The Scottish Government

« Previous | Contents | Next »

Listen

5. Technical Options

This section describes the possible technical solutions available to extend the reach of broadband in the out of reach areas.

A range of options that could be used to address the issue have been examined, including developments within BT's DSL platform and local loop cabling plant, along with alternative options, such as WiFi, WiMAX and mesh radio, other wireless spectrum developments and satellite.

The options are based around the following main headings:

• BT developments in the access network to improve reach
• Other DSL Developments
• Wireless Options
• Satellite Options.

5.1 BT Developments

5.1.1 Overview

BT has made no major investment decisions on infrastructure outside the substantial 21 CN programme. This programme will see development down to the exchange level with the introduction of the Multi-Service Access Network module. This will provide ADSL2+ to replace the current ADSL service. VDSL is the other technology that BT is investigating as the next step in the evolution of services. In the meantime, the current ADSL service is being repackaged as ADSL Max to offer improved performance.

5.1.2 ADSL

The standard ADSL product at 512kbps had its loss limits removed in 2004, so that all lines have been accepted for order to at least attempt to obtain a service. This sometimes includes the installation of a filtered faceplate on the customer's master socket, so as to eliminate poor quality telephone extension cables inside the customer's premises, which can be a source of high frequency noise. ⁵

A lower rate service, such as 256kbps, can also be provided to some customers whose line won't support 512kbps. As explained in Section 4.2, this lower rate would use fewer frequency elements within the ADSL frequency plan. However, as a result of operating at the limits of ADSL, the resulting service could very well prove to be unstable, as the ADSL modem can fail to synchronise between customer and exchange. At these limits there is also less margin available for handling noise on the line, with most DSL modem equipment failing to work at signal-noise-ratios of less than 6dB.

5.1.3 DSL Max

ADSL Max is the name for a new range of products from BT Wholesale, which will offer downstream line speeds from 160kbps up to 8192kbps, and upstream speeds between 64kbps and 832kbps.

The service has been launched in response to the research undertaken by BT on the performance of ADSL with their existing fixed rate products and a desire to move away from the caution required when delivering against fixed rates.

This service uses the existing DSLAMs already in place, but takes advantage of an increased understanding of the performance of DSL through rate adaption and enhanced management functions - known as Dynamic Line Management. Both the downstream and upstream is rate adaptive which means the ADSL modem will communicate with the exchange hardware and negotiate the best possible speed for each individual line. (Note: Rate adaption has been used previously in the UK, but only on the upstream side of the Home/Office 500 products).

Once the line has been provisioned with MAX, the very first time that the ADSL modem syncs up, the Line Speed is used to set a temporary profile for the first ten days of operation. BT has added extra monitoring systems, which run for this period to determine the Maximum Sustainable Rate ( MSR) for the line. Once the ten days are over, the data speed will be adjusted to be closer to the stable line speed. The profile for the line only changes when a stable connection has been maintained for a period of time.

If the line will not sync at the chosen speed at any point, it will rate adapt down to the next best speed it can manage. As DSL Max is defined over a wide range of speeds (288-8128kbps), it ensures that even for many longer lines, lines will be able to achieve synchronisation at some level. The very low potential sync speed of the Max services may allow people who currently have Home 250 to get speeds like 384kbps, which will be a welcome boost. In addition, there is a chance that if the line did not support a Home 500 or Home 250 service, it will now sync under the Max products. Although this would provide some benefits to users over dial-up, it would be provided at the very extremes of service and consequently may be affected by reliability issues, and due to its reduced speed, could not be considered as a broadband service.

It is worth noting, though, that there are still 170 ADSL enabled exchanges in Scotland that will not yet be able to receive ADSL Max. These are largely those exchanges using the Exchange Activate community broadband solution, which don't have sufficient backhaul capacity to handle the Max services.

The following guide from BT shows approximately what proportion of BT Broadband customers can expect which maximum line speeds:

Table 5.1: Broadband Max Line Speeds (Source: BT Wholesale)

Table 5.2 shows the relationship between the sync speed monitored on the line and the resultant data rate that will be supported. The table shows there are very few steps at the lower range below a 2Mbps line sync speed.

Table 5.2: Broadband Max Line Sync Speeds (Source: adslguide)

As an example, for a downstream line sync speed of 768Kbps, (which equates to the second table entry 'from 576kbps'), the maximum throughput of data provided to the customer will be 0.5Mbps. The only benefit for that user would be the increased upstream offered by the Max products. The large step sizes are most likely behind the decision by BT Retail to not offer upgrades to its customers on very long lines, since they felt some might see little benefit. Additional steps in the date rate at lower speeds would be more beneficial for longer lines. As it stands, a line can rate adapt down to 160kbps, but BT will only support lines able to work stably at 288kbps and above.

5.1.4 ADSL2+

ADSL2+ is the next technology to supersede ADSL and will be introduced by BT as part of the 21 CN network upgrade. This upgrade will provide the service to all exchanges in the UK. At the time of writing, the BT productisation process is underway for these services. Whilst other providers are already offering ADSL2+ products through local loop unbundling, BT have the added pressure of providing the service to as wide coverage as possible.

Based on network noise assumptions, BT estimates that ADSL2+ will give many customers 8Mbps, and give a few customers speeds of more than 18Mbps. These speeds will give good sustainable throughput suitable for video, for example.

BT pushed for the recent changes which were made to the ANFP, (the specification of the way in which DSL technologies may use the copper loop) which now allows ADSL2+ to be provided from both the exchange and the cabinet.

