuIY#Xn/ǔG/z Fu|*77!dlO*ܒ l5#s=AcaBkńڰݽsc`ع+o#A*5Uj?{t>#S eЫ6-XaR"otV|ۦ%偮Sk_*oK}V)zz>c=Y]Ȓmd!;^ v/`3*̪mL [H6.Ejlk_Lu.gRg>*Vަȿ_vZƣ8}l?eYw[ٳ 0Aŧ0ہ7i}7Y-UL6'8x[1C$dΖYޒeSeMߐMkb־і#hzǏ1y_c-ŒBXY?hܿdY 5]zϩYZ<]UNϕRq\LY7uX%1fC_tlnt-rNɬs@5{LlZ^ڬZU!; ;swuiID V f|Ź)ǩ,[%HxM]شy tK4dZԹIǵW˾' C!  d͔SyԝJ,p=9rqu e(u*G r uqs=Ph;GfQվ֔92M(9GLO~7 MaL@?[w;s+ӳ;,{EV&M9 sb&qqdm<ê T_Tw3]c~*!mIICqVpBoEDaw̭kkjzܣ|{@~W6?;3sZt-GԺjRno6YFiiVs+A#Aqiynp߳~}H?A Ҳ^)1Xɢ~oqnV~qH1kvύFC.4 gdѸ:,'=/AYXӣؐ-57g9t?B ˼}?:?,:i^"GpZ==[./u&qYtqx[ݕ{P1?W~X؂,^F ,{NA *4L' r4 sj{7c,Wu|Zdh<3 f.A[W1W5@FVAG=~WQfZZܒ= 9wuitxt}$ nR*=l`0s<.˜#H{~Rq3UU?/kǝs51pVeCb=Ա>>Lqc֞O.^{NGFeŧ6 Im$IY`ѸdO݇vwgףT.x# c}_Ȧ;Sm0GӐIVn*"ӯQ|Xyю8s@? Ҹw62 72:WS{CZr%+ E^wb/OŌxv_VK:?9aχ!WBDaKuY+Pecq'=ivоFB>K14%ǁF˦iׯ[+&9nڏK±st4:䱲ҌG@'0:0<MO^}]sY8kX@[耈 7el٦kY9Z?<~$!mmsGL_g?ǏZwؖyErեHC큎K3$gK#dkB}앰V6i:2W{ߟ,,4H\rG؃??̾Fy RzKgI'F$I# w;w-%wwv?Cd5Xޮ1# gdϰ@['&1iŦF1Gn> O^g=Ln8~3LgY57 {ޏO:Я|QA'8c>jo;9GjW^ Ïq< ]dP[ u 6GOY{w`m-oN=?`i:{ZTti9|&/QZ9M2b($;(,˴ګճѬڏ貵܅i:'3Y?=!Ϸ/r쳾ڶ';ҴΣհ)ij氭+2[AN~(<ңQq.Qݯ#9sO X *頃[VZ~rH,=ÖsOӮjlCfk6ξ1Ok3랈0.b:hzFn$kNT2ditFIwր\[jA *`Ś0x̐dtONV7v0]d}pɨ*lvk>LԳҹNٵ$Ŝ<3C}p+Wf%y6f׼WF3 A=@>mF*i첍c;G#pboe~?CX)Ի]WŒ? `=I컋OwrY/&-`8Pi͋naH66fqe*d{q)$DD@DDD@DDdF$pc :W%ono:~Cx@cīG|F`dDu=럠ٷt%.k#|1C5@ZȤo[8}<\GboTiիfͯV'M32UQzִOMx^O֟]`g>#d 6n/xg 7 :!v?"x|~?rȥHSx?*ܣN)VaMPL>[^{N- uD#{A/sZ\s@q*FՊ=sc0:R8czAѴ`k}0Ni@1G#c_?FIH.O k_\gckҟrlⱈLr<~?Bg\&єB6 >-[Up&#?+C coeL]꿋. W7x1v}qʓAɨwG V$s~+Ə5B&z~?Tպ4G\{J(2j5L?pwx(LǀRۍm=J~<}_Ԯ2h^$@Ǵ}!}=,:&$9t_֏=XRh";Ȃ7G1ȃ_e76';Y#wn=w?yѻZ]Ocgg&hp.i8HUx{#~2Ao;MeENK[ˣF\}J6t͎֤/2 Lq]Bm7>_Ķ7$<|T.ͭoFO>0=g=J *kwU]f&vÇ1$Y}{y.Q,4+!$z<AvERo{QoE.b91И ?ǐ2ۚΦe4`e×L*gwx8km-~9qST9" 1?/U{ukpZDAoY!᥸ll`,hÚ9k *GfG,c#=z-;=z+0 0R#̃ hφ:\U.d)y{K[8p r_fQ]0[9w.Ϡ """ """ """ ""r),0Jk# Y0,qTvJٹ4f:һW/gעj>s&簒C ,eJo>镈ձ<>#3cz8Z0h v W^la)ⓏLhE}H]_ap^< EGFדXE]2띲u5pLAeH[zgo9{Cm'bףfi;m$ZE75I:2R凊,h8q(⚃iWҞ& GԞe#-h> =#dmR,'!>ǐWoi}V[g~d]fA4_.ֱǦاmh*Lɨ֐lE$8g.AۺYuEff3g&~3B_fya O2RIcdwsO_R[msQdŹԵSqV+tP87t`۷&}eAּ}c gZmlk62ޤ; vMѠy<9k=g1+Bٹ pq`]Yϸ~s^qG垢I!VΕ&xd5w:zZTk˱kMFk~dG:Ik؁ ?# _HY^6q91ytyM[<4u9:;̱ns]9)(C# &jyL!p-xHpk>>V}E.ue!3ɉh6UquRD,WG]dm~˸%fk,Pپ駹%Ϊr> Ƿ-49ݢ4<2d8MV+M-ynF9x{z>)< # QrŝYxg7p 0yzOBр0=AZ[_nQ~3;Q]ՓGdete{)'IMs9CYD} _}V슶,,osȝ ;E+'tUf4lQRukَ7j1I 8KHqGEV/տjk;hUq"(.c_s|UJ˅Ly9]Te~᛺\fi\ǒKK˪1{C8ei A +eM]5f)}>ǡAQ4ݓ:ٿxH2~Y =4=s\2ڍ1Bıw>زUPXtoc-sOY`[XOgͥʳsڛvZͳ]$pˈ'铏|}{B8z';=̕a̘zۇ~[z*1ΒGq t. ww*ZɂGHwU6ӫb;3Xs8?ZCqo! ;V'\:h>ߧ(24宵'<0: ܯ5˶A8,B7~7n!1KݡSa܌MC@!zV 2faϥ#%בֿ$cӉ鯤3^;ps:G55tvrvbN\ĞGOgm>O0%|!AQ-mR[1>@0[C_XSջO[`kgVB+yg c*V&YqUpO@nNE^' me5ver [B;_oiMa?JE3(پwY"GqhUal#]wG7><\8}iS=)f\MF>ұ#!Ǿ5>h<~ioOp6Y䜷>0Mld7f xȈ+[ ^wu#i$"wX 8w?R OPũ H>XC%4xi>&R͗d$K}wԎƗ߲=;Nȶ״<[ğjܛO}y_^Ș c53D wحe!͋G9USK̯^ZBw\xNg$zo~!p <> &eWi^w2#z|D:G8A;;uaj##YQ|c!sr]-ϙNmf45Y `.랣tXtOn duzF}%{?u} XѲ6[2,a頌VџԏOb+{9<;lcas󶣐&3kFW{}}j*բ-/5(-w\*ѳjx&3Y clf2C\- ˰Eܷekbp`c\I.?/NzrѿlIm]TDq'XOW|$TVpk5庣|_<z|}/9jSM^A秀Z ) :zh]d%XM^D3Dm0>T}gm#Y "dlrc ѥKq A0G('(4;kOBgb8R,ɭl>.!lw46 u5,-YjubKl1ٵ,eain0'821<ώd},عk,-~O\Hn6RvDb~SQʂ"" vVkMhTً75iσ먨" """ """ """ "/2Hȣtc+^tC>\s^4U_{VM d%kн۸} ?-ǮxrP R(=stMdY`9 ..-{5ԯSE,l8J߄:qsh&QW/wΞTc&#Z AnGܭ]׷:-]u Dk1Q(Sז912I Pbh/GbVc@ R-=f֎ֿғfp99Etqb _#[~YЖ;"T% ͚dkxd+:" """ """ """ """ jВŏelTUN61" Z-=|4S}xzPFmD  &f]+^)Z=\h\vbi˙IDQa>>c!-8pAڲٲԦ7!쳢N<ωViuՅ%ĹΑi!x7}]4wʿ.lU {X%{:^vݹey}{]~~Ւe(%?[$s2B0s V~^L6ēA"%ޥjvQ?Wn7A`#|yr湽s y -/򥑎vL?cfprCyy'=~oKϟ:2gǜlv~AXZPcF=e:]}Z^2'xEe8Ȉ!bG>MvϮځۜw^~2" """ """ *?FGő)wQ.hL?~C[ǬXXI9AQz7#bHCX\_O>˶ӷ%GWd~r"GEM2>_W)$E,j$cc [ہ9I6d`= C*߱\nНle< KOawhܲ5ΈGcݓ{:׀5;얈ƷOS}S{,oW8N Gr;+neJj݌&h#=*4N챝KPr#' )['%ϥP[z-mvLfhlK#A3Ǡh!M-Չ~@6?!}J{t|\a|>w.s/Z ޺mܚ&I'1KwgkLmkGiOmh]0Jdák[d{g\agF}h .ֱfl>@83HƧ]N|ڄW$ 2>iv; tǞ!1Gֱj^M~DɔAx}G-{+|uȞ SPo2C2ܑǹad wkɺZв!mvg,\h ܹb7Zzo803&ӏRn{ ށJ~=8ϡcvkء}hlL; COӌ{ӇKga [hW6:hmShJW9>yɍc,㳩W*-\0pwm>N}Ӧ$?ֳMgmgGr2= %4Gǧ=UЏ16@ą˳>]k:HɑE~c~DW;u:t>#~ivw{1>Vס[e7e.VZ^0ޞ r):*I3L<<<}&͛Y 0s 5ZFHblrH484%_ikPٯ72Y9=C_~ĒK$bsKYGSn٥NCmV5X Έ5дGwE}62CY3Iktø[Ȁipy^N*o _74d9Nk[&;:O~?J؈*I)?Γ=ߏR" ǵOsknt~?+b ۝'{_dFT E x>'N1Ԟl"" """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ """ ""}DyK _Toc364245657}DyK _Toc364245657}DyK _Toc364245658}DyK _Toc364245658}DyK _Toc364245659}DyK _Toc364245659}DyK _Toc364245660}DyK _Toc364245660}DyK _Toc364245661}DyK _Toc364245661}DyK _Toc364245662}DyK _Toc364245662}DyK _Toc364245663}DyK _Toc364245663}DyK _Toc364245664}DyK _Toc364245664}DyK _Toc364245665}DyK _Toc364245665}DyK _Toc364245666}DyK _Toc364245666}DyK _Toc364245667}DyK _Toc364245667}DyK _Toc364245668}DyK _Toc364245668}DyK _Ref361132010}DyK _Ref361663954}DyK _Ref361735769_Dd 5gf  C AèDiagram showing where the HMU may be located in the broadband service supply chain."`b]&9y}Gᙂ]`n]&9y}GᙂPNG  IHDRS`TsRGB pHYsjIDATx^}Ǖv2VL )K.rKl';ee13,h]I$QoOCꪯ{L&EZ@Z@Z@Z@Z@Z@Z@Z@Zഷ/,Y@"WY Hzfľ6$Ŗ?C%-- - - - - -pZ@"3Rރb *+j:2*cVsۋG9JXGj%U[ ƭ7˯!`pԈW/W]9s?~!V|_ӏ^pgdL HzvSɂJ H H H H H ~QPw܉{miiAp@XbT`4ڝnwqeIUE5&GAkF@'wuwtwuT2` El<7wOvt|ӟ-//?H܎X [ ?|hY7qוֶ8, HAB bu[lٵkW{{;+BohP1M.O~iUi1KFM8<:DRIM "P04T:_qΛw$ݪR+įp:oB`a"- - - - - - -p, I3/W*;vT V@`R ͠<L<ÓWT^SQ7zܴ1S &aVdo%^h3'd k2IU D*mL'&_8k>Qd:\)q& RO}=%wwwoݺuݺu6lhV&dRUJ*jM5qfi%5c\e ~WbqC_U=S ȕS(ؗUyj!*- - - - - - -p(4һ~_ɊBjcQ~q挙;nWy>iiiiiiiD#fʣ??} C &,)֜.wYeu;uvݸ%eF 2Ne `-Z3?mHē; ʐ4kƔ OHm ID't+bjUhϦu]mx]Z@Z@Z@Z@Z@Z@Z$Y@"דdafZWW Ǝ6lń~7jp@ڃ1۟hh?"''q H:&=s5j„ FqM-M Q:2z,D<6XIǁcp iiiiiiigr@tvl]Vo;DNw Ơ>K H H H H H H Hz2|k̞=oj".>ZB'Ѷ '+!7p@ Neז8$r=ujc"ƷzNGh]㽥t-cc q'@nY\NgG[֍k w:BnUgρ;H H H H H H H d Hz >r+,*Joz:uD*fd8+\sK ݽM.X@"S\)&O W.7tF'p!+a.s"//- - - - - - - V) cs[7I0M@zT*f@r`H$rJGZ@Z@Z@Z@Z@Z\Oq{" BV$^jy&}1j(Jf2'P'fX`Kgg)6P HzkdX 7;@*zR X[; ʋH H H H H H H $r=ZKΝk0lҼL4XN4Zz"=voݵǎv!xCMå` Hz=<+Vlii)*-_pѕH@c@ TF3z6ؽYUt 4a$؂JD*,,0Ld $g'Ʀ&Wpz<ξ`uF ӡ~Ͼۗw$- - - - - a`0o~oۤs~DRi{ 4JjЦ 4'3QA{4l *hxH9vKJK -K*8`Od?  &V^VVR^b4N:u555ƧΞR o><е76.Ë& yy'Ga\H'q{{^(CϽHTi" @Td:r,m,h;й{mU}x`0L|WRQTTTZZH7') J\z=8W˙+(,xSL=k֨QL8'Pg4aMݾ :_\?aE0j]?i1.oJZ@Z@Z@Z@Zh, XaÆ[>[ĔƁ?̤\NJVx$:ض차붕7Ұ@&ڱ6ֲ:dzv/8@CWzQv=~It2n4[f+ Ibx, S l<b;#YoX,_$!ѬјqJQ\R]>ٱ8ڴ<)!x/){}=d Enq3G.-+{/J H H H H {wyetƻ[^5~ꞮHD!Wxh) дݺg6.o=LF3 -RJ_zܮh,QD?kO4  FQAOtl`ȰKdTk"@ ax򪪫" ̝3~Tn/-+-,*,>zooA?"9c'Wחu8dXnݺO}Swӵ 8V{s[vn&l^`v*7>\j4ф3(TDlQfBAݩ9i8{a W~mx>'|kOwFx4Ѹwڥ~ajoO mjG" +++KKJfS4De;$h$ϧ AŪeb7 񰀑j@S$;&0`]S5y)S kz*///)+ZE@'ըS.R}Ժq5bjkNӮI H H H H @ Hz{ܧկ~ vxekOz͞qBı"`9g9*+*jjv;DXdhbr[m%+ՆX'x@tD0qX-h)͙6u"dͅ ,B_ [z,-/EB.Y| l>&KǕ)nbC"4Y\5I H H H 5cʕ]ۗ_R D)bd`X zSҽ=>XV@FhRc8~p Am6 P&&!u kZVx\`0^&8f%M0jtĉ̞5uTcD*J TJV Wf ʲ֧L/o̺M D$ \7:De#  Jn6AUSťN$r=&? "g?g}V \) X-p8ۇpP*L0"IT c+LR{ t|?Wvwn= qn^8cT)Nb_hm v&-WH H H H H [4}޿o7uPQ^Y^^=?`#~ehHP8XBV}*X&E,*b#ػp8bX9a Dwk4+/ 7w WK.d֬檰a\ LK^_dK{Z6T\}եyijSQ`.G"+gR$p77v>j˛MF2B\0IULweS5N HzZ]Mcƌqou1ބV侟NGQӍ`(dB,\X"ԭ\2̚7eڔ .`Μt[@ /]K/?~@ tr6h{;}^`BN[ A UonobulYg.rqٟn4t{ڭͽpu-TUIYiViiiiis4/MyFڶW^y}RDx>SvR_SڝNu"Wİ!kd0CE#w`I^z/`Ϟ=/Jkk~EjiӦ]|х*i {|{{U,OX.X`g^ڲvApzZLIͳkD7C'!V<:|c]$r=uz`T1e|"KPBh@Hj98VH 8"`K `@Ogn^H:p>OY +W ze"[{|I \}M+Z}x믿vŧAO]]]?z0SчϟXUau-}Ta͟?c.V- ,=HyeiiiiSg\O>̎=-Ə*v}{ˣ0 U4p!+ŪOQ^0+4GV.Zf޽OWʷ[K47h3-'Ӓ>g$r=妍yف`jE민zZTӂ_?bĆYtzSn7 B-ѵGyd۫k? 5ֶjժWUU->oQIYY[Gu[,]~i;w>t;B|tzH r}kW"+Օ!(2;zc΃}a\mFofw=OU^_Z@Z@Z@ZZ@"דj~Ύ}j'̚2z/b{{jAXVD2͓&NB|W, T6z͚gy.t-^ly{[/.*.iiYjݒ%ot̘:78Obc|]~uii\4 Yķm]>chЮzOoYt̙UBK!w8- /܋oz K/Z4ǢO 7n8qE_wִ}CgN]E8X"ʝw9\7gWS]*0k^T}τ+%~uӳk"sǖ8usƞW8?r>ڍ;OtE *K <dEep*rJyH@x/2]!V$ؽo-_,Y8w?xYK9c3eohMY4 WɵU,U_NɦR5zeӆ}]ŮO]<4k ~%%_}|gnڝHQ"Vc?`XNBzV f۝y~~l꧿g*cAn3r 77tr"פB9zk$ַn5]0zjVI.GkDd~&{__!p%$+B dUDV=ӟΚ<ꏿ~<`dHN/ @8 Ƭ&Ӵ"['o<r:q@ 802%g8-.YlX*NgnCwY~,=aҥbf/?20~KTnn}KZ7.-S__W_555yyy6LxhL43}a]$ )w(%/==? Faw>>}ě/;vyE򼚪bOi :1.#^q5@Tw uӖo%^p78z̘/<0 }ܿ. PHL(ֽ->,Z8J;guZO\r%~ȵ 7,-/- - - -pY/BTk(AmUj\<1mhڹm뎃;v4ioFtO?RɝNtIҊe ʨ|lܸ=oٻwo{g7517u&b4錦D23g;i߼c+fpk&ud&Pt"N&2^fLW` O1.}rJ{bo߾{pFByNy f_} {b/,~"o^}ތ y wr`$w6ʕ+\?B W@؏}c_җ RA?D(/1eR{|{K8 JieXu>%%Y)5G0J#&Ȥ˦D'X&1ڳ&̹"鵱=/<{W[&M8k,Gul bɔph ѐkuu5+>zG)փ;H ĭ[67<Ɨ_ߺcWK,V̘׷lvXl Uc`YZ EO//>ZRd]w~T(8B jFX!2SD*.D‰PX5]<̚;k Oɲ4xh+VsE4_M|DrhE@->HW}5jr8W]1ZhGHnz}JIy?*0zQ2$"Sɉq],` l&]\iWt:9<'(,j¼Şeœ(AćKBX?>\Tb.ɸ ~0 C FbtJE uXov&KU.T5׎6lO=?*n(멵XU WqM$s"yɵZeTuO=bYj2+bJr2m ȧ2& x.B{"Wx|&45SNLyf",[|ōs}KHʪE4HNW~ӕS/>olt(`5  V˷`o[?=g<{qu=~Ʉ8 lP>;}L+"-0زe pj0_Ѩ}+_?<אxx5ÇvZ}/}ra"Agm_kSU WF?EY?yV<"K% 8_dV/F7{4NLT}M WLI%Xa ٤৶N=Xi}@Z~+ W˟~/@#i bP)\Z=Jy)@$ Bb^I`E^VZ@Z@Z@Z`Rԯ~n[iQUxh1[6mJ`tk b(X8pPZuU8/ :}REi@XX>f><%m^#W5OO<{:ɤ8,V;,fj6JK6t+#NM?X7$SDnD!xZMZ̳jGc˙wgWc@^^8+N x־~B,<*qG8 M hbݖfH\\`?ovįbmI20srϿ}o&?' 