Access to bandwidth: bringing higher bandwidth services to the consumer

A response by the Service Providers Interest Group

to the Consultation Document issued by Oftel in December 1998

The Service Providers Interest Group (SPIG) represents the interests of businesses in the service industries that provide a significant element of telecommunications including content, data, Internet and mobile.

SPIG believes that fair competition in the provision of telecommunications services to customers should be the priority for regulatory action in order deliver competitive pricing and innovation. Fundamental to the achievement of competition is access to infrastructure.

SPIG members believe that the principle of unbundling all network elements is necessary in order to bring real diversity into the marketplace and to encourage new entrants to supply services without an excess of environmental disruption. The extent of unbundling should be dependent on market demand. In principle access to infrastructure should be available at any point for which a market requirement exists. SPIG members are agreed that effective implementation of local loop unbundling would be the single most effective way of further promoting effective competition in the UK market for telecommunications.

SPIG’s view is that the UK’s policy of promoting infrastructure competition has been useful in developing competitive access in cities and major towns, but that it has now ‘run out of steam’ - operators will not extend their infrastructure to small towns and rural areas, as the likely returns do not justify the investment. Wireless local loop remains to be proven both technically and economically. The local access copper network should be regarded as a national asset, to which all competing providers of services including BT should have access on equal terms.

SPIG welcomes the consultation by Oftel on higher bandwidth services and expects that this will be followed by further consultations on all other aspects of unbundling. As a matter of principle SPIG would request that the consultation be widened to include access to the local loop for basic telephony and data services, not just for the provision of higher bandwidth

services as envisaged by Oftel. SPIG is aware that previous Oftel consultations found little interest in LLU for basic telephony, but our view is that operators and service providers are

interested in providing a full range of services on unbundled local loops, including basic telephony as part of the package.

SPIG notes that there is little competition in the local loop in the UK today and this lack of competition is keeping prices high. BT has 85% of the local loop market. Unbundling the local loop of the incumbent operator to permit access by service providers or other operators on fair terms will enable competitive services to be delivered to customers.

SPIG members in each category of service provision identify a pent up demand for unbundled services.

• Mobile

Indirect access to the mobile networks is the subject of a separate consultation, to which SPIG will reply.

In principle all mobile operators have bottleneck control over access to radio spectrum, not only Cellnet and Vodafone, and access to each network is necessary in order to introduce a significant level of services competition into the mobile market.

Such operators should not be able to use the bottleneck control to restrict the connection of services, which they elect to supply themselves. The result is poor consumer choice.

Independent mobile service providers seek access to unbundled elements of the core network of the mobile operators on "cost plus" terms, so that they can satisfy the needs of their own customers, and supply combinations of services using different mobile and fixed networks.

• Fixed

Service Providers are starting to deliver BT service provider products such as Calls and Access on the basis that BT has an obligation to supply to independent SPs. The conditions applying to these arrangements should be preserved whatever the outcome of the current Oftel consultation.

Service Providers require access to basic telephony products so that they can be bundled with their own service offerings in order to satisfy their own customers' needs.

• Internet

BT has control over the local loop into every business and residence in the UK. As a result, BT has a unique advantage over Internet Service Providers, who are businesses offering Internet connectivity. ISPs have typically been able to offer real competition through the use

of alternative networks and approaches to delivering new telecommunications products. The inability of ISPs to access local loop reduces their ability to be innovative and competitive.

Access to the local loop is essential to allow the continuing exploitation of new technologies in the Internet market such as xDSL products. As faster access technologies like this come into use it is essential to avoid a situation where one player, BT, has a monopoly and can control prices. With access to the local loop ISPs believe that they could provide their own xDSL equipment at probably a much lower rate than they are currently charged for an equivalent

leased circuit. This opening up of the market to competing xDSL services would benefit all UK consumers, as the price of all "leased2 bandwidth would drop.

The European Commission has not mandated LLU, but clearly regards it as key to developing competition by unblocking access to customers by new entrants to the market, both operators and service providers. In its 4th implementation report, DGXIII reports on the state of LLU on each member state. Germany, Denmark and Finland have implemented LLU; Sweden will implement it when it resolves a legal issue; Italy has adopted it a policy; French operators want it; Spain, Portugal, Luxembourg and Greece have no plans for it yet.

Oftel notes in chapter 2 of the consultation document that LLU is being used in the US by competing operators to offer xDSL services.

In December the Australian Competition and consumer Commission announced that it would force Telstra, the dominant operator, to allow its competitors direct access to the local loop. The regulator said that it expected to see ' a greater range of innovative services delivered to end users more quickly and at lower prices'.

SPIG expects that LLU will be increasingly adopted in the leading countries in the EU and elsewhere to assist the development of competition for the benefits of consumers of all kinds and the economy generally; SPIG cannot see any logical reason why the UK should exclude itself from this group of countries.

SPIG agrees with Oftel’s assessment that the demand for services based on LLU will be greatest among SMEs, small sites of larger companies, teleworkers and heavy residential Internet users.

SPIG agrees with Oftel’s assessment that copper loops plus xDSL technologies are the best bet for wide and early access to higher bandwidth services for residential and SME customers.