Commercial trials are underway of Fibre-To-The-Premises ( FTTP) and Fibre-To-The-Cabinet ( FTTC) architectures. The FTTC trial involves cabinet-sourced ADSL2+ and is aimed at extending the reach of the current broadband product set. This includes serving customers on TPON cabinets that are beyond the reach of broadband on copper.

The target market for this type of hybrid delivery will, however, remain the more urban areas, in order to increase the coverage at these higher speeds.

5.1.5 VDSL2

Within the wider market, broadband developments have made 2Mbps access pretty much the entry-level option. Developments continue, however, and telcos are beginning to look at much higher bandwidths as they work towards offering real-time video services over the network. Convergence is the aim here.

A big issue is what role copper will play in this new scheme of things, as the bandwidth demands get higher. Fibre has a big role, but economics of deployment limits its widespread delivery to the household. An alternative copper based solution, VDSL2, is a new DSL standard recently published which is gathering wide support within the industry as an important technology for supporting mass-market broadband access at many tens of megabits per second. There is the usual trade-off between speed and distance; typical figures for the UK to comply with regulation, would be speeds of 12Mbps/6Mbps downstream/upstream at around 1.8km. A key attraction is that VDSL2 can be deployed in mixed exchange/cabinet modes in parallel with existing ADSL/ ADSL2+ systems, thereby providing a migration path.

VDSL has featured in BT's thinking for some years, and they are considering a possible future trial of cabinet-sourced VDSL2, as and when suitable line-cards are available from manufacturers. This is the main driver for the remote cabinet based solution and would be based on fibre to the cabinet with VDSL operating over the existing copper from the cabinet to the premises.

VDSL is likely to be best suited to provision of bundled multimedia services to mass market.

5.1.6 Local Loop Upgrades

In a number of cases, the performance of BT's local loop copper cable has degraded to unacceptable levels due to the poor condition of the cable. This may be due to deterioration with age, or excessive breaks and joints, for example. In situations where the cable is so poor that normal telephony can't be provided, or if there is severe shortage of copper pairs, BT will replace lengths of local loop as part of their continuing investment program. An example is the recent cable network upgrade in the Poolewe area of Wester Ross which included 4,375 metres of cable renewed, 26 joints remade or replaced, 384 metres of duct laid and seven DACSs removed to a total cost of £106,256 ⁶.

5.2 Other DSL Developments

5.2.1 Remote DSLAMs

One way to increase reach is by moving the DSL unit, the DSLAM, closer to the subscriber. Pushing the access box deeper into the network enables service to previously unreachable subscribers, and by shortening the ADSL loop also enables higher video-capable bandwidth.

Remote DSLAMs are environmentally hardened versions of the exchange based DSLAM which are placed inside a cabinet and then installed in the field. The remote units can be connected back to the exchange using copper pairs or fibre. As there is no power source at the remote location, the units are powered over copper from the exchange.

Remote DSLAMs can be useful for serving large numbers of ADSL subscribers and scale easily, usually by adding line cards into a chassis. However, there can be significant problems in connecting up these units particularly in more remote areas where there is limited number of spare copper pairs. A customer requiring ADSL would have their copper pair fed into the remote DSLAM where voice and data would be split with the voice portion having to be rerouted back into the Cross Connect to the exchange. The DSLAM would also require typically a 2Mbps backhaul link to be delivered, likely over the available copper in the area, which may prove a problem.

Figure 5.1: Local Access Network (source: BT,Mason)

Remote DSLAMs can come in many variants with sizes to address the specific requirements from small limited port modules up to standard DSLAM units. The cost of the equipment is not usually the main consideration as it is usually the installation costs of obtaining suitable sites, which will be the main factor.

5.2.2 Loop Extenders

Several companies are promoting various forms of loop extension technologies as a lower cost alternative to the remote DSLAM. Their target is to address small clusters of subscribers beyond the current distance limitations, rather than large groups of subscribers, which may be better suited for remote DSLAM deployments.

These technologies typically incorporate advanced algorithms that allow increased performance and overcome different types of impairments on the local loop. Unfortunately, many loop extension technologies are proprietary to a specific vendor and it will only be through standardised versions of loop extension, such as provided for the DSL alternatives of SHDSL and HDSL2, will it become more wide spread.

5.2.3 Repeaters

Repeaters amplify the signal to increase range. Using SHDSL to feed line powered repeaters; some operators have been able to achieve DSL service up to 16km from the exchange. ⁷

However, repeaters need careful engineering using good loop records and ensuring compliance with frequency plans. In many cases, installing a remote DSLAM would be cost-justified merely because it simplifies plant engineering.

5.2.4 ADSL2+ Annex L

ADSL2+ Annex L, also known as RE- ADSL2, where RE stands for Reach Extended, is a variation on the standard ADSL2+ service, to provide higher speeds at the limits of range.

With this ADSL standard, the power of the lower frequencies used for transmitting data is boosted to increase the reach of the signal. The upper frequency limit for RE- ADSL2 is reduced to keep the total power roughly the same as standard. Although this standard has been ratified by the ITU, not all operators allow this protocol to be used on their network, simply because the extra power on the lower frequencies might cause problems for existing services due to crosstalk. It is understood that this is the position with BT.

5.2.5 Conclusions

Whilst there are a number of developments within DSL which involve bringing the service closer to the end customer, their relevance to this study depends on the development plans of BT. From our discussions with BT and our wider research, it is clear that BT's focus is to move to ADSL2+ through the 21 CN upgrade programme, and then to VDSL2 through a FTTC implementation in the future. These developments will be focused on bringing video capable bandwidths to the widest household coverage as is economically feasible. This matches the plans seen in other countries where DSL competes with cable to offer 'triple play' services.