0 %kWO'~O~}M M@"M%Z188}WGZ^ %}?/ gM-M*QB̾>EUx-'8u@W9 ,+Yw˗_^j՗rIϙVm'`Nٹ첳gLb=Ė꼕 h2k65b҅c]NBeUeh~ mY,֎]TIG0N&aUYeUů"VoQD,Ɛa%e,ҼB0Hxqojʄu5V~ٷ @0L`}}3oަn¬Ȓ r/}z1hsY@q^=/s_MW̩+q(8\p0iShCV0?E8g-klvSx'``ܘ|Mp`W;7*DŸ@.jv/=‘54 y]߿K[vzx\1Ycss0^,!2CҝRP^?{m*ZSްŮgr2aiEQBJ|%}-卟VZ SbSO/%eM+)N$0e3Kxj{l÷ܫvʊrA2k0Xcq`^NCY%ƒm_ڽ}v[?6 ǽ^oG5!BnB2ܜ!v #eQmmY[;,nC@JLƛT/5=HZ"JHoA c}쬡]b b)ll;D+g1!.#\Yw5~x8oCXP5 \TuL"k?8z |_;5wvr uۿ,Vv> U+M ǔĝxn޶34)F-B' n%f0}Yo_"I!TvS9=n{eMǯy%i4 zOk1?ʠ%6(T_[fcXV%Z3Cm|{/(($aLSBؔLBx }O,bp_X>F𝻂">D# WypwźQr]Z#[eَO=ƍ xu:p v%8v <Ԛ_ ,U5ƢR1\%?рxRnsƔ+W!iEɳ(.&xJP"p=DE@/O@# ւXŀUn}qMw NKQf: Q088'$AUqᲰj>Ng'kOX1nl[z4Bc:5={ƌԮ{%;"@)B C[$B.. ;L4 Wh8eϐL$|0 lF}*2ՕE1۶m^`Aaa1D,-0'?jlF}~ˏ!J H H ؼSccXUK1VՅ &29˪JVJ_l_")Os( n??`B>JnaBʓo>aޖ>?&inaaw\ %d~^OCdUpaQή*z3'G%y T *kScU k .<ŊS žB#Y,e@tF |eXc+aYFl +4iUw=A} xte1x Ȕ8AFŋ g}C?x.r})^HJZQ/͹Ot9.&"|n/@dZ ?.ys\?V1T Lhdr9+/[wEnl$Ln m;Z:q$K.re Aa$8I]e|jE@UF%~Ugd*^ Qilww!)HJ $@܂& 8 jRB0@AaUdKHY^[P`"8.rwٚk+sLF>*PR2ҳT̷ճښ꺚8ٗ#ēpٱӳ纬6;&1Ͷď6٢fM!8v <AVӢs89(⁓kI FPFhZQ AQ"V3k1[wݛX_ϞW`T @(ըuH-]ݶ@w ?m8N TD (,g6^\"ܕ ՟JQ WA^ ެmbFjBz-׾1=]c?h۟Z0.?K ״`2EÚwl>\sp >m|դC\U`Uł\)+D^ aĬ= h*:]Brڡѽ8uY;e.>ע ;g(/4{.ȫ 6ݿqp }&O :D~@[9 _3'x=;(1xؙx#` 컪A^o 48> EFE  BK w\pbͪwih^y+WU ?s=8]-Gt{7C6nAt9]nwaA~aAAiYYiiaq!fnY`&aa^zL:w=}E3kj+Jn aVClS䚮%;viDb>4_M*sRZǎ |ێԬJ MG/(G⃗Am{{`4k'͚P <ӻc&`2fDf2PİPӮHxL 4 rbP{碅c_z st_XQY 3@'Or5\Eֈ~X`dJ&K#.j1/jH"$;*0dRn?9[[߷^T]',}ieÝ\HuĈ=xp@TkT`xjZŜ6q* #xþ?])ϴřu #Wį:p usP C<9iv%o-P_8+A ^V 񠚐BC9!$6`j{,bk!Zx@ Uo,kχ۶mE/I hԵw|Kܱs*;UsPJ\!6}Um3^ūx<"|1.bFÑG [j7ݾ; P8W?*& <` &M\eQv|qJQH*H *Vց?7[eeuEW*bq .Ho$ R&`ڪ_my7+THp/7rngVaZ"ͽ#U oHK,JsUx &V['{"ΧY{28sh *|9n" 'N^h *Jj;"Wq@,rMG@$*4\Å38|ruZ.S5g)eDDU K/jh  U;*Y.#cK^p!"xhjvttWꪫF,^5vWWc hPj8 W ArHiƚI &Ӱ>@P}#h! 3ngzY C9j@e.ˆ ixD4\1xeDd;P?sj?PeWXHgdifʢI?>ȵow&݉,qZ@D?O>wOB3ˮ8PA ASò1G͛az1г e {x`6/ϓ_ACt@u%:$q`!r@E R5=s ]۽=o K)8yƎ|Z1L**h%{:BŎ<[qcs,䜛@esI&ҩ*U]t/ (oQ8j *5W~^7h7 ՊFeC8i`CpUÓ Y.A;Dc,?&!dP m A['>۰~s6"8Zp38Wń'VJ(+p.vIQSЂN/l6 , DkaVN3գFBX`2H*:;;͛Ғ\[[[|U}Cȥ3Ѿ?w/HsE0$mR rO " \]q?hsv6 0łnz¢B[B_1Iu )1ϡ|oKt i*Ѩҷqgm e! qr`zn+-; *r_zO8/ rO}U~] 8kxQtTbG7TwtSo[z),J"S|fg_{X`tP⮧Ռ8$ )"%aO,?Y5JS"Y*9PQN[.S_w:lh <6J}]|WLPMz (T633Y7!>p,g%4qmB"i1Y m<z}8nޟwЀ) q-vzW$gm -~l=V__x%Rn`~ xp%y[dDֶݻv&p8_RԷIgr P.ii_ )S;e8: pG4_o#%lD8Q|F CF"EQWEY`bEVP(mim߻w֭;֬]b˖xsK/_jæ-{/ !9 aK9Fupt{"&.&z6V_^~}+6liœiȩT \8+%WvˊEYպtckw_ J&Rhg ۲guv4uGA9& "L bტhmmEc nO(cX<}_gUYޘx\SHSWbV<w":2Sx((5`BP0M)UJBW~pŔ 6w/1BL4]fTGvDr3n W׌XL䛅x[UyS۳.lT¬z t?@Nrh=d6 hr BqRKDfRą*Ǯ  `"rh 13rVmO`P@ULQa H`1m=}`zpcxhc(cw{X4_TaO. &-p[`i0_{KS@TQpHE[Fωd3ϙk75>8MRo!ui,&枙o2\&Dkr׈=y }} 2:=yk/[%^_e+ho|FcN#J~?'2aePЯHA+9[2c #* 8yajt&@(K=֝Woj1t!D"#(UDش]ۛR Dr9ōRk @alt Ю p$ywgmEAm;l:m ȫ O U/5 v8ӪlvCe a 8Nς&f!E@D7q\ѡegΕ s@+5+IZ䄮dDbf 4h`H"IX7j$&%vCXo:(eEղ .cf!8a|Bb@%rֺD2( 0? EY 1"LtͣNvuu("sE@\+Vǰ:|zgi+ ok- +^C+|GrD+}^qOVM=yF`UEd%3D񄠐XKan3U W1/e6TI u,hsEc%eBQ"@ X .#Rtp}6o}{ڕ+^b9ڷVf۶< vT% 4Yv2L5: ԞT![A!}.OgKFG՚ 6%NNz P$Ucbx~}ۆ34bٸmÎVb%RRQMOГd]<{}gBNJG`D ') t{o(0Hh6jE!]e: "yҨX"s!Ua]Hʾ 9p0Ufx_ceX w1 _ 3S_8eNTE\+N8r + 5;cl5:hR CF_» LY`B5Vu6/~D^tA: V3PcR@.jCrhLyG ( Qr됷‘3ExbAcvJu޽Ðp 9!!Du: t芇! ̙LH }@; mx03|PZ*-"ZYjfzi;vC.#R@*O&.I.y>A 4رk+Wͥdɒ+Vq];w75遂2?Nթ<f8S"WBG" 0r{W%f Ս()~T,2E# >,As8Dj1^08.x;wކ]g@eP,Prp.˞WTM>(%E΋fL] b&)+9.S|WI4/ݺ5ʃ(Qڽ­eTG"XF}1C`]̦t0>0kzen_/u}\Ξ#8ȵ'\npꬊ ڕ OqĮ2fؔ$0 TXjm pC50iEC4!+S@XEDǃEAB!k$A#z@[Y#n6b̏) .T+:j'Unڴiƍ5 Rq0m۷UJzu'P\eד`yS`٥dǯ:\ FN@zV' .sb*rhx>ڕ:8ٷzй:Jմ$<JTV+.pBbKyc4auXԎHi гԱpjQE(7%Xd x`7)|X w;) N(`^Uߜ :| 9˭fY~ /3 d2}3N V)r ґG^{{{(ᆍDm E 8pyRvT|SP]I:7ڵgӦk֬[re˗-`u =6n$pDY0GQ߼b 3i 6[P9SE*ķ2V4ϛ^UY$Lsoզ +P` 5* Rin5_TB$(uEk!pb;9TYh5t|m>lo X ͸ D8"'Xr˞1E!rWe' !6*^(&U$uO*d?0KTw%ጤ΃x|n'\s'C L4N8x m>-T H$C1(RD\<# Tn5e6 Ѻl:WiU!&ZPßߛl&}7 6vHtO0M) hDWH H H H 7 K 2J/A) RĽb+/RJdhA7zq&ڋx/" YE#~Mh:*9c]Eܡ/d90Ε/07ig-fi)h{[϶mǟ_C. [ţ\FɄs4?&2++f؍+SBA0$NW 3xǪ鰐dGy2@|cf [8"Bqt۪3g^Yؒ_gw6ԎkWTɵ_=[vJ#Jf'tC`4qSf0]2iT7mH'Y1\,B3 i {oL]m#OA*K;q{CϼJd!1B tϞ}"S*3 _.HPbL8J33X-x1)l(L" vqx,nv]D_f`kq?Rr$]O8!4p3rWlCrܷk"5#ii\D^G[ZXLL(ʫo4@L̆#;TK͢Go х|#&hM?GDX͗ M40+ūD]H2 K:xpKa7{=ş4f΁\9PXX:zmi!эbn7MxÉ`/ӧ{_{}^?F/G(x/ DbF]xaz=Y2Nj4]N[!ER Y* N#P\6m&XLJJT+ yq4)c4Ux>zIsSb-?Mm~d;d{G#/;@"Ab̨|>̈́&rb^ůw+7 -}fј\2 \D",ժ)^#hUEU>^Dړ?؜Be +6þgu4N,r%YaA|*pj B d7)TGbI" ,.#N@O1Ǘ1X>?-i9?4-* LXNbF xD"q :Jy;tZv7@ ElN$~=9vW9y`e*?<ȃW "@L" | @Սh"'h WinwT{DWDP@^sVPԯ\5.cV+PdU8/h {}>owv{> i8đQx es7aF1AAw ]c{9Xm[l?1:l#AY~SB_sB7D(?O[:aY3Rfj3ߺ|^w3{Ǘ7n;Ñ(Fi@ᦓm-=Wmy`)rFP. h a O) S>|芊EĬר+uwǝX nd4_+MW&\TM +z aGk k/\⳯ G avG%Ŧ"sx+Bi k7`^Y7 #Q `45 K[J~{)%4fE 1 c5mFR,17e2[Lԗ[Ra.2J< GGԾ}{Y#ը׾5 19e0hd&KQU V)@vàǕ\ Wyn(Tu*u%,yyc }wk#D #0݆yvAV"-Om}lU[g ƹ8L._uҢ4Rep@oc{W# Է rGi$>z=3iE#䐅@hEĬ=@Lf"4Ui '}aa)n(= 7#Љh_`aqH8`.$x{dvԐՠ;\`6c#h !hsMI0GWԼagu+ oY@nr8- zh>܋^`6;D$̒Ҝ*9^*,^BJ'Ĭ]?n`>D`,ڍ!D>Ls#O @ܰrA%PEl L+{U(7*BSI~gQR,x NimpQ%Z@f%OE3 $ &Dؗ J¬Ku.'(!Mz%bɅ80?Dլb/eBD 1NOC"N-\вGjcuy,r"lnLE'=.)txv+NqQm|i>M]@\p~G0qM3WfϘ>Jo1 jIOGW5z1ՊUbq8?nJ`RWU]ŌF0*؇K\ew1]TW#+! \sbփL Q*0'B\蓨SA uj 7$A"`gD%JnGaJV biO@vV|$DG{=RmPj>&d{ -pr,|r("-p,Xo@*TF^yRXLfiW 3Jj^q@Hupwv"J!;3Y?-KJh#bNd]SPܰY]NaGET1h((!5nɗ?Œ/DR (/ =Лc, J3amꈜar>$fٷ"m=_DX NK ~EK&UΘV,h.rx|C,g?}˖ t8:da c['k7EzB >`Oa#`Zx^GXĹzm$x($ 2|W)W+Kf]@e?Y5+؝"[jct2 V^ 2$@ u(B$#A%`(ג꒺$sO+s`#>NDK a@^Y}'-PX.27ʘ_ & &8? Yp"6Y(Q'y_Rgܱ^|F 4=5ڽokrnnopBÍm1~(%j+Nx{^_oy~CPN 0O؁Mfþzs( |MQx3Gp̩㫡Xo\Z7f) メ,6sAd)e U9ᜫp}`pHRI TJBfE&42(&y//^8t,E<,rSݴlDV<ِ49` z|2iXNL dPYs+o2CC tHH}]6 #kU¢=+-9ב<۩uSEu, CX"*p,p*TsB es ^V]#D%A bjv{oOCrQ%٘9$FQ=# [y ~ NI@lxZNI2^D? ; cVo^r'uN NXBtn8CT</j;ww PŦb_MI#y%q=j>l,{姆id_(1aS@Pp eb{\f@Q9Y +B#G38vj$ڂ`eF6Z0p_K]Ņpcf2~ᗷŝ} S$OX1d}^Zгoj2Ԧ<s'\q2~En.>CR!f "cKZEE"z;Q][vzMWO..,pR`?5HD-Qv.C=9nS a⃩yТDiw) ;!5+7"QB m >vS =1}0y6vZX MY{ٗ|A 1(Bazc1ȱRPS5cN TY,H8 +9yd fB0ESuʂHOG{r d98z{ P$ !:5v !XJ$ghW#<@^h\P *,$ E\p?a&:EFŊ xp9V/1 @6Ԝ͂y]0de+YJ,"!*C1(fevfb `KX~R)ci"¡`3 pPXJk4 ,9}ꀗjͫ,yH"d:sן۱Kt. ,nb>Ԉ  "K%yϜxsMAX*nHl|&ŭA^QbxPM^Tg=*wtrU"(G677*..K֚ zq.x#-S ;r$>S CMR?oP_8闌np`|Nu)\8Xa;M-P΋ PaǬ*hU釽{w̘1-}0G R[[ƍq[V2&^|RV^NQP8v,< .6?ǚ yn" XAihhQBn ӎ<7`:tS\Va QsXVy,pJJNzJ؉[9rA&;6pД?ʑ[o[gjP#3Bl{q5S$Ub2T*(qnh/v{ PիˆpWl뭨on?oLJA9d:C N|57aSl%R*ĊPzsUX!83<99 xVڗW7UYv3 O8,_zlXDꕇ^; d2VL<s-D0Q$*DAIBzЃY3@ E(zw9.Vhu 0 76jMj u%'7"XVx{goGGO_?g7hS in1Xy?a" L[>Ϭfe[eA-"'D WKJJ yui =KO??3tѣ$`dMI>dB ǐ\Pv Y4f}k\-(\;9w=c"׵#iVYY٢èY"s{ю@Y?AVљ UD&Tnݲ蜳HQQ,X̂9bE#zr``.܋ޅcvm:fʖ{9QC' lݗv#%B,x ^V8<δ,W' QJTH<!C>3:i|ŕsU  " X $YT?ftM|7 ծ!qo>vCBRrnCd)j*+EczY6靅ۻ-W'+:RHB4'+^0F*SmNxʢm!0@ń1b }0O~"݂\bkA*=PT~Oӣ°q΅cH%kOO\W^8uuuT-:σzz0ُGMtxBFQ$5DGǙ]Ss̭!\P ODپSUZa" *J(fbaUܢrf։+jZ@1 {5jG? r&K'AmY"QI)A|VHHK,U9# W?pw**aw;,0=V[!&yIODh̤ .D%<"'ȢxMj;A܂H*댟L~~>hQ$:)[?2m}[0a{u폄R`ӽ|PJWpD\2:fiؿ pYƩt8o|eUr]ge%[W֕V O{{Nz0A3 Ls-+/n*Ҕ^Ӣ*uX'lE7b՚D 3|$9cCvࢦZVΛ^ ,hB])bȕJ y`AZ̈́hQV7oK@}}3f̨hoo뭷zz{Ǐh"ɹeɹO=녔 8U(4$>lc <,皋-k5w}r~ҘUVMǯae⊹Y+硸V<[38B(Ă(-ןg\ry*x;w :ÎZUA %)Qrsd !T AjPzmp8O4Sy4*!.jG7j8t8ߛ Qs9W: O4b6^ 1 @8rYy.|?cǔ|_Onܭg`^H%@uH/8eMٳU5<6,,W;03⥣f ÿ, $oZ0:O `n_g <|NU+E,f\y5KpW1 L 0.T+bs]1c. 1{y=;&"aMZ 5b.)-*+/C%,!"]H<`(Z/]xr02 `,]>lv "fgIFE_ovv8Iմ(fS7ͩn W0Xl +U<;pYCaHxXb܋ pUDk:W|Ǝ{eR59aEyŭ>`/+T`Sf\9*#A-+yоb,yfm_uDmUއVT빒b cہS%`+jA3r%!*Ag Y \TmP9Mv;EEA:`y+ <·,Q~ U7lhmnhi2s}`6IA123f;Fz:;Pd'LAdH{i:ؾMo2x !,Y``#I@TdYOZq++ Y"QY0e5lL"=,gE^]`vLtuuCgv$䊐Ҡc%r=:ܑEГ֭Ӓ "JP/ 4&EbK.PL o #svm2@x$zA V58[lQx.\_>Z4)N(nPS'wUaEۈ1ɓ'RgULئwK$J1| 1VѡA,$ljJg^<N yaSfqkB$Jkc|Fi;),##J2MNv!b%heIBS $xYg p a Q bf}#AYO̭ bzq [Oh\"Y<>"̪4:fEMZ\fo{ݴi^ᇑ;/4|&1l]] ujGݪFibXZm^Rۂy8bbQ\|#sbRP V*̢9vm}aI= l2D6m1>ݹ;y/ƛ˟^}.,LМଶMI lc6M2k &Jd7aBe$™jp&:~W$95‘twuv!K9mvsp_ǫuw7̝;~lSێNC<<Oә 0|MPʜ ϩH | Qɀ\He|OZj 3 L+UIW'm+8tf3{*yP+倰VUVIn}`[n :Lت줳} 8աߚjV[qE+ `JYA UԚAM5xG (kѾԆz9xjՀؒX8lgqb@P~9 +Y@ Gß}k֬۳g?qZp*|G{$8OÛ,2Z`(~ŝN2 .u v1ZBR1K#+(W(iTj/@*Y`Za$" &Mv_ue'0;BO,9 &zdmcc{TEfE"g 4\Xc m9Q5G&'X޾ zB^1YªNHMYDCͭ-j]vI)5ZE<VvJT2f( \HN/ WgLx'pļ7P/3@J|]M>`C?3p#Y`sl".KD?>ɦPytѐABz&9 Hh)fGL `S4Đ<VUh]{9fWDp,4IJl:M`%x"*wŐA^s_0q8#;/tGPé9(RڜxaQZ@1¤V1]nߵG&VnkC*Â5A^ei?oyQwRA!G%Dn Q&&"1TD.P8 W-]oHO׆jybU~/Q(={ m6ꃉ\*DvFˑ>ӳ=?AE~Cf1N"+7"ɴ/5;~D)KISH? ԄYszؾ}]\43 9s?rZ% _{M"ףֹ3',гz̳yK25e͊|#+۴gҸ|#Hnpyh}.,Ղ@hfZ9̪*"oif 2 RSBY <>%J}-(V225-ӧY<}49ZI$hjc@Evt/++.R͂۴nt [k@߆&UW 0uz{I0qS)ODL * c4F  >dCi>g4-"u:'I#iI4;#dX*PӸf܍#;$8J /R5<ٍvwt#f9^px5† rH$ b'^|D:m% HEB$+Hx ;Ĵ}pNvF+?I $9Zࢍp{X2CA~\Έ߇d1^ooooo_/0Byщm 'gemE!X0fѡQR\ 86=m`2cLT2Ddc9jvpL̾? |,B!E Rh鋵&]@\dhwp~aX7!8x(D߳!