However, we believe that pressure should remain on BT to develop appropriate and feasible solutions for the out of reach areas, but with a consideration of the fit into their overall investment plans. The upgrade program of Openreach in addressing issues with existing cabling is likely the best opportunity at present.

5.3 Alternative Options

5.3.1 802.11 WiFi

WiFi networks are often used as the basis for community networks where a low cost backhaul, such as satellite, is used to feed out to a cluster of properties using WiFi unlicensed access points. There are a number of these networks in place across the country, mostly funded by the public sector to some degree, and operating on a small scale. The attraction of this solution is that it is relatively cheap to install due to the abundance of equipment available designed for this unlicensed spectrum.

The wireless spectrum used is at 2.4 GHz and 5.8 GHz. The lower frequency has been a more common implementation due to cheaper equipment (although the price differential is reducing rapidly), and a longer range than the higher frequency. The 5.8 GHz equipment has the benefit of operating in a less crowded spectrum and uses a different transmission protocol that allows improved handling of reflected signals due to multi-path interference. This means the technology is sometime referred to as operating on non-line-of-sight. This effect is better noticed in more urban areas that have more reflected paths.

Operation at 2.4 GHz involves equipment conforming to one of two standards - 802.11g or more commonly 802.11b. This gives a throughput on the wireless radio of approximately 5.5Mbps with the 5.8 GHz variant (which operates to standard 802.11a) having a throughput of around 25Mbps. Range for each, on a perfect link, can extend to several km with 2.4 GHz having better range than 5.8 GHz.

WiFi can be used in point-point or point-multipoint topologies with a typical scenario involving the local distribution of a backhaul link through a combination of multipoint hub covering several subscribers and onward over point-point hops to serve further customers.

There are limitations in the number of hops which can be deployed as the performance degrades the further from the backhaul point and the normal radio planning engineering requirements must be considered for the design of the network. For example, trees will attenuate the link; links over water can be affected by reflection; and buildings can cause reflection and diffraction, which can also degrade the link.

From a technical perspective, WiFi can offer a good bandwidth experience for customers with symmetrical bandwidth and at levels equivalent to ADSL Max products. The level of service provided will be down to the design of the network in terms of capacity, contention and backhaul. Contention, which is the sharing of available bandwidth between the customers served, will be introduced at the backhaul interconnect and also at the radio itself.

As an example, consider a simple case where a DSL 2Mbps service is fed to an 802.11a access point to feed five users, Cluster A. This is shown below:

Figure 5.2: Example WiFi network

In this example, the 2Mbps ADSL connection is terminated at a location as close to the exchange as possible in order to get maximum performance, but within radio line of sight to the subscribers. The ADSL service provides 2Mbps download and 256kbps upload split between the five users. The radio access point provides a throughput of approximately 25Mbps. The radio works as a transmitter and receiver and can't 'talk' and 'listen' at the same time. The customer will receive a throughput at worse, assuming all five customers are active, of 5Mbps to the radio and 400kbps to the Internet.

The second example is slightly expanded on the above and adds in an extra point-point hop to provide coverage to another five customers, Cluster B, as shown below:

Figure 5.3: Example WiFi network

In this example, the radio units should be dual radios, one providing the point-point link and the other the coverage. A single radio would offer much lower throughput. The point-point link will have a throughput of approximately 25Mbps. Cluster B users will share the 25Mbps coverage throughput of radio B and then the point-point link. All ten customers will share the ADSL 2Mbps backhaul link and so will receive a throughput at worse, assuming all ten customers are active, of 5Mbps to the radio and 200kbps to the Internet. The availability of bandwidth in the backhaul connection is an important factor to the overall network performance.

WiFi networks are increasingly popular in the US where 'mesh' technology is emerging as a preferred option.

5.3.2 Mesh

Mesh networking offers an alternative wireless topology, which consists of an array of wireless access points, connected together in a mesh.

Instead of a series of base stations, each serving a number of remote customers, the customer stations become 'nodes' that are interconnected in a mesh arrangement. Each node can communicate with several other nodes as well as being able to terminate data for the customer. Data traffic flows over one or more paths between succeeding nodes to reach its destination (so that the radio mesh has some similarities to the structure of the Internet). Traffic flows in and out of the mesh by connecting some of these same nodes to the core network, e.g. through DSL, satellite or leased line.

If one route becomes un-workable then the node will automatically seek out another, providing self-healing structure and removing the reliance on single points of potential failure.

An example is shown in the figure below:

Figure 5.4: Example WiFi mesh network

The company LocustWorld produces a popular mesh solution for community networks in particular due to its relatively low capital cost and 'open-source' nature. Their meshbox is a single unit combining a router and an access point with a typical installation involving a number of meshboxes located on suitable property which route the traffic between each other. These are low cost units (of the order of £250) and are in operation worldwide. An additional attraction is that the software code, which makes up the basic meshbox can be freely downloaded and used on an existing PC with wireless card to allow a small operator to make up their own meshbox.

According to LocustWorld ⁸, typically a commercial mesh becomes viable with around 50 domestic users, or 20 business users. Small-scale community meshes that are run by volunteers have lower operating overheads, and they can become economical with around 20 domestic users. They also suggest that a typical 50-user mesh can be set up for an all in cost of between £5k and £10k. Operators with very tight budgets can save money by doing their own installation work and using recycled PCs to build the mesh nodes.

There are a number of small-scale operators in Scotland using 802.11 wireless to provide services to local communities. Examples include Speednet operating in Ayrshire, cmsbroadband in Dumfries and Galloway, and Dick Fleming Communications operating around Aberdeen. These providers use different technologies, but are all based around 802.11 low cost wireless with satellite or DSL backhaul.