\ҡ`L'33W\qqQQ_sus 4p4'h1{Oo \ eJ\|}þa,~Dhi Lr*ELP.تϘEax|s o5+ ٵʛ/ɎBoE\h+sK{mąads&!'xbRY${ A,Z`3l+}]۶lbh>s@0r ή䪵,cz.cF] :ZMX1r:#9ET ~D=w#rMe@W(BhQ2JRWu@Bd5PɂHze=WB$-cX٭P{ȱ,8Qx 0ȓY(;ΎEGjDdCq 2 'fD H.CR@@Vfၸ7Ť㹝߲|<>S"8&dž>;M2mؐ9W&T(P|X-H+R Jr 3wmѰ,N <3NP h$RtB(Ǹ pPMtp-(,@EPsp֔V+/eII$$n=lf(F*Ӫ2f Wv g_, 8lkm5"?GA,@B' Sq a"NKڕ*} %,Vΐ?*J+ cq>5ćUiq9Kz]%۶_im W[c$e-(‹g34x-[\Zp0غ N |L<!tf1D}B'lo9 \DCS*ODɝg Kiýk\XDٙ#ϩD\_LK*m۶KLO b W-hѢaZ`:QoҥНՏ= ,Pa5X,PѰv * &PoU03D.` ~:nN왔 /r$Ȍ҇_YIw2aQȰOe:c@oCVY W2Чѫ7tuEzaTv׊ A W za BH0ԦӹM-XJ'3+lJƅ C~R̮rh~eնUmQ!qd`Y¦(;`*x:Z6574hi9&7csW1[m-c;[wއNjmNJ̦hu1w\DN|4,b}Yݼyxj郑nik/vu ӎX[1;/Voy;슃uN=9p;S !FHGD8` & Xs׿o⢷ ~ 7IC3D}aamċg>Aև>ח쮺psL5ڦZu #YM(%`ݬ(r\<1c=4{㡔6- UCϭ唡˸GS V0r9\Bnj)4BILIJEq-8%E@g7HR_MS,6$м##xa-n5iƎH?zL[m}{7;tw/C67UQ" <76?;F$H# <\LhTF+U1#1g! HV(X*uxꈸɢhrE}4<*uRU#]t!S#\4P#+N0.Yƛor=f c *~KlD>ڏSo05d;VHd^M(P)9(B}}ܥ$ezs 0ɨ%1^:W̫{('0i]AyŹn tދtkuQ~Vo[M.3VЬS77]A(6rIɮly#YCQ1IEvۨk̒[q|T;Mo3j˙14\ l BjMv6a!偃~_Kzܯ@ r˭mTVZmoPoA4H%m+3(r CA~!*p': PEEaC@>r4i޲[ oq4+(T$j_!Š#b9ccNځnYl )~րH_2 r#Z cULqV'두7g6W,Xϴ" OMpckbI>@USKt0΅"׽hmp*WQ8rZ U6BTQ@+}70' xQ{ZhΚUωPك5\d:e܂A  mPcGgy х̨艭ʫCQ: a#6<`i[HeaU Wt_]Σlk :A@.,=jzX"–FpkӿGKDG\g9)ީJD\H|?W ufD@ gK2}JG_R(/r]r cрkXF0JUdb8_uefo:ϴt]roRly2p>%Q+V*U^^Y(ǣ + F!``3'p0o U'9{+rݹk~2'%+wCúl'/'݊ՃSwN| pȏ$SLJ\'ް 757{K$%ˡB0 qB{gwǙ%n :IGyV(ݰФST @(\qxEaX!0UA˃ 'ihg:YT r?, (J'fEoTj%߈=i~X&rah@iOg>S2*׺;Vҳ 9-KC;Tr?p 6#\hN^c_X9B9,eS79W8!~(鹱աԅ5=ݬ,N+7ǔk@ztW>%ձky崡^ۜѰ)>cX44հZiM n}K0 E!UnzW/1"!,)$*Qؖ-{i)qd.eleLٍtUӦܾm; @Y'Bea/)B$ס \ eC& z L ;(րS:CC(Xɨ(U#+".u Gx8o%i jlDXL4:V7~':9WaD0+Wq[:e˖5~\]!š E4NXX) a^XJLebXAYS1 %3}Оs2H"_s*Ff-P2tt BzSI:ŭ:\i,c1n$z}Ag5bȭDFJg܅tSGq qO lj~J*Aбr7SW bE# YYrD Ę/HԀxйniqs]r;fr<s'PRہu_Ѣ/z횷Ӕ72qh}-ZnTP+p }p}r{Ґ07&F*~M%1{z)Vh^PlbQ$V{PRZtXb慃(`b9*o aѠS4g_3Pé=3` h5t  /W]۾fQF/9w\tײ9"a* 1'QSAv66?V@7ifn6·bmNo~Xuw(Bǯ |ײ db( Y+W~wbї5CN^&ttUb!BD|QAQ͒&y~1d;=*z+!8#ۚG{$6X葏(-)TvDsĊ#{Rxp>ǠD$6 z.BE -G8)_5&bTJ_m:q4x47766B.ԀO?`Um?^L7<뚅W@/nO^9уPL|Dx5W765ORm6{_7:bJJ-Y{ή3υj7oAJrJA;z=DN:x8C܋U QÙ'!`kDe p BKmtzr@s CH!8(uYz)*no`K5kb篬+ ܎c&ʁ[Cjru+8'OοbP$!q]q%?csɩ Yc3Y4pc ʠy`}A[xR2)'m`ȱ KgΞ{~xF J n8om dg}[n2Z8c=5-jS z5_],*.?:S(cnU9zلYNm1H8UDѢ ~ԐEpXX"U+<{\ *slUsgvEtׯ6&ؔ qF噰>|(hG9o禚']Ud_Ĩ[b4;~ak/0R1Ez hVHZ ue}baf|=4=2@)-P[A]I^I=[ob KJ]'(D[!7)G32֬Y T^k$ήW7ρ??vFضǨ gȋ]\q3|'#7I N/hKy C`D@,Ds@ˍ~ !*^,UIJuA|?tI^?|g|]rI@j a"9mЪ((I)mHk XR -OL=Fĵ8z  gY@a=(k%pXLiE?Y&3)eo+~ ǺҦ21 `ȗXD'ܟub|͙3EvQmu&q&qONNV&%MF>UWGOw a l2v\Bq3dMYoWϲE[^Vhb4 Ocb2:;Q6D| x0¹ ,-ḗXݞ#aiV 9REDEP؍z{ {=lOPӊ0qXͷŰƂ@^NRryY69 :̏4=^N9Q_ wQۏQa/,h4 Q%z( .^Y;&:^j)q$bU,\@+ejMDmoIQ:sB9)&[H`p[jj4HsիIk0ώ!/-͖4b;uEo#,{8,:rɓ55x ~5(;@Un1fD` ܠPgg'avX:-A %br?̐`uE yNmSBbGx`ΉH/8^ztZp2MiKdE ޏ6ZS BL)"T=S}aD%tqfwz?*~ V\DE7T4Ob(7>$FںB7ߣ.Ղt<+Q\@!`Q"z-3H:9{`CZ H Pp )Ic83Ho]]=FtW嬡 g59U"Y~ܿIt$4ypEQ2zY9}ht 1'zS5ӓ[KloKhңk pY\8k%J\HIr&ң)JؘF1,s46/!_xszo$U6_axz z0-'v؇q2_T^%?G,m姤(y!ԊĐ5S_4[fy"gKS9g: J`ՍѧSvGqJZ ]/48#\ f80‚%)ދ۰8-D Dߩp(C$$9ȌJ3#85 7R%k8C 9XwlFbEjiT()).--EnJZ8 Tq1Zg X^rMQ) -tHP%1ZGNbs7vlE譝Tx]Dpd54 X >343aN8z2+F6v+-?g>Ҟ)c Qˉ2"M)F11cՊ$49-_XO&9%WXU/͛7>tBa!qS!*J'tKF5*, Eà<鷱άvi>BUN.b2"y1*SGO"P9SiBMD ߈ )֣)bY~$""RǕ%iر_?| u +8"G8Y] =XkK{ vU+='++C/AJ177/`p1_ ܒG;k>qiuW` H%W=JlpXF*K$YY _Zn XѠykwk;rҼD[Ũ:Rn~upFA9lZJR5tE\ *NȏyzG#i6 N'>2ˉY⁋⦃O4e1xN;9]#'~Gi_2gnnlC' C=lii -=͋RnwYtQ!s=} `‰cQ6LLY1" BG(N-X,=$(co_ƂfB@xUBe6*[8.[lժUYp,ƍ7]Rob"AZ\.Y0vZE :*22+cгAFu;6$2vClBN }dffrzΞSLkNI' ֬jG\Dw) d& ƈJ/v{ۿۉ彶A]8=Mmll_^ʼn,(,0`X<Ɉ-\yt/'(ďh 3YtBv .:ӌV El ),w]]ᠧ1WM Yn-XRPXu{^S<tw9gM)-h`Dҥ*BV>GOnrs V\=0ԓQR\ Jf8q^u _i+aaDݒ2(cQW~W"iqS' /hז\Q:>Λ^ K1@F̕RXV j4 A]LDĶgv:3H WhyJ,$"!J.(~YZ*V~w_sִ+V.5\_LPoCwt|ضhGgjbh9r_yP9h;smvIFҰ&Vɵj0 WAit~xug݋Dͅ?饿O&u sOu^D,[eq傶TeA9=.G~#Lm0F"2f ^D i c&U=BJBG:\WV gfdb[xIha$L&7q{>ygѝpOB' Uyy,&_!"q^bJcve6xVdѢEKQ. j8jɸyyye˖-ګkzժ-(6%';W{ "qn "f+;n1Tbbh&8,$y&{d r(&~|wP1iY`F#lP]3tp2'ؒS܌t)a&%|R SJmoi[84GiYU0YQ:ȷ":G'&cXrA ayRR£zQ.scw7WSIC[EUԚd#h-k$X@֢;Df( sS2$OqxK\w՞ȶik5:1,TFbE ",1>G>AO,IssORTxyee%0yo JT^<ޒ]S X RޅhWM ;D<[hAp1W܄}q`N_g{pDlP:A5T'"qE6  l`{N>V%%&޿?ׂ>p6?!>ۉO-D kD^d i2 P6w$ h1-# mhQCoQbE$0u}opLyecIgUH&FU<ݴ4RoTwT:ȼ%G t `ڈ^ #8<e , QӢs(!w%zUXjCr_j9T!* m_(K qtuz/?3??OUuξdZnpJ@˚cZ q˂`ijR,DzJhqڀbzokwyy|Su?_迸7+%aM'%[EcD3gcC8{gaHYz뫯ZXPPVV+={iM Ť[OS `~o)K[,z0{Mn˳Ǿ=*j;䣸\cbCjR]^hcӋf#7J?&C0QW$ިkm xu͈I;+Rs>8 D-غhoʳ+GO )O"r`v0P?n\kd:f!e|0`12Qq:.㮢g9KXOI)iSK475تh5QA˂cm͉ .Y<|Ko[ w,ziiu'Mu2BzH2p/.KQ1WtEhN0 *JigRBͩkL%%'dQsD{寧mFlkGuv;A80 L"a <*ܡ5#g h69] 0Nq$:dPw]S]S_ڂ[v)jOoxhex>I[ylt^ndʺ8MEܭG!VM) *4 ^́AN_9O3 R M]{c#Ј0M&x쉷R s] J +*`$U8W&)0,r}ݪ߮D%'f_(MĜlo?pEuVemV8XnqeEJ抟 2~Cf"iN|,bIISK. BqD鞚1}J,]A[x;QSD뵙 IIִ[N%;&߫&:=ZOC'hrq g>gdzɲηs83 +U-j(1O8ʴU0W;FDݛ.x¯P;{{>)h65:zwZRVF3`#M9/>) mm߲lмҗAJsqwY7`6'6Ui {C$M:^Z: Ummo{0V h?: *h|-_O5~䊅_LE| QN5-h7(\.7JQ*ɂEwށ@qI EUkQw, H$q)үDe@M͍". z`*@((JGZ 1gJ  -^c\BHT{i# 0,67FHGbeiYYs֌ /P1syt0w}ϼau Rdx~jDX?B|% r֖$|,QU zϿفz՜ZcrՔJRP@pJ{e_wL+z\Fod-XWBC;GD)J$\8lu &"m~ӛa6̜Z`8Z'4+!R2C>.XV?|p'Շ<.;ϡ~D#k(}/"0n䛟^BvviK,xúkf5˗-5s #0.Nzuq!ςjDTu5QU)닊 <ʀTBFBXJs2Du+OD#wssܹ̹EU\t>EZmBI> Vj4_ Y"^0>'(J0˞ਘ2݁ȴt djRD}+,B@Xs2s-L|~S*jKKt2fNɢcGD6w;C}}ӳSg!QշtTT]Dj(72by;Ӆ@JZ"cD={F*/_lqhobx~3D9f9L )r02 0\*NXTIEWYY YY^G ># W_W$B}vSo8cx%!(gژ y>l Λ" fI 0A| lG!:vECcŴae1!f:D[JJ>4 ڊE@Fy&H$ B7oQz{zp̯68@qM Q$틄*SMAv,HQ Aا JUC" rpĄ33SҧM/yMOK _2׳ (Oհ"[^[T$Q/L辸(#'88oCMJH ^ȾB8qQ ɄA{*[^ߐ7 "a b?F 4GC„KhX9[IVb& {4YŹI Te @U}sC{0W0H,%8bK{L}Ӡ/.//6WqVG-@ini߲;'+?lF& "h+SXyl\DI1J{*O"2@X1Nv (j/lq»oG X'O.\޽{PV*:x7zn{z{ 5A7N Ɖ !Z2e VE"SMB{\ij#GQQR\ I>kZZ*R_ QA#A.m͘t81hb,5FcA(e+Vc斖ҩSGYb"IXi ~V\X" H&!,qzٶv(ekjpA>r'DL)ElZD)P2/RLᾬe!,'"9y7b^#t9njYٟ}O5g!VEYJaFhkpa 3;f%GLDS?z7[K 7<ʵLڕCX5L-4[LiQ% J,zKz ~Q` #pV= WԸqSz+E}}tx!$/pQ \R勄".q 6D-D[N[pbXՖ^_"\=M-w[t/+W s BE=0(`!@lz#Xl3gnQq1Zye  #,,RRRN',X.bK(Jl=gu=RQ:mZff8"Ţx&!P "J H1Џ Rr …<ւ"s ;ru,!$@|;BEN4mt0`l騘S*H$,;?T_PS]{.TCĬ<_CS Ƹ-Gّ̐u]b"#UgPs=Oo^2;E4."pgy+{ DQ抣r+KU?,w4ϟc 0kIe'ä(A>8{-5+'8 QUjBX2O?8xh[ NDg+qMUl+NPAصb=YHaKBA" ꦼug[gnXZ**6k18c~K~Q87O݋:\| FWn'*]y(9Fޛ[Ifʉ^V U`,(5`{* wԴT`jHfgaPUQ KEּ*Q|/DW9Eq^ÑbLJ(CVdK/e2Zu."==>шظ.ѳ0Np??O=PóB'y6"<2/JBdMԅ3Qe:ASX=@ E:()">8??] -}Ġ̱G-_Va;w:Y:pHϘ b*nA\->T9G/(,士J|Td EoҼS*pk533y!H$ }CKV0#^&p@AMYJ Za"F3 Bof51WT ]rIDAT#3zVWdoz2sьU 97e`l#EZJXE_)6dX1WfH8woØ?JBMn|ARO@B>}WޯL9|r6A7 vGU$-B8 D&"Sho  RRT;RW &7jPGdBtOx'Rf  m,99]9 ԁٱŤˌbq>NB2PI-R~(c YL R-˴tdGokqƌ)FD`J|Eii* 6l{2kjEܕ2zq!ẑ*s@ \UJ7⯈*>G&S:Z:z( 8bTvE9!X(6D5 Πǁgad| o:ܳǭ_7Gn1Wc ((% 0@lHPX0ATU `Ln+@O3N"xdԧQMHH*_t1HC?E_? +AFd0P!`@_$Q-] 4@O6:eʐjXΝ; ,ZBH) /iA" H. LJ}DB-(d ḛQP_HP7ZZ|!"t JHU&'Ae |KϽ"cԌKRU71W} @qhމx8liEyFр"엢bs`V6?bЃ8/!}{o( "6V +qRUD[) Cav{MdYXରj9^MfE[)_jr1U^>o_?7TY:kي,\wz)渙됾xv4DJh'X`D.K(Ӆ\bE(|EO) \;Q R|Ӎ]2~l")F1mQ9J Mr'[02 BT䓀44qXAY{ z=}(ފc !TL~eN# EJ^vlkgr11LYLliiV*,,w5sHbC_éXbLl]sBܵm%P 2m:jsjJm1:ρP [[69Ғ D̕ Ps5g$gFO]}>:ϑw`)R>`}[H7.X,tUh02 p!l&ْlwzz] -vEϠG1mDi+.,Nq,Nc,Q G^CA֔ĒR,J53c9Lg\Dz IP[fY{f4K@THd51~Fr҄ q(/@.|X$w3cQQ[3C@؈ޱk~(Ql:S9X8 2n ?,zA*BJD/6i+Q1^OQ͞uK@zpؕƤ9xbOv\U]9Js7g O {I:Mv59+X"um&l$^3W #" i! oYAYb.w['o-,Y_^|L}015mL"XN47#kEQVJ8K,c`\.>P@Vi)Ljmlnپ{EKJH٬^#.@\+¼%grGl>5ys\/.*DEZϯͲL=[]ם|^ڂYibTGɕ1{//'*;Ѝ_rN"-(%7˖` o8 A@q B̂QP2@pW!FǗ&ZP@^6BΓNN Ǵ)Y抋mDO,!,XrLh+@:ۻ29E1RD>@s1%N)ɂz άС d!_W!8d .[\uUW^… 0X|9eVdqŽ %%jE&50`u֕F[{z>8շx.豫+X"6';z.Ńlٳ,ol7ŗ7FX3}5ZFo_,5ec4ŠUcCZ{D`R"o$@908_ᯞ#UN1"kDNkмGjuxx_` Ա<9! XH S]=P9BJxH5y=x&բ;כkpLҽ&𣴳>՚EۥfĸbP೧t5Pב<|-yB>TT)Mxx_}gZ;YlD@"pq!`!F?|Ub\qJ+ɄP0W1 H|Jh~f$po/R`Kz F& ĨW,,iHФ%!d҆tTBXNrބ <̏Ho7Y:v}2''mVg r,>HJ:ATʋc)RrBtQQRz/_:V͇ccaD4>-؝ʤ8{ׇJ]gI>|﹭zPKcR4id]k@bOW4`6`ȞFF[F8ku=1UEÓx +9r7$ HZƷA1௑e9?58}κF.*Q18V)F$S/h ج? F F #s"ʌ߼: c+@k A>[_UWW-]]n BuK= n+wڽ??ng)YNu h]$F%:l3׬v޴fvEc$h84Zk =D@" @Y+Ħm=bGsw-x{+|=}R> N9*HYbThCU(@@,&rA 4>-<'[:W37]g'Rs ?Sn4?Kh tp{\vꃳSsrSr0L[DQ+TѾGk=AsJ5eg<SR[f]lI콅pP7qӾXwڐm 6/L19b14(`|g7$p s\.GmmIToljF5Nr$*{)OI6ܾ[X\]'7uKy/}hR6 ʋ? G]I^SF7<ĕj{<$Ӷ¨r Z{D`#x)?u+u:w(D4Q1tE:W%*ZJhIlH"mY3F ³R]A2]nАJY&0 Ѫ&`' Tlٺ.ĒjMJ՟YL"JePI@|ShWٹ=fWqŎ\KjԊkO q+kH23<5sƌ#+!>kA` Z%Ss2Ғm deLi18M^)< b1o[>ۗd&sٷ,wbn Dg/Q߀~o=%t>0k-%1juuIfaI[n_ASja6A^+i\ TԊgIqAPaiYfeO @7zp! [+>g@l7W$`*$R`]\\4UiA|stTۍ?p1kb/~|m7~J;''uiLTIu.Ƣ! b TJ/RQS6#11QԹ&t޸=EEI5U]IOr$;{?0Y"5&)D%{ZO_5sFt5uvw;GXcLL@F"dzEdb V#& mekY ٗ%$UKh. U=c?՟k F^>v:YV4V4&Yڳd:I8m w>mϟp_๭݊kRs?} G~o[N" E2t0[C9՜~r sy D 3D%#(1YeOΦ>'3s( s)$'2=K:>CĴѦ?派/.^'b8%+ұ累$%)61ஏpR1ᕄ+Fdo#U" @5d M=5C{}KT[6n!Ux6*@ɩM6kjeui_G$iQ(Lw@#.bP 9 |Pi;xcW? 7QF:_0,+6IWi, h UTIb>ȕ9XE"E!$lՇ;:_E[u%N-"K #_]5ǁC*Hz{nAմ$uvQ y-Je4?hJ JbDND`La?eXx!yG ꞦCj yEC"缘Dvl2Wsxvye6s͛7nS睵x׼[Othq^֣/vໂ,+IP:ԥ3$eAGRzS"bmx"3VU)@b+b&+AX|Z 5kǒ?'b@#?YU5tۙ3Yb(-[mM/qL ҧf'նnDW_4% ' n- Hd $ u {eW}Vl;ѱXֆ`Rl4zya{wSSo vx45vͮ]S!M='Zf"JX9`Jf*TAd4V \q\w} `ZuJ3o'Y`Ⱦ&vkk9 gHApf]   wQdGD+rSt=8# F+hRl}n XS#0S%@\ɱ(FDuGU#Bvt<U6t/ **_ f' mAr!v`3ΘRZ<`RRw8 CFyhat81.t}`}k?h֦77wu}X=ܢovkl$$O rCګ(Kbֱ ˟y6oYnlo½+KkSSdw"u=j`NrTϜ^4C6! {zz?u=>|-@MҜqPs57fFA*5m#2-iJLq8zKcja;#@iT+(gV Œց2Y`@GeW=d_ELVG5+ջ2酨l0[t&K` |'su8Q_.KQP~Ook[EaP0B.pjP=/`Jqâh+Wj_H'GjqV]]ZQXZw7 w5ms乽隟|\ڜqc ?yLUݼy3ݻ{:}׉'ӋrF}gbb`}x@8|D9 iTBU? 6`B~FtAm3 |] {V탚Ns&[~fFGwnU.{9"@#Z 6z9!yBKZ T l2=3yGSu`pBQi "K%F l3,8Uo-V3[{}BO*քĢҹawLaqV1 ܎&`[WhG&!ۏkݴu^]8yn iz[#ݫۤY/NM*Z"Kj%mmmvaÆ`O?_xVۮӄ &>eJB&^5mKC8Sy .sP,1!6`j6 dg{s]ŲZ.C Yk{:Qu2 C-߉>گ2GĀrOx)AT.{" iz->rNjl$Ct suwVTumi!*cȿ+x>YlnEZLz䂧?^fj5H~es;Deft^ck:f| RR`)HRhHza@+5޻:{Mm cwmѝi':Ç[MDYڇVjM<`/^zyo >{Wu;iЈ"ԫl?c^uO}}-d5=7K$وcD A9 `ߌ҄,Ӗ7n`wyч:{#5d>=>} -Oe?`>U8UT1K&H9ID("?d>񣻿Iƶڦwu~N~_Ro„Ux_G*+Us- &bmws&J7dΟ2'ޖS !+Rܕ"~ltV  #Ԫ p56vnmY/|hssu+u NM^2sq/ `}ݿ}կv؎0ІPl ٥kW~.G~4WtG>T -e$}./'o vh{}$[I|# NZ$A|UξiMqF\c6H>L0}&YӶy둳I$٠+Kap=JZA[? gk{+^}\ 8#\XJoE}$tx E,Ұ==vw/%xi3 s;;{{n7QVa+Vy^j d0 GU_λ_o;XVrk &x!(jsd܋m^7|g?7|+_ʿ뿾 9AbKΛ~q֭-~-si(A\__yg3uh}>,oM`Ͳ̻{uYi6v;ܾd%fWNg+՘ܲXQWP"["wJEA0[~41Bp6q=r6D@"pMԖfR4f9>~O~[Xu]j^~7z+*~Ԑ7NVJ"*@XfU=㱭hR2sKjq] 2JWP=K[e Zut8GZN0>sײ?/mG2 ۨ@/g?_}J/pє]\%sy][S6'$a 7?Sq*v::;{zx̔.X255jx -5E >T #Gg |ÿhV\eHZ mҥ}I$tmQs# ffEӿ}߻:=zH+omۜ!xGŸJ1PR>Z>g{ VXc%M[Quvv}~rsŋv-FI>ʰʹǷʴ~+~'.M'հհj @*r*V̧nhs:L@'A7=)5 ڊNVC[K'̥D@"  `XS"Eaumko}fkz-/ךcȈKכTV\ q\ũh+U.ncٔBSҤh߅: =mm^ۋ勊oz֚1d'v\'6V;:::,{ /Vન+G`@l5>113h f&dMw!O/aAiwءj\T*1%yܒD@" H&>WMK*23327[kAֽ憦nË%)ޒbH3 V+\ujQ-+k"g jJCY@K1N]'tWUx{"bN^8+eSV-2$c$t#df0S\'SW /} AVU|e1N^诐3s̘;}Y sKR FyCxcxA#~eY0 sH `z,:){lD@" He{xVvQbB~|ߨ1z'$fL1kƼE,*>+53`"~UpV橬U $tvvA|f728%są}\7}wN@" H$8`i׶k~M%ROY)>?f8Z^U^]=.߇vɨ`P^Vkɚ4s|VɥNsln$jm$d.wWSթA 5EɄ 3Izap?˭kvuual2edN5k%O+lՂ|V1$1:ae:z᎞.$\V*0W!X?<_ gwrqD@" \&`Ty&NIjS,= / =G[ǁvEvJ @!E% zŘoNˀbvڴ`Gu^/Da)J "Q[=Q[0/b^!MnRO0V?JY/)t7}>ׯ>OAMTf@(\Ebk4,ۘ+ '$w}_;PmI$D`#+Zճg+aH3IIsCz4rpeKBb1RY^YSvNjb؝8F 7$$I" H$-us?SOkyno^IYߥb{%OlQ` DȻ#-MFC1]wu[Mob[nz j/ 7ѕ,5xH{Q;w̙3AaOT>~U6E9+/Ċr}Ɔ܈D@" H$eh7Q$u**6jDeQ g,-7{ @D"V([6_[ mzRf$6˛yuqLK-yv3}yA0ۻmXF"=}v(N=0kvmjk{GsSӘ3J$D@" a`Q@͂b_}:8QY>q4aHBҬ*-ڢlzPבk4kBTmxx"gEص0߱u 6kcD7BP88|{NB" H$0ƮsluT ]W^ Mk>zHF do=jډgѢE999V-!omWaf֧u9j``\D@" H$uŊѹt44Om~x~*G eaPy6m!i+o]Z IwGY- jiӦj;O#j"v{㕕k%$D@" 8 礆D|"c=8~rM/khh @(Ry D + f}8ȕ^QB_ -epWNᅴ|>簮$Ҹك'(l[+`LU?>鑕; H$s@lyssrNu"! 낱+J # !Jsrņ\5ހ sAH$!SsxωYӹ\#fSHς`]Xme:uj6I Fܩڣ# 0zGEFI2f,Y>Xr7V0Q|ӹj;=ǎ!o\D@" H'[6n T1ڴ~mJ$JJ1- +qVC׈8nm.X LByc/xG9z~fPޞ7r{4u՚u1*mFih+5mFl P2նYW\s`GSȷ꜓V޿N=^u\\D@" HX œ'ֹw>³硢L2<+1,RSŁDD_3t=Ks%ذGYkS{mniZL.Kzz9bEEyCԕkzF6cZ{VR\wOo]ArH$F@չ"hY6f_ݫACF'uܞy2󠕈M\iBq`JWξiMq6U&CSa=;{jH-4=E2I}* 4h…Zb0`]Gtq,0)BSuw+wC" H$@ &9Lu Iz̬.GX:Eõ~uŴkD]A{zvj֠#*TF۟ѐsrsFs_eNJP!XzujjjoW]f@Y@M7j]a?б5jp:3 r!D@" \PDr"?%Q[\iݝWEs$pz=B &On&bMOnܠ yG۩QZڧ5Kb0\GAf 5 XƲy炴j_itvl~;-0+kiVi,pP[}PGS{걻{.;/rD@" HT+5R$zͳח'()F+g?4JxC6G&QJƨ)~чygk&a|ذaHwMHxV12kV>=^>0;Bg[cn>5 gyqȮ/_m05#+YP]/ †ajC!|P(-&i<~}w^V8q!!if6SmVkVVeK,Y1!gH$D@"02j2z*k`2kD+26S4CԖ3윴p6#eft8Ų|ƺ;ubZAg6Ao4k@k͝kC7p*L)R6 9nlj~?_JkK5A.%H$D@ 2W>1X|8lD+ѬbiFhasؖ Q6 WN1)>l%` :j`}MzEg *?h|7nfASRRΜ9(!!> z^ =mu>9D@" H$X뀄,8Qł6Bl481لppKF!;ÿŪg2eʜ9s'=b g2-ᐿfɷ9?Vmd6C@jN2ueeSS]NG| 榖5UU'[𷳽6Oŗ#1I" H$xt_mc4 ѬFh#ok[8%#Av.nߌs1o||ҥK1j_]yC(hk DYa5Ym8V#;-7WAa-4&%,Xhܜ` Ӈ|,z!kkO q[m/~Ws\D@" H$$k40F;Ax<' Q}χ\nWׂ|ۅ+h+8+Ԩ+X֗?x?jo87-Z Z-''gyK/VZj3 ݃*ook?y0t:jMJJl +NTzw]}/cG+B&/H$$@@ejIhOm5*sg60he|q#S`8-%#lE!(E$G(T^-h!抔[ Fd'>//owwǙ,qfJ "ڤs\b̙ 6[__[k@YuttTuv=nd2uz5̐s  ;zq` i}_Kʱc}}RBp>܆D@" H&-tlpbѩ-F,Nk5^04`b[(0`H;Zy \RGi;T#Q[w裷ӯ<񵃛m5[V`ORz+VX[?vv4S+ Ax-Vو^ 2hD]L@}cCӡVڝ;wيG:;!<8ɭI$D@"090~3k%Z.g/{RVڽj;G]8htt:՚jk?t8QO5Mr9mx`ka  ,~(K.7x"`- ρ^,|`bRBNn.h%$D@" daB<_.Y< M\5?]u-_ڼ7Z[u|זc߮9UU5,q0DRT0K b,+̀+LJ7}ܒrCa X&ނn bf !,ZjB -fϙ=kVٳ}^hX[iRH63;2K=H$D`2WrG7OqEEbaoyq:}sgSh' Q1sŖ[T\W^!V ÍՀA *'q8]HBW00[PBh#8^7i˪Մ:P ʈ+8#& b VEՇ(aWmztAϊq(JjL-1|pl>(2KUD\ART7:A_}9D@"  &xpӵCB%ͤFf̘]h D^kFF֬YW\9o\)n1+t 501KdEY׀X/NgAwzAE>Nv Fbne'ͯ+]*w~?V.TC/Dxot"0iwCE6^" H$G.u\/D:D15k֤QUg6YO+[rՒ SS;;;RS])0@i'h5bP Aff +AnDZX109ٙ+W.җ}?`|ka,*fLXB`Ua>C( mms_}s˖9D@" 9dj٢,OD#~:-K^쏼;icb,j/iƕ~N;slŒ-v[aAt鲫e?^yxwWl@ A m|0+(>fY}C\ ^ c,0\dPd FeAjNHκ֛ݿ{ʠ85Q*W#;%Bzghі H$(uu՚u1*4ߴF7 A7s֯|~fTgZy-y!8a]DrEwj5lX&e՛?tU2KDZm%%kZ;r}MH'/+ʾHD\Fj" (Wlq2HPFC(,6Kvn֒ nϺy˖fB BRaU";ju=w+9;ɏ|KhH$K.q 1w_<u&һQjO4 M? MQ=_SOixUqMhEQ]8ԝgv%s=3&R3Xӧ}=]= ,<@V%Q3Rz0i Z z-P$11@}.ʲ[3V\~=/k޳g쿶煡 (A2!`p֖7𯯼 =?y2*'D@" \Lj]kW*XD O9 yO:xӃ߰6*_(OݩS7vڝu MY r쯰߬\boA@0b6[dHH<?%]A??~$w(V " [Boθ_=Go\lnX~_>|'8+D/K/k~+G@Z" H.G8u[6V@;QWOq{+5񧜆?Ҽ٢hV BVǰSY;;"@ >j;\ӟAA OWj~@hCCt>od&cFFԩ]wkCjiiHLLF~KQ,?>RU^{"~#OW\ն$H$I'7V̞ZaMOn%wxTy5fm$Cu*u5fӱܤ2 kwv Y214ig ['B޾M)__1s&"~JE D @ p @Rz"y+V,)(됃lٲ!e|ݶmcerr2?UUuW_]+!tUx/w7D<~+ lˆK$e PX '"g ;FY ޚ'jֈQ/j#{-KAMvFN ^ú7?.vvÛڝKV",ή֖_˦g6E{jƘx$$@+ۋ.amE( EIةS^zUUUUӟW\q &}(-Z|Wdee{nw޽o߁ .ζ^q-[kgMO,_;)H$! 6 yveϟ5^io4v"Sd6 QEZ -;Err1 o7@z=,\poS+jժN===~nNv6ѽ-[T͝]v,]TrIاd%D@"p|#~a^~Zh^z7رsxdB5Wش:]ņT*xX+=g6Py 4W/Ž;F,Y,J 477G'O_rUfvvsk۲niwuDzeK/, rD@" L"$sDkb>XQS6|'*+^:R`Ȋ2({ŕW ZTTr$@+T*֠\TΝ;n[jU{G[oin)R|->{;ϙq-ub\D@" H.#$s~oo椴5t7nܷo?3֬];̌ UdpQI+殈Wťյ^|B|FFƜ9W\QS v4._:nYTY/^'Y" H$gdgd]󻗺+Ο9YS+v bVT׋7s%Q\@SS] ;mZiBbɺ{=-Kν瞻A'+XD@" \Hz ׄΎ=}Gf̼v@0*dZ\vE((,V4p*(`2=@w|֭Ku .p{),H$%dȁ<݀oǿl%b%!VrO[#bBhԕin`[ݫ/,q3>8rAD@" H! C~ۺ\+W.r܄x JVg1y:Wc'pm[]XO={iJ1K%D@"  \ץu)#%M" H$ drpظZ%kZ8g1N&٫8*F8Qۼ}];>҅u))MN0ꂫW.ۋK'vK$D@"0 u'N<4ęHriE7xžVMI$D@"  $s4|q .LJJI$D@"  HzA`H$D@" 7攓D@" H$D` $8HD@" H$@@2W $D@" H&N$[)H$D@" D@" H$\'qH$D@" aMpÛIENDB`}DyK _Ref361131994$$If!vh#v!:V l  t 065!` p yt[@$$If!vh#vv:V l  t(065vgp(ytXIJ$$If!vh#vv:V l t065vgp(ytXIJ$$If!vh#vv:V l t065vgp(ytXIJ$$If!vh#vv:V l4 t06+5vgp(ytXIJ$$If!vh#vv:V l4 t06+5vgp(ytXIJ$$If!vh#vv:V l t065vgp(ytXIJ$$If!vh#v!:V l  t 065!` p ytm$$If!vh#v!:V l  t 065!` p yt:$$If!vh#v!:V l  t 065!` p yt:$$If!vh#v:V l  t065gpytXIJ$$If!vh#v:V l t065pytH>$$If!vh#v:V l t065pytH>$$If!vh#v:V l t065pytH>$$If!vh#v:V l t065pytH>$$If!vh#v:V l t065pytH>$$If!vh#v!:V l  t 065!/ ` p yt|<$$If!vh#v#v:V l  t20655p2ytB$$If!vh#v#v:V l t0655p2ytB$$If!vh#v#v:V l t0655p2ytB$$If!vh#v#v:V l t0655p2ytB$$If!vh#v!:V l  t 065!` p ytB$$If!vh#v!:V l  t 065!` p yto$$If!vh#v!:V l  t 065!` p yto>UDd U#\[  A?zPicture 1Graph extracted from Ofcom's November 2012 report showing the differences in performance between particular ADSL ISP packages during peak, off-peak and as an average over a 24 hour period.#" `b eSx]xt*SneSx]xt*PNG  IHDRW~usRGB pHYsjIDATx^ \Te?q5/!x! %-tTTn)T붦jKXIl*F薊ڿjefCTj R\ϙ33s>syawKׯ @ @0 A@ @ @@ @ 0.A @ @@ @ ]@ @C@ @ K @ @}.A @ @ K! @ @% @ @Fw@ @ % @ @@ @ `_K; @ @Ct@ @ p  @ @/% @ @!p : @ @@} @ @ @ @ @  \>@ @ `t z @ @` \@ @_GkʒyS+.Q!sug\k.[(lH16|]??j*̸}Ш;";qs%/IrN3(h)-X}֢n)3slXx7kf͌GHܛ'#.:?#k++3bt38zVd^b~e;$'3~ ^'~-+m̏i h޻@;=eIDvVi*$F]Q7:*1ڵg?o'I䉓'q9S.6ţ^bH>塇<0<:L/QIam>g~UnJ۲6[f-JkOTF@#m-c%lr9 `W@X)ܬ@;Sw4]bW~ƌMS١(:K%3M,*W:~Ѷ+Xhb =REE 83aNE{촽<|#iT^U{MU%=Z{%ϫkN3UKVVj݊%߳Tp|=T 6"Se @?E Et xMrv۟{DF@{B3R~:u冊3vB6g{OMKrnFS)\t-ɺ7-KܵO-nrxGf7r/ 2Y1Niط甍M8=kwD @^4>SAu?Wۭ>۟B-hi̷Kk>,sKY՗^l?LyGiOG{~~LW2" SLIy^ i;4f7`ҧDЊ>z/›>¿\DU#{>ٞ7因<5[R؛5rbA7F5-iyoϮ>5հ<UjYbLڍ;ƆhU{6H' /7ҡںkِٞdQ箆{]KWxGu=fG Ѯ[{3 5 @A!1<_1Gs`H"GƍwJiվO3}& M_qgq&z_JK뭿æ|"k K3fwxeoġ ɹ 2DLm~ƜrZ.oN-hW+tr&GF.:"Zz%;俼0cMCɢKL?Z.[v./mRӶl X`cqx]]nU6F0WzW@:`D.lÓ\ڸMmq>уVBboidɱ7\BF)/}1]\mˋ=D>gCu16xL>KOyݣk%/۶I5F{h=]-Nz])\Qѐ_HB[Uc/H({wwf۵1bù[VH"w6vu!{ W<0aql㏣oӫ3Z,ݕtXA⩫%^\|ôݳmV1y'0|sKacx_Yoۢb ]Wm2< OYm [~`! ({&#WZ;֓]?.t{Lcw3">ʸ򫺳Qܢͱ'?åq:jͽ&")︽Ԟ"cC<y(孇)W>Η<=Iʆ3Q*8%d$üS_V9uk«idQ ECB.Y,[vZ8zխ F>fړ0PО]Jd_%8W oL|!?~SVwtF rg~XˆzqfF{F<^WQtWoLIM{&εhe5%nt˘\p Iܦy[_,\=ofuty*ox,Vh6<>+|qx7sF v͹^:S$+RWh ;3Dbc'.7eq8^랥'$Ta3-VBBI^Lp!V=<KaܜB;鶨(ʕ?iwKigǙ߮hO7}ԡDsF^Mqfr?[DGS%7'+e!8zYײ{]Μ Wf1Gwut%}1b uwغd?"au17Xh]fEqxpK6FV*W,ն[ljȐYw] kR{UUck\R2yj^hu Vo >n)bӇa-6f_? xålOi}B I;schqՇ6оt7}g I3&m/PTԻZaE'S1@x]úsHWu&1C k bϹTHi^57J'/2䥮k7Yz"ݯ>[X*Gxw`g (2W_E~&7V/G]L*}0=sh@[O.Iqԭƶ}S-o [ sG66f"5Zڣ8.ձ98"ktǦcW.^PpDANbd?˗-&wwTBq禲[n)׉cP3DT"a f,.11mQѹ᦯˗X~9I_shKc@9'ݺC.]ֽcأO>c@x`'o.'yg}s` xАa8c?.aȜg1qO &5mt)qeӇ0a\y'MxSdE+3BEm]r c+=2\b=RnTȣÂ6\B{$ܒ̔_@#?