A recent report by the Community Broadband Network ⁹ found community initiatives in 550 towns and villages around the UK operated by 260 organisations. Of these, 40% were social enterprises, 30% SMEs, 10% public sector and 20% were not yet constituted. The vast majority of these were less than two years old and almost all are deploying low cost wireless broadband technology.

5.3.3 Practicalities of Wireless Deployments

To provide a wireless solution on a small scale to serve clusters of users, maintaining low ongoing costs are fundamental to the long-term viability of the service. Customers will not accept a substantial premium on the monthly subscription rates and with low numbers being served, careful consideration is needed on how these networks are designed, implemented and supported. Low cost often means using low cost radio and routing equipment and relying on the goodwill within a community for location of equipment on rooftops, the use of household power supplies and an often 'best effort' support service.

As an example, consider a typical small network of 30 users provided with a service through 2.4 GHz unlicensed wireless from a DSL backhaul. Customers are served from two hub points, each with DSL backhaul, and which are interconnected via wireless - similar to Figure 5.4. Typical figures for a low cost network ¹⁰ would be as follows:

This simple example shows that after 8 months the capital costs are repaid, but takes no account of maintenance, support or wages.

This example also assumes that suitable sites are available to host the two radio distribution points and point-point link at zero cost, and that power is freely available to all sites.

The low numbers served, paying a typical £20/month service cost, does not leave a high margin for the operator. These examples can often work reasonably well on a small scale, and which may be manageable to a local operator who can keep costs down. The issues occur when this has to scale to provide service to a number of customer clusters scattered across the country and the associated support costs have to be covered. At this point, the model comes down to a matter of scale and external funding support.

5.3.4 WiMAX, 802.16

According to the WiMAX Forum, the industry organisation promoting and certifying this technology:

' WiMAX is a standards-based technology enabling the delivery of last mile wireless broadband access as an alternative to cable and DSL. WiMAX will provide fixed, nomadic, portable, and, eventually, mobile wireless broadband connectivity without the need for direct line-of-sight to a base station. In a typical cell radius deployment of three to 10 kilometers, WiMAX systems can be expected to deliver capacity of up to 40 Mbps. This is enough bandwidth to simultaneously support, for example, hundreds of businesses with 2Mbps connectivity and thousands of residences with DSL speed connectivity. Mobile network deployments are expected to provide up to 15 Mbps of capacity within a typical cell radius deployment of up to three kilometers. It is expected that WiMAX technology will be incorporated in notebook computers and PDAs by 2007, allowing for urban areas and cities to become "metro zones" for portable outdoor broadband wireless access.'

WiMAX is emerging as a last-mile broadband wireless access solution, which has the potential to make broadband service available in regions where it is currently not feasible, particularly in rural communities. The benefit of WiMAX is the open standard which has encouraged the massive interest in the development of equipment. However, the competitive environment for WiMAX will be tough over the next few years as it will have to exist in a competitive yet complementary fashion with cellular and DSL services.

Along with stiff competition from these established wired and wireless rivals, there is also uncertainty at present over spectrum allocation. This is a particular problem in the UK and the rest of Western Europe. WiMAX depends critically on spectrum allocation, a process that is still under way around the world. Of most interest to the development of WiMAX as a whole are the profiles developed through the industry organisation the WiMAX Forum, for frequency bands at 2.5 GHz, 3.5 GHz and 5.8 GHz.

Initially, the lead band in many international markets is expected to be at 3.5 GHz. Due to availability issues around 3.5 GHz in the US, however, the lead market there is likely to be at 2.5 GHz. The availability of both of these frequency ranges in the UK is limited which means the unlicensed upper 5.8 GHz band will see the first developments. Being unlicensed causes implementation concerns for the major operators due to maintaining quality of service and for a major operator like BT to consider WiMAX, they would require licensed spectrum to ensure their quality of service and enable them to build carrier-class services.

Pipex, who own the 3.6 GHzWiMAX license in the UK, has received an important boost for its WiMAX plans having secured the backing of Intel. Intel has agreed to inject $25 million into a joint venture, called Pipex Wireless, which has plans to roll out a high-speed Internet service across the 10 largest cities in the UK and take on mobile phone operators, such as Vodafone. The venture will take over Pipex's 3.6 GHz spectrum licence and will be able to supply Internet access to laptop users anywhere within a large conurbation.

It is important to note that the 5.8 GHz band is not yet certified for use with WiMAX. Where operators say they are using WiMAX at 5.8 GHz, this is not real WiMAX and remains a proprietary solution. However, by virtue of being unlicensed, this is hoped to be an exciting and fast-moving area especially for innovative wireless ISPs. Developments in the US at this frequency should help similar developments in the UK as equipment costs reduce in response to increased demand. Section 5.6.5 addresses the possible spectrum allocations in the UK suitable for WiMAX in more detail.

WiMAX could play a key role in providing connectivity in areas currently poorly served by wireless broadband. Initial implementations are seen as backhaul options for WiFi networks in place of DSL, satellite or more expensive leased line solutions.

In other geographies, particularly in developing countries where existing copper infrastructure is not available or is very limited, WiMAX is seen as the real opportunity to provide broadband services and is likely to be the key growth area for this technology.

At the unlicensed band of frequencies, where more rural and in-fill operation is likely to occur, many countries allow higher power output which makes this band more attractive to WiMAX applications. In 2004, the US regulator, the FCC, adopted the Rural Order, which it believed would promote greater deployment of wireless services in rural areas. This Order is designed to facilitate the deployment of new and advanced wireless services and is also intended to promote access to capital for entities seeking to serve rural areas. Among the Commission's decisions that are most relevant to the deployment of WiMAX are provisions that permit higher power limits (up to 25W) in rural areas. The Commission believes that this rule change will benefit consumers by reducing the costs of infrastructure and otherwise making the provision of spectrum-based services more economic in rural areas, where spectrum interference and congestions concerns are not so pronounced.