<6ˍ"d@ӯ\PSɷu6޻Udf' }<n`-1C˕u'=f|NMMOJ[Xcwrs&C{?)9YO1&˜B]nC=K<,o"q8ƳÓl;bmbZ Z'42LB,k$pd,{0ӑDL5EW_0/gB&}p`O|6 k}ܻ9{FRĜ'o_0R9ۿOrW+"e>$+=D|f|)m^D|Nx }]bp!eVcٯTzo/ rdo/nrş.IVPma)$yb>oJxuaκd'Hdnt';Υab>[4ں:2]% />ixAڹOCo X\^(9^Nj s⮍y_!SHZ U`f,G};41n刃 H;>"O#,5ݲ+ʱǰڳ풳sff؃ww3{Jܮ=X/Wz+ձHŵ@?ނf̴Sn!wض1/7NE(%(/;/yvke=o$yxY_+9OAd/oq!qOB]"%?aN )Bv_55[fVSi;A,^#tV8kmj*^%sPI<mQ֋Gv1VOd5g~/o*[ĉv{eyFz~kGÔodDwDk0 ^K̓ͫfP(&Q}m,)96[6 "謌[/!)-Vb%-r'IЈAmucTh~uYk}Ebkx2o*m}}OGyƄ mw'&=x};Qknݮ=X/{WWc'km雠u fc 岽0E(G]\GI~O3{W-#Uد [|c<0qmJ㒡#M _ 1.q6XSzFnv 䰂e}>Ov0 ćEdL2 >c8z F.Ϙ`&>m8\W^RDw@FS2k;(OI-T?rabbF qo,{{;bhGddoO JzNrⷳ,vOj[5阌"Vۋ6ܢ N5Nݐ$b@@6h43w/U9ڃNaA>΋M%immpҝ鼇; c-LFnuS>i=ž>e Gg/INWoXjkxj2%5^\K?{S~<k7%3O &t0x~%?~^˸א3?ivMY96㚓yRϰ~4F2d\=mgPr:i{x#7뢊/]@+U}M+O?97wն^"Th7,L3L8$푘n_tWJzZ,,-}])5X%]n@Ğ;ukƦjL|XYe[J]٥랋DI?u;Aٷ[wA0/NIĒK}4$4k71cF6cCרXS@ٶ7Geayc=`0gDO@>۰۵ž:,4أ\B?g%+ݶְ;\ű_1q±P3P?賦>TۢѲ帶"596ʑϖiЙ(ײԨqժȲ6s>=9̧Y#3<Ӣ:ӊ#5d[jw'9OLNV21Kmי\ړ_S K9bn_7#Y,G$ Ofsm?o#}a>\8rSIVжb]Ohܳq XgxnùFfk˃q'u Ippf뀿alm憳G%9=pҕv>mUӪn8;~\ti} )X~aWbnwᒎ7VrOMVsmrG2;w]C'%Ӷvm8l"ջzk|8b"P ,W4+qWxF: ʹmiF&՛?UgCn)_ cT'L;$nzynCrR#.k/ՋKut$.[_E/MY;{P2Rdiq'KlS-ޡ.g8_mr$f U:ȴ~Efai};&촟>yㆉ̭L@kʺsl .u t!cdʭ3q\O| STdZx|?VeӶx8@B#P#z #>́Y+ڳ69Զ<:jO}㽳ܚ"|aXO3M+YHƧ ..WY,Cu~&$Iʈ!)hw~6i汖wd}a6ڼTϵY9hGY^s9-e8H7w WCAu9{^ mOݏ,O~7Na _w: NIAlQfSYq͠H>͌S<Ϸd?ĮyH~x+WWڬl-- SN>0oݦ*OwⰄa7*WocBvGy(Sݕ/lجPo}]%C,Vl2[FkeoVu2+9v[  6<~9tAA~JC'QW+YM[5c|SSOhJGv-VSȥ乵!)~:Emo轻>[n5bRU]Wҟ♍iv.YT{qg>kgoHLl|>?5P6hjg%ϱ_6hjWmy/u}K~I[EkaFJAwZ=r^M/V]Y,O÷ +jA/?ڰl%гPnu-5eɼ/EɊ29,LW-2臤WΕG3*sØ]_9?tvGNv\, OyXa,Zd5NkMK,nbrsi-Cunjp$zX6ƌ) c'o\\!<퍢ކC,;d|kN$CHy2;xL´9#;*WfoT~Rt=zlNmI-I[Wdʌ[Xm\q5z9^T%D^ (4- k3IOx7kf1", 3g$+b0~pыvZB}"i`>vs0NIzzaERs7x9þ\z;t'EAN)ܝL-]}ۅų ˕?ЎKhoAmhy,1h*aqq91s,⩫_:MQtێTL_Y>L9̀n}hAwYW'ˌ-jmh̔1a wm+tH}zHI|b%)zg.J\x\@u_{F e4v$7N.e)_pla9 'dwoX]5|-m/AmH߄ZT5pFVKO"Oo(_̽ElI"q9{I 0H Gs&8uK58σ윝gĵ_I_}Ȣ^k#tŜ+;vt@Eȯ=Es"kʩsm <1h>D48w1H_ =rG3Ⱦ˱ƶzkN9Hr>!;=7-ǰS7L>a6O6z9.k/3 Į;uċٽ[+|:ǦK>dͿ8$)3jhÛT] Цx1&,<=y +gگđw,};ewQy$b@ "-1C{Ыc/5۹6m)q- HJ)2* SEVY$w`a_bn~^\֋)tlO[PLܮuv!#0 <-B Y(=ҝ/-l+F UOaLY km4$aH]1@ } C R⬒EC]Ik}m3rttO@]͖2.?MtRܮQ~@ A.Ek2MK񻋆d̐ߙ3Pw !_ b^=uhg'`sSH?.7oʻFB  9)2?,u;Obv͏v@) @LBjrOݫu-|Q5]D]{` "eצ F@6.8 @  d**@ @K8T@ @Q`lU  @ pé @ %ت @ @.ဇS!@ @FKUQ'@ @8t~:ӅviEŊ+jHܛ'{lբ#l$Y 1[DFQboo0~mڟ~@1T2Ade5;?lBѦ;&4fΟǎ#rJyuBt C @ h\~XZTm_Rx^6~{ ?*7JH*嶬,}Yk] m[mڔW 2?6J1mu?`+̩W8F*l*JXY= @ [K|X޼ZD43UKV3)X[ݪu+|o1ў|Fx|eӖ+lY_+Sɶ.$C @  \SndN@i6%vЪ+?}YE4-ś>LU09 Fh76,E`[Ng!@  \{sBw44mE8zŪq%1sqhhrQZV}So'ں_/VB^)i95UkVjSmuw6vT(*Jreab'i~JyF>lx| =Vʜi@y%Fo=c%y[c%+Rж.eBר,Z{QR#vzƃ<=8U|>Gڰhi汫#Շٸ<Ö5O6gnKػ ն)wn4G JLm@8ETYUdBW,ר|`E9T͵UK5p#\lK>\RMx;u`JK5XbGq̰Ls5adμDoؙKU{W'uey6~y%^r{W u`#½Fږw-k  F䩉_,Vz9x]>mؐnﵧфEt 3%9xV?ʲΪC @QD:{B*ƣt=[WUO3˝J@y8N/Է*HqaBXkvlj[&#?Os +k*>B[\K6(^reO]iasQlm\#dIk9WZTNnSexAŮvJ[~ߡD$)3&iΛ'*I1,'7o~|uҼ~ab^nk2t6&V)IOS4eT+n:N9YS @ (\j UAcBOn5z!]qgq mq`meA]/:V{jE)-j)iy_zKQ+lW)+|~tm[E[uZGdR"!RUc?ך?ؓCMđC^2,,fΣOd/a.$N+Mɿ7ܜ߹6G]cycb;dgdɮyR^þ=O̙8~6L5+w2p$o\\KӴ瘽: 2)m.4Oֈ^Ͱx8"4Xsit f1&9BccQ6%77e0^ @ a=x`yKoN6~5sП5Me[~vOmvF~ZĘHrdԑP˵Vs2r\F'ƯbMKhS "\TŰ}QVmeN[{)So37LӼ%eXL \]ԃӖg,2&K{h7t8*Q͉4Dv0pŮ!Ҕ-Ϧ N6cq }SL8LXcgu3 *qL)o=m Cg%ۃ{e#H$'gqEX݅Q_Ո/bd&@*$8xí>7rzH @*pIX|XzdOO"ŴЛ(t Eh75M?m\@1y`f+"cC&#SӨ gΔ ۓ76eUkGk<̔VđL6臖F{cq;'fk;73/+Q}=lھz!fy,|4Ij1B*{),p<۫kks+M*z>dðF<^Q2:j)v=jsŇYyj!CϾY/v){@)gk @ | ^*jt =Ue+~vqOVFb40枖 d0Hui& ȿR2Lqϛyt^T愂Co~p6.5=ga]C-**{s'OFӧAQ[l"[dH3fL,^VYe{!]b1הhc)f@Ø# zm8X[DE0xa$^z3;cRG-#f-ցnjj8z;6γoebC-'y?p @@ FOԣ;)Q]z%J֒3Q{~PhQ(n5|ɺB$FDb]@qІ˗PQpaǕԙrq2fgвU)e}X9#~yqYtHxs|@l{\k;O'Ocd:#f>MSVN-8[4Ma-n=CX^lLxHKlw`9dKJ=~>#Pj#9@ .T^\󟢅ܙT|mG=[C7{^QӎҜmP%kzog]Hmc@{"0gc4;ωSi˯Wb7kM9Zv$i) Nk+aN(/Y ylXN8 @@ KȖ4KV,Qw6IL#|9}K3fn3yt`yU;^-j[EQ7YsqhlBrnz%n+|we'l]&s+6#QJ3l0H?~^˜:?j ųϨ][3_5"k2Y:M Ҭ'1jurf8  @'15mE5{aSY5l(~D(t`3#Ȋ =p¨v>|4TϲI0ưִʱ%1VڶϏԑ/ݣZ<#= s wR]scyŔMSA~a.~%o# c]O5۝C-b<9KNkh1*b];IFkJӧ箙\\}hkX_k+IFؽнrf8  @a 0G5f'OS1Nvj~]~WVvp6JaQNǑ @x.T{Jv+WSTDn%6~_YTW+Yn8 g2%+el,zf9"YDvzQ},:8,auU4-7*gPv\GEC2J$hʔ.V]Y`H÷ +ja/?ڰ93}gj ๳nwnq<.7~ 6OjWmy+u}K~I[fak[f:E!?^5w /W;nФAzN*6Ҹ& jtff܍4JXmd-E~I&EdE_Y˞]93!@  t~ZӜ2y-X}լgN Ӈxo\go`U7{Y}xj2 MlyNN+2^/VHܛgx"ת+͚Ye|ϙxP=UKh1#'Z]_i^,9Q䈞םDc6+Cɕ\jQ)3$rRd1VBp:G?:S(8=RcbO;Td}X!L&>>T~ g8r;RLj "RkKIs2-}s”pNڝNǡOc[^%XFd+VBNggy#;)Jwikt~g*W,7q3-Iĵ8ӵ p @ K>K:ǦKbkߐ1U[(r򠎟7;k|) rovš҇~64Hӯ\5Ն18<~O_`,xgt2GOR#L^q<ؘ$|Jڢ75q'tG5g사+a,'z*ga 21yq On@b%[f%Kcښcm#;Y/g 8]8  @%]s欛,/) ֛`DŽfg,K4ֻyd.YJyIcyκ'!*)=^B:|C&"ͦ ǢvNlmzP|4{4/%=n `QFq'3U9$fx(4fΣWHY{}n;Y;Q(_47107{]~dyâGWQs$4ft::p,} ;3bމ‘R>X EkiP}fְ.me[O/v`tJ8 @.q{ BKǕJjE Te @ IgKDË?|~id5/H ċj @ @@߭%6JꏋJ3  yvӂ&;[*tA @.8%Ҽ"Uz{91y!l|2ǥVQ!_9(|=U] K;{C^'>j#;+8[ߓGoUi4@ @j \95$4&;sqn8} qh/Yz8~mN6 ҆Th?(uzgJ *9m;ݷ4?~|sZ$=/ DdJΚj!@ @9]#MV !=8ACe=6Ex|+xNtcCa8.QXQ]x}WDÆpҹ$D3E)H' @  nECѹ'}a̘!xHmQ.-e4gojNPWm[;YU*-,ώg;1)$)#tӷֳ'Ϊ0Я#\̺/LsH}wQ-uh!jNW~ۤE^l1)lْfdݍ+Vv5cbgA.-/*ߟ\fwk/oUڎtu[,#j?RT@ @7\B6i:uô;~4Ƃq͐>0gBڤ˛Z% Q^ۥl\_i/OXѼ-ԯ{O泈lP+oUGHe)Ww!-~S4l_2uH> Iv_F?y&1Ha/ ?)642 ~ƴ?͟*//Cyѿ׫uHLh;|Aך_/;'-mHEMڏoڃ>tؓ , @  r,yҗ2I"C<ʋJgְ͓p7O{ ]uI-uFY>^1Q}Ui%s0KM[t]bVg&ѐHuON6J2zrbv?N~(&*"YCƧF>Y"$/355_oN4-;ߜ08ߚt ѣ,B$,W \9}Ω% @ ! siKQ!SRP,L/j̼2vg2[S犣.!u'|ߤ=gˌ8rÓIā9YI4̔,wmi4gԲa88&g3qHV%,C3"S&PCWDj0*v }vIA#_eYvIwIpux~ux @ Jc4d}"ޟKPfg䧥:e^fĄ6Qfq̰'&h/mנD!a7pLCO_Y{IuC w?GDbKCEr׌Y;=p8861e`wOО0JS)m;pjq߻^96Kw,?(âL*nh*VnaErD8 @W @%ך)^^K7 (i\_KCu|]g8XtWNO-3,8}] @ pIzmb`a1lj\8~fkMeQ7Cz}t3~/LcHYb\;pi'ō|f|_j4^kX7F7nLd؝?ȸqVׯ1OeVW϶6NIp, @ @ @%0وx*xic_6x}(˕u'm=V >Ez'N[؜ձmY"dtwPqHTRbqs&IBl 1w;,6&Ey&a*>_WQ3 zDFfQ*W@q  @ @K*Rtۚ*7.LUA~߈mqh ! mWuD$mZl4w뱍oζ",q-C DUKjOTpywz9LsAi㜡f 4)^r|gO>IVi%0r"ՠjpQXhm˳Ԝ쨷 rrg\jzd+Rp @ ¼Odݮ(M%w( E]_U**=B?8^?'զZWų! 1VW$gmZm[w婨{Ft6f03Sm).iև5e%\riHh Ș&{$+|mۦsUXE$;)º]9ؤoRk!CdCQߠej*j+[0N^  @ ׯ^#~38/kĝ>)հʱΖٰjٚ²!Ҕ-t۝*97ˢ1/ױ:ڻ#4'ƌ,+hfٲ#«I0`53ɱbC-ИNaR$$Y= Ű䜏&ZtQ.ZAB ^Hq?@ @n]Bڅi#y+_R#ڱohNiBGә*Lj,w5%h#V,q|s)r$Ia*bo huaȸ NU我{ 0`49zx&- =d 4fߞN+!=eQ!v @ @ t*hls=nlh̜Geܑ$sbFg]eF~YP*P"K)ƶXSw:c2;#?.'g|+/W-!@ @5Zmew* 4yyTśZ'R$ǚVTeY}X(L|Xd@ @||mKwħ#E;Q ^D/X8 @ Jᒠ?.'[iʸLWaxlRצ  @ <K}d5`B -٬PUKRВ! @ 6Q s:E$O~||fmKVxW?D"O A @@ \MVQeJC~@֭3`mWye*_?0 ? @ 99A@ @@ kPT @ p @ %A֠ @ @\.*!@ @2KAQ@ @ \UC @ dY: @ p@ · @ @ . Eu @ @*p WA@ @@ \d @ @ U8 @ @$Ձ @ @%\q> @ @A&pI5(@ @WK |@ @L kPT @ p @ %A֠ @ @\:\~k=_۠]sEV{=?SOyJo; @ @|pI뱍k4!%3UKV7ٔz䡑.(  @ @~8ekX×X 9SOjZ52 @ @O.6~73kj~a\I95UE[- ןg, @  bBxv^:'J(Ջ?o^R.5-!:Xi3ޯh@oC @x$h2W%PD1s32 qĩ=paw3ՑPCM Pm]5 @ @.\Ҭ*׭*=R@<0w7L}dl!@ @@K-u+wd$z{WZ[P-#2}x( =5/G<PU̐f(p{ 6P'?(F9=.@:+d  p8ҒɉѡT{b$&y8vC"@ @YךV-:" I+tsxy) c=<+GT-A wg\+v٥`"N_Ns3g}h_"$p VԢ猽cd w@]Ic.ǰ$|Τ<+dI @ @>lƵTd+bəszjvYA @ AK͇~+l\ ҡ( @ @ vbٻśtspLҸ0K#Kğ=0 {~kF{y=p>ɚ=O+< =ؤc=W@Pm[mo9@M g{d/m~ @ k h/{wIR.V12+aW~4 @ P@aKsZH KJ]Qe@ @+B6 I٥]]1ho @ @h۔;7 &@,*n~cIrT$dGrx SH#?Y};.wR}r' ~@/mPcy6ZsL Sp$Y$=?M(ڢV7l"خA.M)Oo.W*:'=mR('-Kup)Vmg@ @@yK-PP~ _eBi @7E+7+r @>׸( >A_r&vPAEYEqVz5#{/L-ǔqs 1'UG|tG @ @@DP͍B @ \q@ YL'R x\K.AO `C5[48}tnR)X!W\, x g@@ @ `CK- #t]iN\ޘ/B @@DM;X*&IN8% @%> @ @O@ @ @ׯ,NZ.DF{U,\%loK\(O>յٷʳڴXX5{"8ƥ:<{' @@MY|MOj~n`޿K ʯRDܝ\^  sKx@(.2󒐤yWW2~-7)K]\Nǹ~p; @@mG ,W>/ _E_{ʱU8 T;8%d)b4-(0#l=3@~|MPpoTgdU5/7Sq T9?Y f8ݟ5G @  \A @ @%PG @ʼn@()@CҪW @ 0. vC/@6錎~>  t £Ҡ( ?Kg!@ @ j E :ҕOҲNO;uΨhp%Ӗ  A^7y0A$X]Bo  bƀN\tPx.7E7pA{AP?%o"@ٯg#g >"H ?yQ·}OK2D`&odLUaߞSǗh; M. DE<,@F)*<( @!`>cG-~!# 4MeN8\9`HSS/ I\EUp  @ iğPqԵli P)=j\.C @!Ѫ|HRu3uB̓A["|/z7K.O6@~%wd_!\'݀D $G/}PB8>?0襮*_qIC.٦ocmSnZF4qSru~JqT;ZVbZRVcڶZ嗅g3_^UYoi׬yMYu:sW c+! UXE˅ 쌒/-Y<."Ol9fe 5rQr.ÂI0&˟tv "y|GE!Cd헰lBY^Q =6iT^<3 ql´9TjʍjmW%U$*]2\bўzuiɋ份豵U$br%IE% kն)^*ʝ~U1RN?s*9K3@-QEM҆RoK $sz{H aL? 3q'.N12{Fr  ^|Fʔ;aC,MiT,ީ=ŇcU"ϔMIbD]ߚ=boѕJ̜>j{hs˛>ߩyXp*6՚5oJIeXCuEP:`cێ"{T6*d)Ɋ`6mٴeŀ%_&FYQJ@}QE0ӮH >IXLmݡ}HA=}洄03846!)tщCuMȀӖ ~C䕩%&D6 G>i2*H|q.Wut'֗+,zܧV\hllT"T`wNb/mpf$x .ל].6~# w !0=?_$'!PG@ @ M5,gH"2c'Q&ɛ68$&{O"'"dJV^ULÐyY7wZ),~pPqؐ؛I ="ca|&aֳɣH?#ib,ûxv^*2Yu0L"4!%GϞ_Hu" @&ҕ~ Vpx# j=G?'jؾn>S;(*F9ֳܳ~:Y0鸾ļ䪸?Jq1,[S<;nDۭIzE</q%9ð7[ zsb+9n[]}=^I j/=*,B8=1%ɉѮNҶp),v]MZ՗^NZZ븼^,q @ۯxpyp  0:ʉ]Pل ,kBaoZ.eݵKh;?}=\?Nɵ?xCs*-\.[?^rΏ]$?N:G@ +`Qd @ǒ_BOoW'ϔ,JZKzF30S ?0w[M޾9m>ΉԞ8#\ݞ簮ֳ< P{z1v0Tz/c_a9W}d ԁ5D.g.5ZuD$^9Wi xLH2eF~nnfV]n*l.su؊+Ja]ٱ]I1w{ǸY?Y.)ϐX%'^o{rIuo9$ _)KM,>Wm" JDK_L7oZ_rݞ'׀ɱuLZm\,OvxPV۠rv O"yEB[cq%u"Q9WΏc>x|*7=h#F.X$TmپlqE!CTs~Z:g:t'! / tVU/IC:;OY^+ xJsX}X-I1渦y`.Ҩ" ߙ3b%mrZ*V9ؼp|$v%KV-ː사:Ǘ)f#pImg\P q׺ 8wdp79i8kˬʼn$@{QGN"QXlMwHL8uH;Y̝6AOx'$uͧ+[ȿHS &w6+~xs޷bܨvlpBb|7=fGKD.t` 尅vwdWg @_m`Uo qHȐYw-no>2%|TB"|s&Gmhm̹TaY*]$̥SDSSs46noXFĵ";SĿrS2cЄ순Sy9%"a9Mq:Oنk^.Rnvd7)|p Ox)4L\rbv%KV+*Ri\ù }x[Y}Ysʳc_}Jzmf "Iy\ ϽsV6^ɺ wDe3-+njr¹LrK@_=>|j;ʱb}z@_\js}# rDLKEF<54mښBjX.9\@ @S %7 ~+5ǦH6o-IuuN`4J % wnr;.T@APU qMO>OQA(?՝*?s{ǒ!le!VA.1ǫ.srI뱭_US;Iz =Y5NMͫ"0ѫ%$֠%"tAU ;.wg!~+RLA.kke^ؤP6hk6& 4&)dV0@ @@6.JEE  W?=6;`  @ A !\"M-K$<}ڋsb&=* @(@^!ޱ~Xxd @@ B:pog/KvֿȞϏ>Uڿ^+06}HD5!K~8(H 3~Y[[`_!^zc׫֫G:X+OJ[XMֿZ8Od# @ z?(  @ @A"pI4$qI"A@ @%o#?c?y#W@ @%~G @ O J@K e'ʂ"Gt^7'oW\~Eθ'y3E^ ?>۴<puMZrruE䩉Oן.)<1Sh:<0K =lSWU¥ d'֧}LJr:%))ώe ?Qk#jsϷ.Ne#*{HOoYYteO}҅5 _/Wշ,寯5a~} pi@?_}tF> # ,Ȣ {+ @%h @ `2?q<G?ī#]9%2q.nY_ t>HbѣW)p߹xZc2''b2@?寯>&p?`2|8~o. `tP8  X=2S<}OeAR&qysx^8{.[*K>G1#?́@YCYoQ2/ p霗3A|Vm~'ox$@n\vcQMP@%p6 @JL=rzsTuQXK+XDB@ QMn&oO  @g5?6~vJDY֭?%ihTm O> '@ 8 sy_9RΈ{v\^nJjO mK-!@ @0K`dZ,7PßG @ @$HՂ DFP&- *ǡ9ɩ -ʠ ?y86KEQO T^2Rz@ @$KPV_ m:ܬe  @x"p Oŀ @ @/%P@W^PU* K2#}#kM)%kqG.rYP@_l%o6å,'k Ȯ%'/gZ,*B_QB q. }r9۲X'황a&>RnR^Ny6d/SG*]) tyiO8ͥ{fe]n㥝pGT>۟]?; G=r:Ϩ'_ CKQѫyjH% k{ p!@A/pI71*@%@!\J@(pI0* 2(GOi@ @.;r @ @iP0@ K5sywwˇ!@ ?Kg! ߇lC~\qNoZ~ٓțeD%@.ٽ#:.7w H.rnHKPF]~VyB@q ɋҠ/3e/J'\~K R%GP~CnRˈA =3;돢C [~J@ HLRT  @B hkv8 {vFɗ;jo @ %\m~]++m&}$pQzEm+%6˻SAg$(YvȈok^."o .86\m^GnX7|FD}U-*~~KUU>6.&˻ }W6 @ @%W?ުJ$)/X*ҽ׬,2EjObR`ryK"yW7r @ %0=خƶ&$gfh@E\km>qI! @ _@rLTDU}<3uKCѱcyT  @|% pInH]C*"nn‘<;@+u@ @.),%3nګmmeeM"48턢A=.qE<9.yR} ηpE|\.7| 7!KDw?C1!" Idc0;cܪO75D#~D֍om@z&d 'L-r.$_\C`3o(L&N+^h+!]݌u^qA@ @ K5+ V7\ç { kj/FǑϙxPv .(]@]Aق^b[V|C7d}:FȉMpO'b5])D GhO$'g6+! MH~mNx$Ѵ޺@ZՄ\~7W1A)PʪRNp;% ֱ͟kP%VzGϜ'RuX;D/ Bt]rlբ#"Qu9{Z7vGN&$";+w pM@QVQU^ɥe(ģZV]w{fj؃Owz"Y:=?M,Q`OVc8k1#yipqsjA?6Ban>]bY\R_yY}m}nR^NyvtRw綴}K:+[rZ>dkL FڐWCl}#9l3,\8.eo: @  }ij\HM@"n=%* @ p&l8 /Mu{ o֧;=xwN+!.Dʘ㸙0' Bb2N 3]ұmSU{Fw:7 gsc!#"@vcoJ\)?X p+a])@@0 |@3A7lST&d#**EK'LM@ @ N 'P,SoDn(DeV+;D@ @8R^*'&x @.!m&Jl4!#JosQjxA @."&wZf)$Ĵol!Ÿ|=tNY8@ @A' p pqO]*ҿ?1 {B}%{1$ @ @@͂B@ $EH26&OK#??o?p B@7M@<(*W i}|r.>ʂ|Br ?T(%~@G YȾU\B \v)C\rǹE@i)  l̐ |.0QbrXo;EF ~YB iFz~<5%E"@&d]pD-H~.&J-\ޒK.! #:%krrʳcmrsr)Ð>۟~Ǯ>1#Q;Ĺ)ı0ݰհ>%Ķİȧ=Xı!E֑<{Nl"e/%A"/r'NߧKH %|/p͑c`/e`J @ @  H2(ȷ d/3(jJ@ @C!@ @%Bhe @,*۶vd{b2eZFųSgT4;;횲9*,-àxgXp W"6H>a q%L\lwz}}PWU~@1Rn*ʽ{ŦfB&Woټ \yS@ @ Wbu_TyMIAaZ7..҆Y`G5@!@ l>82E+b%)UEwAκRF_;atqݘ Z@}rʳ)/y=<(R @*jI9s'&v5֯ktYsɡuWWR͆ έC (p+pIJ].!]cOf!@ МnLI Kyq.^rTjS{V7BzG#\L@ @/:GO{Y7XqqHK=dɒPe,Ѷ*,\>۰/K mf3o}Nv?i)\ -E5s+Do6;Dԯ}KYhi/!))|X_^H $I;JzT [<$ \T)(.ROA e嵀ZBFIu寕EӒ>fTղθ-Ud"ɢNju>,.2̾.(F˷Xz[/$)+˰<3O7^u9}ȶ-R$ @GcɛGBQ|+@S}'r.ʁO<ʖI7QPО޿g*rt VU!KM&swtٚiN1/,dǵnպK\D{rF2  SW/ˣ(FT,[ïeF@}m=NF~"3@ ?EfH"2˺[vxa#WΜ,9۠4?~A1:-cTş! - @ |jrY]; Lngd+m*dJ#'z&Ɩ:T2%92" 040U+&E"1~@jؔYׁO=:a~ں&ںXjMw?C (@v-SJp 6ǹ@0 5D T^r)_n5.wR}r' pScDeDM۷l&tB>3|€0복vܴǴq;ݲ;#KԱ˲ѱ-&"6/ѭ뵱s3.٦(֫9TKk;(prŃqX u: tj+[95DѬdA$g9%,*i=Z8"o;KzGȸrnW^"ӦG0ִl*/*ʕca {L"\ e@ ("n$E=}ޞRAܬ*^D ZOx_SQ* ibPqAWrE"gT4*^S|S«(f\_!l/zlv>%qGzq xPTrqFR@0 hj+Nm+ GKR77&ژ&tn5NҩlQ{"=Q8Wj5~K/B$}E~=  @w6W$䀊12?kqnqffꦉ6ͼ]=?8V}Rsq?yI#uh>=N6%qG @  tMh7rm/S?(ǘLhZl.>x!]bXlLDG˜өڴ{TȍvG65Ǭu%CNF'0KhQ @ @`k>,mL6^ؕgZt$UvնKn2nCC&),|gY>*6)` >{pWl85·^:^D$ MmWi[.unfHo7Wׯ%cd @ @ce6q5^@8*)[Յ)KM,>ձp GA$^\|Wo ݬU6"ɓoH4$r<8&Q]m-K~Iĵ V{0k46g kt SUn-Yo\v8~y=8Bd=5/aU m^T  @Оn/D^B_TXhܙX 8l#VR ˹7j֎Ce';.$w]-"%dպ,_1bY_=.P_[Sđ  @  Tƅ#".1jrlY]g_gĘ;nCNnϷQG1%J-n)-XC!Șojnge}%qr.q6>߃|\p.18 @ ALўvb3#ʬ|[1ٝ(Ͱsk3:JV7@+@b V_/prwrI ,CD@KlV=93cYDv'JL~?m?h `=[BzϿg]A/Ԗ7BrwrI ,C\Ɩ@PX=߶:L]B5  d헅\} V G\;Z{Z"(d]K @5T>   ӊ׋WD!GNh;qsgolKzYA"pyR7 n \|\#zvSqx ^lkЊEߐl+˟\GFuѵkwlm%r<+`%z?%E5_n߼NCJ?[ǓKx.Ώ?En@ OUΞ&dhOO5JHiőqΉJ~JyU@|+=a0oё @ U9EFYڈ5V,WR'u3b8@pI^%;xVkk;L},57C @@ЇKDaID"M͟*/Y_F79!@ F=:$@@sxHԨxzMqYMd#aEUO$_/rʳe @D^90]P ,. pH69uEC2tDWXNC=D:aޝvVzrc{ҭ @ I@2覄ύэ1zHSJvRs=Bu~ @N㫯tUwn~v!RY”_tWTA ;7Fxӭ72d;'  =77`۴ jVתұ] ƣpcU!HGt, @ׯYP@ @ E@q\ ne@!@ T6< @ @=K7 @ @Ct@Vy:8}tSB8.!>;.wR\wT+pr@`\q\Bn}Rw\ԓ7  \"C @ `)p  @ @ 4K⨯_ψ 8 @BᒠhFT @XȣH @ \pMC xQ :6~)/恤!@_B @ @w];kpk^.[pF!@ @%qz0 (7+6 n @ `K @ @ @mzEYt+pO ˟t8VLƌl>*% l kqV t+p`\q\Bn}}O+|y 8 @= @ 0.A @ @@5"yjk(@ @. DM @ @%aD" l={*!*@  .C N KvE\QaFMz&x Jfeq >D \";[M~?&L @ (LTs )G!@ @ . 6D @ @%Db@ }N|I듟_Ԁ.prYD| ^.r#?&Hn dӤO >YC@ @ % @ @&ٓ۵cq @ FO[&/=, @  8\mP< iƌf@ Pڇ!+@_}U>eڴm;kOiIF`P QD[ r-[{\H)@:zM.Bna@Kπ3@  t~ m-K~٢/<퍢Pl @ _r2ŕX/G y(7 @ `R͆ Ftkl8 @ ZjS{V7BzGmX  @  hO}qa(Td}},akp!pdE~ VFU/>@n->C)\+L.O%Cl.ܬ$?ٜsBtM炢*{ ƣqcQ'r'}p(  %KߦhILϹ;B8^:A @* M;GK_LuL55c RW{#^ R>y4/neQ( Ztw9! Y8Χ,{ffP(2̍q(b?9?Hp @ `@DUn~F"OL\h @ @>\riͻ2~`XA @(khT\V*_B*s7]"*C @@ %m5ϺN7{z4 ~Tr @ @t@ @B@Dh-B @ D '8<.L`Eٯg8;N g/(Avzg Q+gnvv ~lEݯOjrRxO P_[_` r8v -J`(gWtq@/F}  ? ^@ @ @ @  \>Ȫ%X5*@ @ (޼˜g^]hƒu*嶬㖬-;| p"xCr'NߧKHo4/@6;I<^B\' Ol"e/"J{D&NvFJ"SW+Teťc26<ى-Hk}Yꬪ!sm%upEFDK_Luc[ apuKzo59>~L[g`@ @ Sg+dTZW7.;NQ"o/NUwVmHV;m ë۫LEK*sd3'5uqTEea/)3s\`gi,/  @ssu{NSG$Lϟl&$g !i*;tELHe_BsAi>V"_`~.VB^z|F.PÑnùA0pI0*!/y(1$@ @LTD#vGw}R[a jW%37#s|mMG6},p*^~Jy1cCcH2xY z: %4\Bkqx kG@-BkhR[ED '<; W6uCGJ<wM35Qʈ!ݭFS\ԇ6vi)\DsgI YzlSV4^ϳ sU*mm죩):*)u}Ƣxrn7;W|a-*>PEΊ! u^sO5 |f+yCDwU xOP~ wGlvD"ګmme%`#4Cn^##I{ Tנޡ6Kڵ.snՋlT:jFyiZ60TgQ[M]V-CK^MֿgHĄ 4wnAVA3-y{mN;{sswS @ a:v*kC}~݇bBDŗ2uC3wƸURFwĥoHNu\Rwb9_+JQ٘Cbɖ+l8ZjvRjK bז'No/yzxG}:3B5N @ N%DL^dR`y%jp+/\ILƳ֓291C$Q6Vu Q̜skST9 )m5kZ|XC$+~ferbB+1Y9فRZ @/L,3NcQ1uҷ}5s] Xg$,aѣ?ZP]Ĥ뭿P+X6q5:ucSt~=2ƙDdS_If-9+%|j g=zЏ  @K5W>%97B?bʔ2Ԣպ{Ӱ/>`{TN)߁.<&ם/ hO^#?9 ]$)_>@Ѵ޺WƝeLcQcP 6v|XkȐ! 6|}SoW_jny؁6v})|lI2!\2B#=?GȓcF7/B\\1l!_p%Tr2G'pTV2a[ҫQ09ecUոCn~lЭu$ &*9X=S.q}nPy %wjo%9=іa۷OӰڅɹS彽.ʊަ%<^]%.Q(1mcWSA=.e;MțfxSiC,pCHfo5r%PvNԓ7$pr lӹ-> U?W?r&@~x=P|_ oS\[6̛'m <{^mtiRyH]V{NWM-Qb%"Q体S3v_!b (*fgʟ1xnQtv(_vs( w{ⴷRS< @2ҵ⾉/|>N*҄f%Kz-ssSeuM~-V"F{65JV٫@mUP@ @ Ȼfn{-`JʯK"OL+ƶXJtY$91MH|ٴ3orPRF,POf4=@׽  C2 䩉>YH^AJ3ul(ΣFYQpgTIy9ѱ>AOm5ϊz$fMb1O}B @ WK2( @ @~d?#[@ @0n`t&@je @ @@PQY6SnVw @ doCk⬒* 01 @ K.eP?HȌl%ȏ~ 3#KA#σ/@n'ʂ"@~@o^T/'J%=K#Y yyxuէVBr ?y[B %>@F @$0@:STb<- plډΐf)ZO"=@ @X@ ᒫPWU~@17J-(2 ^쵢SΫt0 @  !\Bb%^+}R%J@ @@[M"|׸;g \Vo~]D>%9н r%Jt׿t2 @ @,hU{&HY;1!F'NU:G:X]w%-S-wsAI`^U^'Nn(C:, A.ќnbC$,.x? 6C!}&P_[(YvȈѱ=zaP$A޾ M?Z!7|)@.)J^ %zrrlg[a}A_J(# @ xR %Nպm¢tsz4 @ @ :\~=+騂?j$d|=='GJRж_+v٥_Ԁ@3]{tL>)ⴶ8<)oϛzTQXczBkĀ(8'k:cDN]:qJ'I@&(go;LdԊdMն),eh@JLu}U??@-I^1s5P#-vZr:+~Yq}[ 1qꓥwec%䈶0Tԫmn,+Vz$}ڋ%c%T*%}P˦Eݮj8cW%t (s}Ct+6sF%U)B{%6ծ R}T:GlDI z(5< ={HI@ N- ~JyQEÿ\;s֋Ga#c|\ύ! /(y&X+ ;wQ0.Uj5e?aaVojmuy%kM&0وr xоCv݇Y VM.bu32W1Q@@8cC!@8fad`zQ0'LCJΝ5%Ҽ"jǒG2cPHlF0YXIss7O~X>*EEs!j?Y5^tupq u+9GJuH,tcBcDWN9 Dח! _ş^MF}jHSLŤDJB&>GeBQ @ ss>1 2lQE0G)#VReAY2δRw\jڂϧIET_;ig+C[Oܵ!Dӂ#ȡI 2P^ƀNN%5:q̷L!I# _q:qt$e ^[ˊWgU$. i iFz~<51ˎ"{@@QVAv'_x -$h+{kϰٯc@k9 5ܤ@KUk]=l亜Y=u9_߸<7!";s^mYGDÒs>hw8ȫXɶ>h P$VPEa!@%0D#zK 011 <-m[ ˃wHm/a|A,ӉϟC^7XE.5&ti=ZȷWTHbk҇tqs>q }+zK&Fk֛!HD'B^2$[~IcF0Z\={?)pTe.8Kt۸7193q6@vyu,ilΠFXוD2Jݢ2ҨXbmnYMF,_nDE u̝!kQlӂqԵlV̥M\rY8(p Y.%I%ݘ"Fg-5ę UD i8+.BZSē`xѫw@]?,j0~D24y6l6l\D'NI7+y:_DUqC[)t iki 홪Wu)uĒEmcH2`i14 W4m*vUWz`} !\B08Dcԁf\pȴ>mD1 @\0Dl&_5'nB6=xVkySM9ך1m(C/Lc(̕+rjYM͟*// /fl8caroֻ_m^n13rk=qOJ$QH8;6x{*.llXHK5f6 &JbaWν @ @YʺO#j38Vc%zP36 vH,sbP_Daq,NŔcB3f./6B3%i іKhGj_-8d1fBȧ}.%Wv֍U&fڜfnjf#3@ @s>|s! c s'@=OדQgR]A169Zl`Ce٘#}{.r|/Yl""HfU7b6V8h{@c`&U!js娻b2A$%V0\ s5/UyȐ/IKiP <p1Q'G3` TGk@;sWCu;.,Vl62Ley({FŘ6awMnq6C{h<(ʖIa,w.pQUkF)AFPNb5b,O2.o( ӃI/(T;tI8bR(J)'"ICeQL[e؟2]~!%Cwoy$=T$fZJ Dޑ}nNq- Y=A@@@@\@GXdZlu"zQyV~ȿO/\w,Njqv#H~5Dܤ-ر^(\H&13m]jm(7%еGWvaX     $s𐵄",-MO7! IO~%yJ#jBHa8 ReEܨP %r]x3sXЏTLv`nlHv8+Sǥ$ z  tÑA!p}ƃd-nB(a#?3@`.i93 Cɏ`:p&ԻnԸ>O8%B>       AL 0"s &%@"nBWLKq@@@@ { B&spz\@@N{  \XGo &߮ە'Lut˽z٢=vU0pK0@,TzX3c,A3ԍ,O![V A@@@@@\%.@       h08(p#$x0 <ɨq VIr @v!A`.Z` E;|'Jn[<|2. d;%        `DLsէWJA2ͯ8J .qEbAhm8?NDHs~MABp8K  |sdFpb    @E/  (=+GP`z>,B#_h7/nQgеGW.ݻ'c `@ue5&vNi     nD8nL K+=pF      K $h0 vt        d.QחV?{4=/%_^d K`݂,Ӿ2v$@nGu{PW owd @@ڑP%7D|BѨŭ@9?bf;]8aK's'gWWg(W␉SL      F@;u`m%YKNgmx:-1=ߓtbch      ' s3sp7c#z (;t<^d~~.? @@@@@@%`.Q_b%#+ע$UN߀?NKʂ #?(MC^!H{?D/c, /S @L{̌.% ې2-[De)!^MJOܽvG4=H:0B'lN*۷Mq[u-      .Mq3S3^cf|,&Krj¬ճm*6)ԦzDEgLXH&     `@HOa~8۟.2tNl        `H@0&eC@@@@@@@@@Kcܮ^vGԸ1 &Ez1!OU@@@@@@܂:xuD5uSpSe%M ,qŤmAL(6%@@@@@@ }C}}`@_x b I#Ivv@v&QL}gowZE6=psˀ ANQqӘ|jE9      NN@.z=4UODW^JQחmY3)I#6c`8|x       >3<)^`|n/IA8]3"@Ɓ I&\MVsd銝(9I>660FN2 89!K\@@@@@@@9XsR8 KF@@@@@@@9\z       NCQp08 @fQE+ r4gft-0iP?MU.:S[{ aYAZV -H-, iגku=IHyEI~DzEÃ|B^Vq 4$b!wKr cY,/sqXkx1AnMP]gQk|ޝVzEW_)iЀN\ ig(0hrid=]s,ڗ E2+w - `.{VQ) sRGRRv!l%҈oGV:e|͚VB@vFܥ斁; R4 -@G]ȯ{"G|Oůk`6ucayrC[ 鯩JYKU^{@  `.q-:r MU{c[6VPVmb+6Aը DAK@]{|9mZijfEf$WLAҤU>$sUiW#ofdu}(ԗ>yS‚mMue˶ɟbjZ@ n@8nD^Cv|۬<Q@@0`. T0j/-xSs d[x&n]L Ч7q}U⬽ˊv_3#ǵ{O75|Tޱؿ/ߑUPZSל%ߣ'jI"[]{ 4@dḦ&uWÇ. νՕ@܄%nHóo/``ќbךrXb+yf4Jb[!Sم%Q*;wEwK\WLV?