In comparison, the UK has power limits of 2W compared to the 25W available in the US, which will limit the potential for WiMAX longer distance links in the UK.

There are in fact two WiMAX standards to consider. The first, which is relevant to the discussion above, is the 802.16d or 802.16-2004 standard designed for fixed, nomadic and portable use. However, the main interest in the market is in the forthcoming 802.16e standard, which is being developed for mobile use. This is seen by industry as being the main opportunity for WiMAX in developed countries and a potential competitor to 3G developments - although there are arguments on both sides on this.

To conclude, we do not see any immediate options for WiMAX to address the not-spot issue in place of 802.11 WiFi due to the low return on investment, the developing state of the product and equipment costs. There may be opportunities for backhaul connectivity through WiMAX in unlicensed spectrum to replace DSL or satellite, but again this will be at a higher cost.

5.3.5 3G Development

3G networks, which typically provide 350kbps speeds, are developing with the introduction of HSDPA (High Speed Downlink Packet Access) this year and HSUPA to follow. It is important to recognise the difference between headline theoretical speeds capable within a base station, and the actual performance experienced by the end-users sharing the capacity within that cell.

HSDPA significantly improves the performance of the 3G downlink and infrastructure and handsets are now emerging with widespread deployments anticipated during 2006. In theory, HSDPA is likely to triple the average data rate experienced by individual users ( e.g. to over 1 Mbps), halve the access cost per megabyte, and halve the downlink delay. ¹¹

HSUPA is likely to be available in products by 2007. HSUPA will achieve speed, capacity and latency improvements that complement the downlink improvements brought by HSDPA. Collectively, HSDPA and HSUPA are commonly referred to as HSPA.

Whilst HSPA provides significant capacity enhancements compared to basic 3G, it will not support sufficient capacity for some future services that are being envisaged, e.g. broadcast mobile TV, broadband Internet access to the home with Quality of Service comparable with fixed broadband services. 3G LTE (Long Term Evolution) is the study item that is addressing capabilities beyond 3G, towards 4G. This study item has concluded on key future requirements and a road map. Its detailed implementation is not envisaged until the next 3G specification (Release 8), anticipated in 2008. This would suggest widespread deployment in 2009-2010. To conclude, 3G is unlikely to contribute towards solving the not-spot issue.

5.3.6 Other Spectrum

Ofcom, the UK telecommunications and media regulator, has completed a series of reviews of spectrum supply and demand since it took on its regulatory functions in 2003. They have subsequently announced availability of a number of frequency bands that it expects to make available to the market over the next 2-3 years. This has been made available largely through refarming of existing systems or from release of spectrum from MOD or other Government users.

In total, Ofcom is making spectrum available in about twelve different frequency bands; however, some of the bands have relatively limited bandwidth and, hence, are only likely to be attractive to more niche users.

One of Ofcom's current core objectives, as set out in its Annual Plan, is to advance broadband development. Wireless technologies are seen as providing a significant role in achieving this, both by extending competition in areas already served by broadband, and by extending coverage.

A number of the bands that Ofcom intend to make available to the market over the coming 2-3 years may be suitable for provision of broadband wireless and/or mobile services. Notably, some of these bands fall within the intended frequency range of WiMAX products and hence may address the issue of spectrum availability for WiMAX in the UK.

However, whether the new licences will extend availability of broadband services within Scotland will depend to a large extent on the conditions attached to the licences. For some of the frequency bands on offer, it is likely that the main driver behind interest in acquiring the spectrum may be to offer competitive wireless DSL services in urban areas (comparable with current DSL offerings), rather than to extend availability of services beyond those already covered by wired broadband solutions.

Within the next section, we briefly summarise the releases of spectrum that Ofcom is considering in the 2005-2008 timeframe. Focusing on those bands that may be suitable for broadband wireless services, we then discuss our views on likely uses of the spectrum, business case drivers and potential relevance to Scotland's broadband reach.

5.3.7 Potential Spectrum Awards

The bands that form part of the current spectrum awards programme that Ofcom intends to release over the next 2-3 years are summarised below.

Table 5.3: Ofcom Announced Spectrum Awards

Further details on the bands, and their potential role in developing new wireless broadband services within the UK, are described below.

5.3.8 470-862 MHz ( UHFTV Spectrum)

Our overall view is that, whilst the UHF spectrum may be a cost effective option for provision of broadband in-fill, there is currently significant uncertainty surrounding the amount of spectrum that will be released from the digital switchover ( DSO) and the timescales for the release. This places uncertainty as to whether spectrum will be available within the timescales required to address issues of broadband reach in the short term, and this spectrum looks to be more of a medium term prospect for broadband service delivery.

The Government has so far confirmed intentions for the five UK terrestrial television channels that currently broadcast in analogue ( i.e.BBC1, BBC2, ITV, Channels 4 and 5) to move to digital. Since digital television broadcasting is more efficient than analogue, potentially allowing more channels to be carried in less spectrum, there is the potential for spectrum to be released as a result of the switchover, which could be made available for other (non broadcasting) services.

The broadcasters have commenced detailed planning for the migration. In parallel with this, Ofcom has initiated a Digital Dividend Review ( DDR) project, to examine the options arising from the release of spectrum from the digital switchover. The spectrum may be suited to a range of new services, including mobile and wireless broadband. The propagation characteristics of UHF spectrum make it potentially suitable for lower cost, wider-area coverage and hence may be attractive to provide cost effective coverage in rural areas.