[y2kˋĘ4[Ea ܹqjۏ4tg kj G/h;-Xb컵/n˰iF<q !%KcKjw5NHϣ=bn["僙kܕ;5nl үVfID%7dVl Sa09k)nVm{X濍 EOMIĜkwlIisL֋Ǻ/Ru %-۹5Ɓ6A@n_C92 ujQ,MK?>n6mލJZq^@VYn(=Z]s~d鈤Y>Wwߍ0 (:g6͙H%tqbZ*:|2=|DI#Oq}[AqTWܶl=ޑ}xlut |.9yn_8TF;|ӕW&<+]SE<>ym%fّ4%t,_9[]C]C6z3ʅT7פ3F+NǺoL ɫ '#eM\Oѭ-Tt4[en"e׉&Fm'dl\ u lʎW m*jB|AFȩ0y͟!p9bUA۪G~K #J4CT䫏6,(>hNo3ETt5F!oM^ 2!2& ! t_]f ր{Dn@~'G+*K}&_c DM:`M }'уXs$+uP gJo\[ J>#Go,hq _}œj ]ҋ=?=n3/-:mewֳ5^] kn4uj,(k«LDjp߈9P֛ zal6+I%51…m6O=`U?seJvq&+٭W+-d!JpH9ذ) sGSG G=,}sseRZCxw[z4ژHªɡ\D9<SHM>o+o*ڧdK ^>}?+7ȤAC}}hQ"¿\CK5IŴ R>m9bh܂ykԪy$viъ}Mn(NBҖd&%f2w֕Ovcfivf֧7njyHQ lIlb̼)b:SJ.|bN3ڀzucّOS݁7xD9:T] T?_6gKi}(%^C`y}WTdϽ_"aT$|`"H-%sBM=(q mRI*i}]=܏Eˤ^J(6gt@>01Ǧnæݻ9ұ7q͎!r1]r6$jlTpԂQ - %&ĞYq#1S=f" ic;1\FNRٹ &NG {So!͊aCN ctA7R(.}q%5 ' GI+f]Q3D @\YAJBAV)*r=UB7U~8[;22aU/I7fGQu5+15rH]˄S])b׆*x<Y85ߥ eŌ&raFyJĽ̼Qg=fc0"nRʾLuٜ9#߲D}Hbs\it/-~>w9SӐ.OHtFF][ 47I ~x\MA><&-L o6'n<[uΪ#qx:-q!TyGZqs^vgx\%$O,[p&DynQ= ; zyqcMKJ+Kg&3rq N)=d{QYYqu/h,s.9.!$@( %RwSᢗo/K\ez}MmfALΝ:yف]Lh 7ND #(઺/Ó4+;tY Hߣ s7͈Ll_E58YWX˧/55)\kxJdfMnӮKg (z0`6hkˋߊ?Ч(1!8BB ;\X T-g 5/d*iTVX:PK޸Zk hc{|~#UˠO2(=1ldW!ǫ' oU --`hSSs2ɒg99Bl҈FY_'Є$ T(mD@h֌vޡߩ/-O_.g>P\d_˔f>?ʜCFȓq̓m/]}P+'K7#WxSm)Kƿ9([4[Fj*Eg謝>hVU5^̩A$B|nTd͉ oM2eӍݺQ&8P w bF-͙]3{щ^Lݭ[3X[4ƭ?.+?+nd1K{rG-|5xyIΉOKRR/H?6gfjFҭ7Di<9\ WOT30 rljE4F̝ú{ܤ;'h+>A=2os>%(YFjV#d۲eڈW%B{vO_0/-IyZsG^lVR{JnG- %hW,J9 3UqYB&ps=79Vu$_!&-ox6Lc[ isЬɦ*uMӹ[Ik.u0:Ox+٦YApZS>  Bގj?ފe!ފ1.>2鲵+]L5<|/|5e@M&QnΔ ƶ h>йܒ6)QO.!o~:帘'qS٭²%%tjР[ĭ۲)QWf2[yJ b"ˤ7Lo M1Y͍ \ TsE,{I;fuiicN|R4/F Ms  C$vcFLLA":%ϛGVuSy^'K`@2mV`O \$e  gW&dKXs%x}y 6ktMC)N`󁃖eCJiF,VQ OPF7yd㎩oU0kF1 \BH@ 3&tnVwvDU!%Σ H    G@AVx l%@H  M  in}103Z  $$     *tRt{HHX8k6 rY8]     "a iX JZA%N     NG{$?˔IAY8 =c=pCR1${\b=pC0R1${\b=pC0R1${\b=pC0R1${\b=pCݻw jCR_*\RNEHGnSחE:]IcxnԌWEآ2-㭬.WVq 4$b!fkYWZ]xw|OH7`pY\q|/cxN}i?;Ef<% Љ93 o:/٫e}^w}tR%NNJo[f~i A>[ՎSNOyqW˟T&bdڇQuceّCSΣG /<7lw-k%Kc -K-pw^[XFhѢ=}mo0XwcO"幃fڹ'~b>ڧyg3'o>wĞEYŗ?>.Xwz'6pwߋ쩬֗Rstn֛O6 QuO%7(]T\+.y}˕?3yk=V|R1}:<2Ssؽ tQ͠Rߟ_^G͒yr559oO=çv ݵ֔Gf#2QDv_o[R]nD M*-pG󋌧d%n;l`[[l},[ `.Ah/ϫv{G>왥;M³/"9MgK=IP,n| :/9TWߖR_<}<;Me儴|l3{\cFCqjۏV#˝;-.T7^iVU] K ' r ivk` [:b._?oG \b#8GUS_*S/K~QHcyaf{I_)kP]t#kCs5z@kTk& >0TWL#p( xo(}|,|:s2[OQc!d3(%4= T;540"dḦB5Rs%ޖ}sS8˞h/cLa0e?n !=<3<"ID:c̈́gל)<|2B |IDATj6Uub A[lB{L},!XxB@'!sI*Bwʃ,>miaVelPjubtZ-oae.G(]뿠4v`.i;=)+nA^ٗ%n ;nRr{}eN_obIucn6 醒~Sm;.TLA1{Deщ& K7DÕ +u̞b9KDYRC翱ERO8ˎ|yee qTT}A}&*VHti_AܡY5v M]8O ڛWO&i9E%lz5pSώJ,]뿠x< 0O1YLK7,כ&eJ ~aګ*bIgUi;T(Rym8ݧD*1deiț23vcwS5/{u[J6r|[aQFYg@iyx|~#tKƌ!l|z蔽yCP~ֺ T?~o;ugE/^y(Y}ر X ǝsI;jSsLK}F /(ܳ_W: 7 {$7 gcy* s3ޙ"}{^J"PQX/-@{hn&w☐VZ.ȫdȤ3b৘eGĞ `sЌwcD#wtr< :>k;~^i-&@S@~$ȈẼ3C"iWdP_T]k  O~)O̜HBޯ+*>2 hc|2@V.-@X!k_;vY ]?VV~@#sX%um${(m@8@I,$ybj!TWȓrSfJ㧓S'ma7Q7V4&k4ϯg,yp?3(f 2wW˕V)ICCi:$ ~8tѼ Sث4h?)+5d!nbA$zEO.zU=xb\ve ! 00t|uHF* +=\X0%"Z4WѴ-ϳ%N^M&7rFfM|.=i.H<8z]T0ŷfN3Z\/W` |p-L^pA'$sI{Ce4nE󯩺vf14&gҊu cӣ_詓iswZ oEȃ5{ KboGU_]u04A;3 [1߶㓍1F3IVU-ŶM@}xH66Q]{3 [I:!5ucEFʆ7bΘS'[-X]՚gn^,3)q`3 ` `U-@%mw^m]2P\No3<|wO=uP*6Jrioްa5"G,9&=u@[,u,&wKJ[&ӹ$|.Uzg j,Sե-a` B7CqNLX"evoX4Ťc1WDeўdޞxEϝ9#O@ 0Ydc;O_X oVO~V"|Y+ͥng`"-@gW ,AX̀JJvsZ{˵wン;6}y=kJgv'66͜+~1mS3^AúU ݙb]O{ WuW~1kNN{Թ3O:yb̼)l4Н~>uHYsi$UcIl K:-[*{=TWaczt}>kѿ?f\]Cϛ4Venۼk93TNT/o~#(6?5KF1=0kxV$@tgp=k/^^.@s/]tj/u&tA୓QٗQXxy UkG+-<-K;ϰ8"{dԤC¬M# hn+qLWʝ:š egFaeE6J.^¥N>g9_dt\M:!Όug^ۯv'^QmiI؛èɬHucetr> F` v T҅`B3 \A]ut/F5eRӋbЯd/ʆ${:>tʇ+IO_6iiͯUlf~U~yr5dA^CK}H8-}0Ց=Fᄽ$GU] k-~\WsX/9HHmXffdeH(E~7 zxeW=W^tܾ߫239ʄ+>s;Cz l#5a  ]@A>E~Sj@6LA1[aøI 蚃?f5IbIn3S80} }3%w ?3xSJZ*$aّi HBRX>dbFj1iH, J.޾]O)+nA^M","hœmlZD gvɦ &mZy}W%4y"]__[X-3m vzX!ιA*`.%nZMφ4F,  G2tog6$ rt/EˊYLȻMF ;=? be-SCoJ7Nm*癉 6eo105WjM ~{[U,UT4Tb 6hp/>Y,9t1+Me'ho>瀚ذf-]jL'`zemgy%QזWtע^=dOREܿ [!J ~/(4N`.icOޓh]_0j/Eԗڟbx\ҰAƟpL9%'u feuI޷Ȓs0q)˿N, Iv4*UuA^٢ԝ%o-(-/ږt|#[<8=X6uWdVQXm|h檔?Pˎ4nX;k ;r\ŃbO f Equ}o hv^.!ƓːW^uGٟ̉Wj`6%>Sm9_\L3Icn,ݱMLpvq-[kBcE. ]뿰.$h w޽6N]m-=Q0wB7u厔qokPa9U-ڳన<35c-ЏE3d>`u$D$+'RSHiԸVL%M;'җ'q:?+g` +Rir62l' K-͙]#IG"{A:;:xB݄!NPk'8 * gC-%}:~T3Aw/k.V.ݴV-ְ-V.En<8);?;騜V,uՉ4n(x_k&F l($٭5Bg 5B;Vb]]|SjI%y4h{b?ٌM='k,>OLMbE-L0آ h_CCewd0;&8 acKFat&!Ucdi7áirEU*91Vb T}*a p]0l6@s*Mǃ?+"`ͩuqae2;V]dD_.IgHi;lx#vY?}+^^ nH^#~\#V[="Aרet1&;հ%G1;>7*ԢĢǦ*³㳧??*ĉ<lWհ @lbg% =eiir|zLLQ(gO2,_b#WkhsŽ -u||#. Y_]q\]@@@@@@@p@5A@@@@@@LEs       NW p0K ́:K\]@@@@@@@`.q0P4       `.qu B~@5A@@@@@@LEs       NW p0K ́:K\]@@@@@@@`.q0P4       `.qu B~@5A@@@@@@LEs       NW p0K ́:K\]@@@@@@@`.q0P4       ?QOIENDB`$$If!vh#v!:V l  t 065!` p ytc$ $$If!vh#v!:V l  t 065!` p ytc$ /$$If!vh#vt#vd#ve#v#v:V l4  t20655. 5/ 53 5 gp2ytXIJkd$$Ifl4rl }l!. / 3   t20644 lagp2ytXIJ$$If!vh#vt#vd#ve#v#v:V l  t 0655. 5/ 53 5 gp ytXIJ$$If!vh#vt#vd#ve#v#v:V l  t 0655. 5/ 53 5 gp ytXIJ$$If!vh#vt#vd#ve#v#v:V l  t 0655. 5/ 53 5 gp ytXIJ$$If!vh#vt#vd#ve#v#v:V l  t 0655. 5/ 53 5 gp ytXIJ$$If!vh#vt#vd#ve#v#v:V l  t 0655. 5/ 53 5 gp ytXIJ^K pp00H02 0@P`p2( 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p8XV~_HmH nH sH tH D`D :Normal xxCJ_HmH sH tH l`"l Ap ; Heading 1$ & F 7(@&5CJ$OJQJ_HmH sH tH X@X ?tf Heading 2$ & F 7@&5CJOJQJL@L GNv Heading 3$ & F@& 5OJQJL@L -Nv Heading 4$ & F@&5CJD@D Nv Heading 5 & F@&6CJ^@^ BS[ Heading 6 & F<@&5CJOJPJQJ\^JaJT@T CS[ Heading 7 & F<@&OJPJQJ^JaJZ@Z DS[ Heading 8 & F<@&6OJPJQJ]^JaJX @X ES[ Heading 9 & F<@&CJOJPJQJ^JaJDA`D Default Paragraph FontRi@R 0 Table Normal4 l4a (k ( 0No List 4 @4 c@0Footer  B#d"`d Caption#$$:h^`:a$5CJ_HmH sH tH 8o8 uKCopyright textCJH@"H . Balloon TextCJOJQJ^JaJ4O24 / Quotes ]^7CJB@B J(pTOC 1! ^e`5>@> mpTOC 2 S! ^>@> J(TOC 3! ^7CJaJ66 c`)TOC 4! ^.)@. @ Page Number4@4 c`)Header 9r ^0`^ List Bullet & F5x<]5CJ_HmH sH tH jj  Table Grid7:V02o2 c@0 Footer CharCJR6R tf List Bullet 2 & F n5<]5R7R tf List Bullet 3 & F 5<]5BU`B/ 0 Hyperlink>*B*mHnHphuFV`F uFollowedHyperlink >*B* phdOd `Source and notes%! & F (^`CJho"h m Footnotes+"$ :xx^`:a$m$_HmH sH tH :2: "Zu Table heading#a$:@!B: %/  Footnote Text$a$<Q< $/ Footnote Text CharP&`aP { Footnote ReferenceCJH*KHOJQJfOrf { Table text indent roman'((^`CJPOP uKTable text heading (<<5CJpop uKTable text left column):<<^`:_HmH sH tH D.@D J( TOA Heading*5OJQJ\aJvov / Contents heading+:x^`:"5CJ$OJQJ\_HmH sH tH tot "Zu ACCC title#,$:h^`:a$m$"5CJ$OJQJ\_HmH sH tH <o< tfHeading 4 Char5CJZOZ 3sContacts heading 1 .$5CJOJQJNoN XL& Contacts text /<<_HmH sH tH HOH [Contacts heading 20$6POP [Contacts heading AER 1]\o" [Contacts addresses+2:xx&dP^`:5CJOJQJ_HmH sH tH PA2P { Table text indent alph 3 LoBL 6K/ table text 4<<CJ_HmH sH tH \AR\ { table text indent bullet5 & F HbH 6K/ list alpha67<<^7`horh ^6cMulti-level list 17 & FxCJ_HaJmH sH tH HqH ^6cMulti-level list 2 8 & FFF ^6cMultilevel list 3 9 & F<m<661 / 1.1 / 1.1.1: FVY@V 66 Document Map;-D M CJOJQJ^JB'`B PComment ReferenceCJaJ8@8 HP Comment Text=CJ@j@@ PComment Subject>5\DoD :Heading 2 Char5CJOJQJVoV iDefault @7$8$H$!B*CJ_HaJmH phsH tH DoD  Heading 1 Char5CJ$OJQJJo!J S[Heading 6 Char5CJOJQJ\aJDo1D S[Heading 7 CharCJOJQJaJJoAJ S[Heading 8 Char6CJOJQJ]aJDoQD  S[Heading 9 CharCJOJQJaJ.X`a. A@Emphasis6]DoqD Heading 3 Char5CJOJQJ:/: =Comment Text CharF^@F 7n0 Normal (Web)Idd[$\$aJ@`@k0RevisionJCJ_HmH sH tH PK![Content_Types].xmlN0EH-J@%ǎǢ|ș$زULTB l,3;rØJB+$G]7O٭V$ !)O^rC$y@/yH*񄴽)޵߻UDb`}"qۋJחX^)I`nEp)liV[]1M<OP6r=zgbIguSebORD۫qu gZo~ٺlAplxpT0+[}`jzAV2Fi@qv֬5\|ʜ̭NleXdsjcs7f W+Ն7`g ȘJj|h(KD- dXiJ؇(x$( :;˹! I_TS 1?E??ZBΪmU/?~xY'y5g&΋/ɋ>GMGeD3Vq%'#q$8K)fw9:ĵ x}rxwr:\TZaG*y8IjbRc|XŻǿI u3KGnD1NIBs RuK>V.EL+M2#'fi ~V vl{u8zH *:(W☕ ~JTe\O*tHGHY}KNP*ݾ˦TѼ9/#A7qZ$*c?qUnwN%Oi4 =3N)cbJ uV4(Tn 7_?m-ٛ{UBwznʜ"Z xJZp; {/<P;,)''KQk5qpN8KGbe Sd̛\17 pa>SR! 3K4'+rzQ TTIIvt]Kc⫲K#v5+|D~O@%\w_nN[L9KqgVhn R!y+Un;*&/HrT >>\ t=.Tġ S; Z~!P9giCڧ!# B,;X=ۻ,I2UWV9$lk=Aj;{AP79|s*Y;̠[MCۿhf]o{oY=1kyVV5E8Vk+֜\80X4D)!!?*|fv u"xA@T_q64)kڬuV7 t '%;i9s9x,ڎ-45xd8?ǘd/Y|t &LILJ`& -Gt/PK! ѐ'theme/theme/_rels/themeManager.xml.relsM 0wooӺ&݈Э5 6?$Q ,.aic21h:qm@RN;d`o7gK(M&$R(.1r'JЊT8V"AȻHu}|$b{P8g/]QAsم(#L[PK-![Content_Types].xmlPK-!֧6 0_rels/.relsPK-!kytheme/theme/themeManager.xmlPK-!0C)theme/theme/theme1.xmlPK-! ѐ' theme/theme/_rels/themeManager.xml.relsPK] 3346)789;;<I8KKS`YaNdBghjjlm21t  !"#$WVI|}uh A d P dt*Czr 8 vvvvv\\\_S  H  pC(/f59M<AFAKPT[1ag%louiyj~cH:#@S! "9J^aooo8pppr .f",4H^r JT=e]xd{0LNOQq?