It has not been confirmed how much spectrum will be released as a result of the switchover and the broadcasters are lobbying Ofcom for additional broadcasting multiplexes to be made available within the released spectrum to offer more broadcasting services (such as High Definition channels carried on Freeview). There is also interest in the spectrum to be used for mobile multimedia services and a number of major manufacturers are backing the main equipment standards ( DVB-H and DMB).

The UHF spectrum could also be attractive for non-broadcasting services, and could make it possible for mobile or wireless broadband operators to offer 'rural fill-in', such as 3G/broadband speed services in more rural areas of the UK that are beyond the coverage of current services. This will depend largely on how much spectrum is available and the extent to which it is harmonised across Europe (to generate equipment economies of scale). Spectrum will be released regionally as the digital switchover occurs, but new services will need to be coordinated with digital television transmission. This means that there are two options available to Ofcom for making licences available:

• Issue licences for services to be rolled out regionally in line with the DSO
• Offer licences for national services to be rolled out once the DSO is complete.

From an operator perspective, there is likely to be less interest in offering services regionally as the DSO occurs, since this is a more risky proposition to deliver services (since customers may have to be turned away in areas where the DSO is not advanced and/or interference may occur). However, waiting for the DSO to be completed means that spectrum may not be available for some 3-5 years, depending on the timescales for the DSO to be completed (which depend on Government decisions with respect to 'switch off' of analogue transmissions).

5.3.9 1781-1785/1876-1880 MHz

The auction for this spectrum took place in April 2006 and licences were awarded to twelve companies. The spectrum sits between existing GSM1800 bands (used by Orange and T-Mobile for 2G mobile services) and the spectrum used for DECT cordless telephones in the UK. The spectrum is, therefore, subject to interference from existing 2G services and also, given its relatively limited bandwidth, is not considered to be attractive for wide-area coverage services. Hence, the licences that Ofcom is offering are restricted to low power use (maximum power 23dBm), which limits use of the licences to local, 'pico cell' type deployment (which could be used, for instance, to provide mobile enterprise solutions within buildings.

Given the restrictions on the licences, we view this spectrum as being limited in its uses and not likely to play a role in extending broadband availability.

5.3.10 1452-1492 MHz (L Band)

The spectrum available for award in this frequency band is 40 MHz and hence is sufficient in bandwidth to be attractive to operators for a range of possible services. Research that Ofcom has conducted on possible uses of the band suggests that there could be interest from industry in using the spectrum for either mobile multimedia services (using technology, such as DVB-H or DMB) and/or broadband wireless services using WiMAX or other wireless TDD technologies. The spectrum might also be of interest to provide further terrestrial digital audio broadcasting ( TDAB or digital radio) services, however, digital radio in the UK currently operates in a different frequency band (at VHF) and so this would require different DAB receivers than those currently on the UK market.

Ofcom is proposing to make the spectrum available in a number of 'lots', as summarised below (these are not confirmed and a consultation is ongoing to seek industry views on the alternative options).

Table 5.4: Ofcom's Proposals for L-Band Licences

All proposed licences allow the licensee to provide services nationally across the UK, however, there are no rollout obligations attached ( i.e. the licensee is not obligated to provide service to a defined % of UK population). This means that whether or not services are rolled out beyond urban areas depends on the operator business case and the services being offered.

Given the flexibility offered by the bandwidth packages that Ofcom is proposing, and since its policy is to make spectrum available on a technology neutral basis as far as possible, there is the potential that the spectrum could be used for a number of different services, which could include WiMAX-based wireless broadband.

However, our view is that WiMAX vendors and operators may favour the 2.5 GHz frequency band over L-Band to provide mobile WiMAX services based on the WiMAX Revision E specification. This is because of the greater potential for the 2.5 GHz band to be available on a global basis for similar services, which has meant that that the 2.5 GHz band is within the frequency range of current WiMAX products (the WiMAX Forum operating profiles covers frequencies from 2.3 GHz upwards).

WiMAX can also be deployed above 3 GHz (see Section 4.4.11 for discussion of UK availability of frequencies in this range).

5.3.11 2500-2690 MHz

This frequency band is commonly referred to as the '3G expansion spectrum' by the mobile industry, however, may be suitable for provision of wireless broadband/ WiMAX services as well as to provide additional capacity for evolution of 3G services as current services are expanded and take-up increases. As such, the 2.5 GHz frequency band is likely to be of high value to both 3G and potential WiMAX operators in the UK and across Europe, and hence likely to attract wide interest in an auction. This may push up the price of the available spectrum.

The 2.5 GHz band has been earmarked as expansion spectrum for 3G in Europe since the initial 3G licences were issued during 2000/2001. However, Ofcom's technology neutral stance means that it is unlikely that the spectrum will be made available for 3G services only and hence there could be opportunity for licences to be awarded for WiMAX services.

It is possible that existing broadband providers in the UK may be interested in acquiring licences in order to provide 'wireless DSL' services using the WiMAX Revision E standard. However, for cost reasons, incumbent operators may initially wish to provide services in urban areas, as an additional broadband service to complement existing portfolios (rather than as a coverage in-fill to provide services beyond areas already within DSL reach).

Unless Ofcom stipulate any specific licence conditions (such as rollout obligations, which is unlikely), there will be no regulatory conditions that require operators to provide connectivity in areas that are currently poorly served by wired broadband.

We are aware that Ofcom's intention is to issue a detailed consultation on licensing of the 2.5 GHz band by the end of 2006, covering considerations such as how the spectrum will be packaged (bandwidth offered, whether the channelling proposed is consistent with WiMAX standards, spectrum usage rights and other licence terms).

However, the fact that the 2.5 GHz band is widely considered to be the '3G expansion' band may drive up the cost of acquiring spectrum in this band (there is a risk of anti-competitive bidding, for instance, if mobile operators wish to acquire spectrum to prevent smaller and/or new players entering the market).

Whether or not 2.5 GHzWiMAX services benefit non-urban areas of the UK may, therefore, depend on:

• Take up of initial services in urban areas, which may encourage incumbent operators to extend availability of services to areas that are currently poorly serviced by other broadband services
• Whether or not a new entrant WiMAX operator acquires spectrum (since a new market player may be more aggressive in providing coverage beyond urban areas)
• The terms of the licences on offer - whether these are national or regional licences, and whether there are any conditions attached to the use of the spectrum.

5.3.12 2290-2302 MHz and 2010-2025 MHz

Ofcom has identified these spectrum bands as being potential bands for either 3G TDD services or for wireless broadband and has indicated that the bands may be auctioned during 2007/2008. However, their limited bandwidth is likely to make the bands less attractive than other spectrum bands that Ofcom is considering auctioning within the same timeframe (such as the 2.5 GHz band discussed above).

The bands are unlikely to attract high volume equipment supply (due to lower demand for services) and hence may not benefit from economies of scale. This could make licences within these bands less attractive to operators. As discussed in the previous section, our view is that incumbent 3G and fixed operators will favour the 2.5 GHz band over other bands to provide further 3G bandwidth and/or wireless broadband services.

Our view is, therefore, that these bands are unlikely to play a significant role in extending broadband availability, due to limited bandwidth and higher equipment costs making the business case less attractive both to incumbents and new operators.

5.3.13 2.7 - 3.4 GHz

The Cave Independent Audit of Public Spectrum Holdings indicated that MOD spectrum could be used more efficiently if sharing with commercial services is allowed. On this basis, work has begun on the identification of suitable spectrum for trials in the 2.7-3.4 GHz band. Initial sharing trials are hoped to commence during 2006 subject to satisfactory completion of safety assessments and funding provision.

The Public Spectrum Safety Test Group ( PSSTG) has directed that the North West coast of the UK be considered as a trial location as there is a suitable distribution of radar sites. Potential sharers have formed a Bandsharing Industry Forum to better co-ordinate their requests. This body is a commercial entity that is open to all in the industry and operates on a non-profit making basis.

Further information on this trial is currently limited and it is not clear what type of service may be offered and who might be interested in its provision. Any technical solution would have to be more cost effective than existing alternatives to be successful. However, this is an area for the Executive to be kept informed and make a judgement on its appropriateness as the trial develops.

5.3.14 Spectrum Above 3 GHz

The spectrum above 3 GHz that Ofcom has identified that could support further wireless broadband services is in the bands 3.6-4.2 GHz, 10 GHz, 28 GHz, 32 GHz and 40 GHz. All of these bands (with the exception of the 40 GHz band) have been allocated in the past for fixed wireless access, with varied success.

Various operators originally held licences to provide services in the 10 GHz and 28 GHz bands, for instance, including Scottish Power and Telecom (Thus), NTL and Cable and Wireless. Thus held a regional licence for the 10 GHz band (covering Scotland) and NTL and Cable and Wireless held national licences. However, we understand that licences have been returned to Ofcom, making this spectrum available for (re) award.

At present in the UK there are two fixed wireless access operators providing services using spectrum above 3 GHz - UK Broadband (now trading as 'now - next generation broadband') and Pipex Communications. Both of these operators provide services in the 3-4 GHz frequency band ( UK Broadband at 3.5 GHz and Pipex in the band 3.6-4 2 GHz).

Whilst both hold national licences, both UK Broadband and Pipex are currently regional providers of broadband wireless services and both are based in the south of England. Both have to coordinate their services with other services (particularly satellite earth stations) to some degree.

A recent development is that Intel has announced investment in Pipex, and it is understood that this may signal a move for Pipex to deploy WiMAX based services. A recent announcement from Ofcom has removed technology constraints that previously existed to the Pipex licence, which may pave the way for a migration to WiMAX. It is likely that WiMAX services offered above 3 GHz will be either fixed or nomadic broadband services rather than the portable/personal broadband services that may be considered using WiMAX in the 2.5 GHz band. This is due to the more limited propagation characteristics above 3 GHz, which prevents penetration into buildings and also requires additional sites to provide coverage.

We understand that Ofcom is considering options for making additional licences available above 3 GHz for fixed wireless access.

Depending on the conditions of the licences, we believe that availability of further spectrum in the 3-4 GHz region of the spectrum could play a role in the introduction of WiMAX services, possibly to areas currently poorly served by wired broadband networks. The cost of WiMAX rollout at 3.5 GHz will be higher than at 2.5 GHz due to the additional sites required, however, the cost of spectrum at 3.5 GHz is likely to be significantly lower than at 2.5 GHz due to lower demand.

Whilst Ofcom is strongly in favour of spectrum liberalisation and technology neutrality, there is a counter view that this runs against harmonisation and makes it more difficult to achieve economies of scale.

A further point for consideration is that it is likely that Ofcom will offer any licences as being 'technology neutral' and, therefore, there is no incentive for licences to be used to deliver end-user services -an incumbent fixed or mobile operator may, therefore, be interested in acquiring spectrum above 3 GHz to provide backhaul services within an existing network rather than provide new services.

5.3.15 Spectrum Conclusions and Regulatory Recommendations

In summary, our view is that whilst Ofcom is making a number of new frequency bands available, as indicated above, some of the bands will be more attractive than others for provision of wireless broadband services, depending on equipment availability and cost of equipment. These factors, in turn, depend on the extent to which the spectrum is harmonised for the same, or similar, use within Europe or other large markets.