$&')Ib~?Njmnp '''01?1RRS X%X%X%X%X%X%X%X%X%X%X%X%̕4TXv_!!M[i4GKHzs< < 0{R |Lw?xXXXXXXXXXXXXXXXT # @H 0(  0(  B S  ? _Toc276654296 OLE_LINK3 OLE_LINK4 _Toc148174795 _Toc213047617 OLE_LINK1 OLE_LINK2 _Toc364245657 _Toc364245658 _Toc361664672 _Toc361664702 _Toc361664722 _Toc361665349 _Toc361665723 _Toc364245659 _Toc364245660 _Ref361132010 _Toc364245661 _Ref361663954 _Toc364245662 _Ref361735769 _Toc364245663 _Ref361131994 _Toc364245664 _Toc364245665 _Toc364245666 _Toc364245667 _Toc364245668qqr  != jjjjjj!,11>>DDVVM@  rrz )f jjjjj , ,5,11 ? ?DDVVaL  '''(RSGHJKMNPQST1LAAYYabɼۼr7?HJKMNPQST33333333333333{==  {{wwGGm%m%''|/|/0@1223377P8P8u8u888wFwFJJRSYUYYY__````cc d dff3o3ohqiq,t-tuu||/}0}4}>}}}2~>~[[XXFQ | jlLNHTdg@E77AL::MMYYGHHJKKMNPQST`d/00L{pq>?HI>?==  {{wwGGm%m%''|/|/0@1223377P8P8u8u888wFwFJJRSYUYYY__````cc d dff3o3ohqiq,t-tuu||/}0}4}>}}}2~>~[[XXFQ | jlLNHTdg@E77AL::MMYY:;GHHJKKMNPQST`d/00L {|~ 6c|%6nm܃_~c3x+ؕQ+~I1%  L1~/53.z^S@5:d4AHHOH75_WI~"I.yRa0Z~]~9O^~X] _*=3789e~?gpj+Mp Q|~h 8^8`o(hH)h^`OJQJ^Jo(hHoh ^ `OJQJo(hHh ^ `OJQJo(hHhx^x`OJQJ^Jo(hHohH^H`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh ^`hH.h ^`hH.h pL^p`LhH.h @ ^@ `hH.h ^`hH.h L^`LhH.h ^`hH.h ^`hH.h PL^P`LhH.h^`OJQJo(hHh^`OJQJ^Jo(hHohp^p`OJQJo(hHh@ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohP^P`OJQJo(hH8^8`CJOJQJo(pp^p`OJQJ^Jo(o @ @ ^@ `OJQJo( ^`OJQJo(^`OJQJ^Jo(o ^`OJQJo( ^`OJQJo(PP^P`OJQJ^Jo(o   ^ `OJQJo( 8^8`o(hH) ^`hH. pL^p`LhH. @ ^@ `hH. ^`hH. L^`LhH. ^`hH. ^`hH. PL^P`LhH.h^`OJQJo(hHh^`OJQJ^Jo(hHohp^p`OJQJo(hHh@ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohP^P`OJQJo(hHh 8^8`o(hH)h^`OJQJ^Jo(hHoh ^ `OJQJo(hHh ^ `OJQJo(hHhx^x`OJQJ^Jo(hHohH^H`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh ^`hH.h ^`hH.h pL^p`LhH.h @ ^@ `hH.h ^`hH.h L^`LhH.h ^`hH.h ^`hH.h PL^P`LhH.h^`OJQJo(hHh^`OJQJ^Jo(hHohp^p`OJQJo(hHh@ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohP^P`OJQJo(hHh 8^8`o(hH)h^`OJQJ^Jo(hHoh ^ `OJQJo(hHh ^ `OJQJo(hHhx^x`OJQJ^Jo(hHohH^H`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hH P^`PhH @^@`hH. 0^`0hH.. `^``hH... ^`hH .... ^`hH ..... ^`hH ......  `^``hH.......  0^0`hH........h 8^8`o(hH)h^`OJQJ^Jo(hHoh ^ `OJQJo(hHh ^ `OJQJo(hHhx^x`OJQJ^Jo(hHohH^H`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hH8^8`B*CJOJQJo(ph^`OJQJ^Jo(o ss^s`OJQJo( C C ^C `OJQJo(^`OJQJ^Jo(o ^`OJQJo( ^`OJQJo(^`OJQJ^Jo(o SS^S`OJQJo(h^`OJQJo(hHh^`OJQJ^Jo(hHohp^p`OJQJo(hHh@ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohP^P`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohp^p`OJQJo(hHh@ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohP^P`OJQJo(hHh&^`B*CJOJQJ^Jo(phhH" h^`OJQJ^Jo(hHohp^p`OJQJo(hHh@ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohP^P`OJQJo(hHh 8^8`o(hH)h^`OJQJ^Jo(hHoh ^ `OJQJo(hHh ^ `OJQJo(hHhx^x`OJQJ^Jo(hHohH^H`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hH^`CJOJQJo(^`OJQJ^Jo(o   ^ `OJQJo(   ^ `OJQJo(xx^x`OJQJ^Jo(o HH^H`OJQJo( ^`OJQJo(^`OJQJ^Jo(o ^`OJQJo(h ^`hH.h p^p`hH.h @ L^@ `LhH.h  ^ `hH.h ^`hH.h L^`LhH.h ^`hH.h P^P`hH.h  L^ `LhH.h 8^8`o(hH)h^`OJQJ^Jo(hHoh ^ `OJQJo(hHh ^ `OJQJo(hHhx^x`OJQJ^Jo(hHohH^H`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh 8^8`o(hH)h^`OJQJ^Jo(hHoh ^ `OJQJo(hHh ^ `OJQJo(hHhx^x`OJQJ^Jo(hHohH^H`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohp^p`OJQJo(hHh@ ^@ `OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHohP^P`OJQJo(hH 77^7`o(hH.'77^7`o(GCJsH tH aJ_Ho(hH.. 77^7`o(hH... 77^7`o(hH.... 77^7`o(hH ..... 77^7`o(hH ...... 77^7`o(hH.......  77^7`o(hH........  77^7`o(hH.........h 8^8`o(hH)h^`OJQJ^Jo(hHoh ^ `OJQJo(hHh ^ `OJQJo(hHhx^x`OJQJ^Jo(hHohH^H`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hH h^h`OJQJo(8^8`OJQJ^Jo(o ^`OJQJo(  ^ `OJQJo( ^ `OJQJ^Jo(o x^x`OJQJo( H^H`OJQJo(^`OJQJ^Jo(o ^`OJQJo( ^`hH. P8^`PhH.. ^`hH... x ^`xhH....  ^`hH .....  Xx^ `XhH ......  ^ `hH.......  8^`8hH........  `^``hH.........h 8^8`o(hH)h^`OJQJ^Jo(hHoh ^ `OJQJo(hHh ^ `OJQJo(hHhx^x`OJQJ^Jo(hHohH^H`OJQJo(hHh^`OJQJo(hHh^`OJQJ^Jo(hHoh^`OJQJo(hH"/53"I%H] _] _ MpI1+4AcQ+9O^S@5I1I1I1I1 .yRQ|]_eI1?gI1cF_WIF {|F0ZFL1FmPtQfRoSn^XH                            x*        r                                                              ?v2                                           ,?                                                                       GuT|g10]I# p C}"L vM+y~ T&FofyUg"(H%;*,DwhC]:(Cwd:Jh~=}{x)++RH%Dwx("L {x)x("*H0GvM+?oRNS+;*,{.{.:~/-6j37Jim3p [68C:zWxXp<zkD>"@>⸿!3r°&sܰ3'8Cܰ"@9$3'PR`F0GJHݱ<$ ,L~=}Z*LJRBwOCwY+RWCPRhXr%S+ {TSJHy~ TuhXI# CwYC]:10],r`,r`"*ZZcy`!A;fQ~tyUg[6:Jh-6j37JiRNS+QiFof#n?oD1rr;f&sQ~tdCw(Cw"(zWxT|r%SC}giVh]aoU{%PAo47o9|312{_Pp/ Wbj{ %:MRUm ,-:Q`.oP E _ Ig  0 A K yk t {  c$ I0 0 ]  0 8 oS  s$ "K N U :Y -)m2BXNWZfn vf&9:FxYw9L 3Wl+~ Clms},/58?u1ro4--~. <JpL7 > jYb.W6FKm(N)0Ou$b(b<FFMIV~[Ln(ZJ#2B\FzU`y ) cJ eZ r r u } !:4!Q!j!n!t!A"$/"7i"i"5##=5#7$l$G$wL$)%x_%&B&E&()<4)G)YK)c`);})**(*0*V:*D*-V*1+i:+|+*,y+,$-,xf,-d*-*-7-~^-b-=m-n-2...b.h..// /g$/6K/V^/0 00%0(0)0W20t3070D0G04O0_0q0211G1o|1O2f2G{2} 3343X3G^3.4>4(l4q45&5eI5R5L^5E`5o5Hs50{5 67"6 L6[6767tg7|7,868rL8 9W9+9p9Uz9 :::%-:Q:_b:|k:n:s:|}::;;A ;p ;2;=;?;aD;P;~;<P!<3< C<"G<u= ==M=(=+=$,=I="}=z~= >|>{9>C>\Q>T]>o>v>Z*?/?GE?[@]@u@.AA/ABDB=fB~BC},C>C%@C*@CkHCO DqD8DaDuD,0EGEPEREjEFLL-3L5L>L0ZLaL6MAM"N:NVN]N\eNhiN[{NRO O3OZxOPDPO4PMPsUP`PaPuPPQQQWQcQ~RRURLS@SzNSmS/ySLzS) TT1T3T44TQTsTtT!{T?U^*U!MUiU@{UV$VCVJVNV#W W6WBW*EWZW|W~WXX"%XcaFcZdRdlSdWdYdjdpwd"e+ezez^f2cftf gFglg!hhh h,h%iIi iiQi%ikHidivi*}i j&j_jmbjkk%k*k5kP>kSkmkHlFl^ldl mq mmm)m5m?9m*Gmsm{|m}m7 nnn?)n,n0n7n o oooEoXo^o,fopp)p+pVpq#q=Nqkqrqr3Cr$ararns3sa:sITsfsissK6t;t0%u8uFu"ZuduZpupu9ruBv"vHvNvkvqvLwMwYw_wtwuwx%-xHxbxtfxyxyBAyQyfyE zsz{,{B{Z{|"|9|^|{_|w|,}}:}z}n~~~A"~Kn~ox~,o,z@^_kYKB}P e2 34,Pz#Lv C6<JOR_Z 0w.?Zg@N!iu jI#:H==eAx"":<k|u}",grm}V@,ce{OMlrssuYwn1<?@mEoby'BtgB9QSZ . g5ACN^jylQ27RMW+gBKPcf? @Rb hq-^Q#f'&UX&^%+b}y!MA!G,M{W]{$!8]euinTr~jR`4b& 2KSiWdn-;lx{0HxFzx-8]k&xy&:lMN'N=SRegsh5 C@ |<5I`I]S[W`yI^$/34Li Y_ah;u "3Tj{Y/?;E]n. BUde:s6ETE],n1%\t{wm!%Ki~~=+36dcnn *0hO{Y?r % &cLp )5HYV{}'O[Wo&/5FK:On) >)HWi3kg?g KQz8{{i}6Y}m{|P >#7j=H>IRStvOM/A!i)]01r.3aorWDc0tBVfi>t -n1Mk$pkP|Tolm{d"*38}9DLkFqYyG@I^VU6C<=:FYg=uw~ |3hBP>^3K6bo|:ZlS[F\gpqP+WT,_ll8=K[ 6?DouwS)555BI`qqy!!v**.D3.:9g]zA"2d"#)*bhn/-1v*^0667>bBDIM"W)Qf7c@Giij5s/tmE OX6p 0S>cISZ&u>Q|L `'s(*q7{p{7=R6 bg= .~RU *bb}N2AEJQSevzK (/J #h`$$|2W`m-E@WVd'rm8,;=S2WSw +0.H\M:+M?Gw* )< !k``opo &:['7tLOPko%'.NdZ %_1'I5|Q37_`2a67C+{7NUAz|t 6>SzX C9[,(36C%E*.:Hoex E ?Xs!1_Toc36424566758_Toc36424566652_Toc3642456655,_Toc3642456645&_Toc3642456635 _Toc3642456625_Toc3642456615_Toc3642456605_Toc3642456595_Toc3642456585_Toc364245657)'*+https://www.truenet.co.nz/news/Performance)''+https://www.truenet.co.nz/news/Performance $2http://www.ida.gov.sg/applications/rbs/chart.htmlw>!Thttp://telstrawholesale.com.au/products/broadband/adsl/adsl-reports-plans/index.htmNUhttp://media.ofcom.org.uk/2013/03/14/average-uk-broadband-speeds-hit-double-figures/ =http://www.fcc.gov/measuring-broadband-america/2013/FebruaryEJyhttp://stakeholders.ofcom.org.uk/market-data-research/other/telecoms-research/broadband-speeds/broadband-speeds-nov2012/2(shttp://www.comcom.govt.nz/media-releases/detail/2012/commerce-commission-selects-truenet-to-test-broadband-quality 2http://www.ida.gov.sg/applications/rbs/chart.html4(/http://www.fcc.gov/measuring-broadband-americaN Uhttp://media.ofcom.org.uk/2013/03/14/average-uk-broadband-speeds-hit-double-figures/jr Whttp://consumers.ofcom.org.uk/2011/07/consumers-benefit-from-uk-broadband-speed-surge/$:`http://stakeholders.ofcom.org.uk/market-data-research/other/telecoms-research/broadband-speeds/]2http://www.nbnco.com.au/about-us/our-purpose.htmlcYhttp://www.dbcde.gov.au/__data/assets/pdf_file/0003/132069/Statement_of_Expectations.pdfpH:\TRIMDATA\TRIM\TEMP\HPTRIM.5460\D13 82182 BPMR Program - Finalised Consultation Paper - August 2013(3).DOC  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcefghijklmnopqrstuvwxyz{|}~Root Entry FP@Data 1TabledWordDocument SummaryInformation(DocumentSummaryInformation8$MsoDataStore9@p=Y5SKJUKIV5XA==29@p=Item  PropertiesUCompObj r   F Microsoft Word 97-2003 Document MSWordDocWord.Document.89q