Based on this, our expectation is that the frequency bands that will attract most interest from both incumbent wireless and fixed operators and possible new players, will be:

• UHFTV spectrum for cost effective coverage for mobile/wireless broadband services outside of urban areas (depending on licence terms, that Ofcom has not yet determined)
• L Band spectrum for 'mobile TV' services and possible WiMAX or other proprietary wireless broadband services
• 2500-2690 MHz (so called '3G expansion' spectrum), for provision of additional 3G-capacity and wireless DSL services based on WiMAX in urban areas.

We also believe that availability of further spectrum in the 3-4 GHz region could play a role in the introduction of WiMAX services, possibly to areas currently poorly served by wired broadband networks.

At present there has been no detail given from Ofcom as to when further licences might be made available and/or what the conditions attached to licences might be ( e.g. whether these might be regional or national licences). Since Ofcom's current policy is to auction national licences wherever this is practical (then allowing the licensee to determine its own business plan), we recommend that the Scottish Executive engage in discussion with Ofcom to understand how Ofcom sees these fixed wireless access licences playing a role in extending coverage of broadband to areas that are currently poorly served.

Questions in this regard include the extent to which an auction process may preclude smaller players playing a role in the market (due to the cost of participating in the auction and acquiring spectrum), and the extent to which offering national licences offers the best solution to solving specific broadband reach issues faced in parts of Scotland and other areas of the UK.

We also suggest that the Executive responds to the current consultation on higher power limits for licence exempt devices. There is the potential to improve the operation of WiFi equipment to deliver broadband access services in more remote areas by increasing power limits and hence range. There is a balance to be made against a blanket increase for all cases, which could cause unnecessary interference problems particularly in more urban areas. However, in our opinion, there remains a good case for increasing power on a geographical basis.

5.3.16 Powerline

Powerline technology use the low voltage electricity distribution network to carry broadband and has been trialled in Scotland by Scottish and Southern Energy. These trials looked at both the technical and commercial aspects of the solution as a potential alternative to DSL services.

The technical trials, which were conducted in Scotland at Crieff and Campbeltown showed that the technology could work, but there were a number of issues around coverage (which is dependent on distance from local sub-station) and possible radio interference from the electricity lines which could affect other users.

The commercial trials, which were conducted in Stonehaven and Winchester, pitched the solution against DSL services and in the case of Winchester, cable and wireless solutions as well.

The conclusions from the trials showed that the service was not a commercial proposition for the supplier at this stage and it is currently not being pursued further.

5.3.17 Satellite

Satellite broadband has the great benefit of wide coverage with the sole requirement being a clear view to the southern sky. This is the solution used in a number of very remote areas across the world to provide broadband directly to the end user. Satellite has also been used as a relatively low cost backhaul solution for small wireless networks. As an example, the Northern Ireland broadband programme, which delivers 100% coverage to households on demand, relies on a small number of satellite installations to meet the final few households where DSL is not available. This has been used purely as a last resort and reports indicate that users are experiencing performance issues with the solution. There are well known drawbacks with satellite due to the higher ongoing rental cost, and the performance issues in certain circumstances, which is due to latency or delay on the link. This high latency in satellite networks is due to the long distance from ground-satellite-ground (Round Trip Time, RTT), which adds about 520 ms to the transmission time.

Data transmission over the Internet uses a 'slow-start mechanism' at the start of a connection to find the appropriate parameters for that connection. Time spent in the slow-start stage is proportional to the RTT, and for a satellite link it means that the transmission stays in slow-start mode for a longer time than would otherwise be the case. This drastically decreases the throughput of short-duration connections. This can be seen in the way that a small website might take surprisingly long to load, but when a large file is transferred acceptable data rates are achieved after a while.

Despite all this, where there is only a choice of slow unreliable dial-up broadband at 28kbps or less or satellite, then for some customers, satellite can provide a useful service if the costs can be justified.

Typical solutions are based on an ADSL equivalent 512Kbp service with an installation cost ranging up to £1,000 in some cases and a monthly rental of £100 (compared with no installation charge and £15- 20/month commonly for ADSL offering 512Kbps as a minimum). Following the recent demise of Aramiska, who were leading in several rural broadband satellite schemes, there is some justifiable concern regarding the longer-term sustainability of broadband satellite in the marketplace.

5.4 Conclusions

This section has reviewed a number of technical options that may be available to increase the coverage of broadband services.

From the review of the BT position, it is clear that the development path within BT is focused on ADSL2+ through the 21 CN programme followed by trials and subsequent implementation of VDSL2 in the future. All of this is designed to increase the coverage of video capably broadband for the next development in services.

Investment by BT in new cabling and solutions, such as remote DSLAMS involves significant effort and has to demonstrate a clear return on investment. For small pockets of out of reach subscribers, this is unlikely to be the case, and additional softer issues, such as community inclusion have to be raised. From our review of the MAX product, which uses rate adaption to get the best performance out of the copper lines, it is clear that good performance can be achieved on some long lines.

We believe that the best option for pursuing BT to address not-spot issues will be to encourage the upgrade of the local loop cabling, where serious performance issues occur and significant clusters of users are affected.

However, there will be areas where investment in local loop will not address the issues of delivery to the more isolated clusters and in this case, alternative options must be investigated. From our review of the available options, two solutions stand out as offering potential for delivery. The first of these is WiFi networks, which have been successfully implemented in a number of localities including towns and villages in Scotland. The issue with these networks is around funding and sustainability due to the need for DSL, satellite or other backhaul, and this will be addressed in Section 6.

Finally, for those instances where broadband requirement is extremely isolated, and provision of wireless is not practical, the solution could be a one-off satellite installation.

« Previous | Contents | Next »