SPIG believes, inter alia, that the Government’s targets for the UK to be a leader in e-commerce requires the widespread and early exploitation of xDSL technologies based on the existing BT copper local loop. In SPIG’s view BT’s control of 85% of local loops is a bottleneck acting as a barrier to effective competition that justifies regulatory intervention to ensure that innovative services are made available quickly at cost-oriented prices.

SPIG notes that options 1,2 and 3 require co-location of equipment in BT’s exchanges. SPIG regards co-location of equipment as an essential part of any future development of competitive access in the UK, and notes that it apparently works successfully in other countries.

Option 1 envisages the complete takeover of the local loop by the ADSL service provider. The proposal as it stands presents some technical problems (mainly spectrum management issues, as described in Annex 1 of Oftel's consultative document) and commercial barriers to entry by independent service providers.

It is suggested that these issues could be addressed by some small variations to the Oftel proposals laid out in Option 1. This is explained in greater detail in the following paragraphs.

A key theme reiterated throughout the Oftel Consultative Document is that of spectrum management and radio interference/ crosstalk issues. Indeed, a concern voiced in the Consultative Document was the possibility that the adoption of Option 1 and the complete
transfer of the ownership of the local loop would result in a "free for all" in the utilisation of ADSL, potentially leading to interference between and outside loops that could not easily be
controlled.

A further issue is the commercial barriers to entry imposed by the requirement for "broadband service providers" (e.g. Video on Demand or Internet SPs) to take complete control of the local loop and provide a full range of voice telephony services. This diverts resources and requires substantial investment in areas that the broadband SP may not be expert in (and may not wish to be). In effect, this requirement would greatly reduce the number of SPs able to compete at this level, this restricting the choice of services (and thus the standard of innovation and competition) available to the consumer of these services.

(As a side issue, it would also result in independent service providers who are building businesses based upon BT's new "Calls & Access" product losing their customers if they chose not to offer broadband services and their customers moved to a broadband service provider. Again, this restricts consumer choice and makes investment by independent service providers harder to justify, thus discouraging competition.)

It is proposed that these issues could be addressed with a slight variation on the scenario postulated by Oftel in Option 1, as follows:

ADSL is technologically "special" in that it can be operated over a loop without interference to any of the standard voice telephony operations on that line. Indeed, this was a specific design requirement, so that telcos could roll out ADSL services without the need to add a new line to the customer's premises, and thus enable them to build a business case for adding new services with a minimum of extra cost.

This is achieved by utilising unused frequency space on a telephone pair. Voice telephony functions use only the lowest 5kHz (or so) of the bandwidth available. The remainder is

available to carry higher frequency signals, such as those used by xDSL. Indeed, it is these
signals, and the power levels used to carry them, that cause concerns about crosstalk and spectrum management.

Thus, it would seem sensible to manage unbundling of the local loop by transferring control of the frequency spectrum within the loop to a separate body expert in these matters (perhaps the Radiocommunications Agency, who have historically dealt with issues surrounding all manner of frequency allocation, power levels and radio emissions and could reasonably be expected to have one of the highest levels of expertise in this area).

Once in control of the loop, this body would subcontract maintenance and engineering issues to the original owner of the loop (who would have the best information and expertise to maintain it) and allocate bandwidth on the loop to interested parties. This would be strictly
controlled by the new body on a completely impartial basis, as would technical standards for connection of equipment, thus avoiding many of the crosstalk issues and eliminating any "free for all" situation feared by Oftel.

Leaving the original owner of the loop responsible for maintenance and engineering should ensure there is no compromise of network integrity while the allocation of bandwidth would maintain full freedom of choice for the user and minimum barriers to entry for new
service providers (either of voice telephony or broadband services).

For example, for a typical BT voice line, the new body would allocate frequency space for voice to BT, and any engineering issues (e.g. repairing the pairs after a break) would be BT's responsibility. If the user then wanted Internet or Video on Demand from a broadband
service provider, the new body could allocate access to the higher frequency bandwidth to an independent broadband service provider without disturbing the BT voice service. Similarly, the user could switch to getting their voice service from another provider without
disturbing the data connection (subject to any outage in service for reconnection, of course), as both services are completely independent of each other.

This approach would lead to a far greater level of competition in both the voice and broadband data markets, and ensure the integrity of the network is upheld with the minimum of spectrum
management and engineering issues. It is for these reasons that this approach is put forward as an alternative to Oftel's Option 1.

SPIG notes that the rental prices of the local loop will need to be set by Oftel, and that if these prices are too high then LLU will fail to produce the market developments envisaged in Chapters 2 and 3.

Option 2 is effectively option 1 with BT as the body responsible for management of the local loop. SPIG considers that this combination of poacher and gamekeeper may look acceptable in theory, but is likely in practice to result in a playing field tilted in BT’s favour; for this reason SPIG favours the independent body suggested in our thoughts on Option 1.

Option 3

SPIG considers that this option, in which BT owns and operates all the xDSL modems, will not result in the earliest and most widespread adoption of the technology to maximise innovation and the most competitive prices. SPIG sees Option 3 as an unnecessarily conservative approach that will strengthen BT’s stranglehold over local access, which is the opposite of the desired goal.

These options, like option 3, leave control of the local loop in BT's hands, and allow BT to determine the nature of the high-speed access. As we understand it from the press, BT’s trials are somewhat in disarray at the moment. It seems that BT has focussed on a Video on Demand-type solution and their own "straight to SB Internet" service, but it would appear that is not what the market wants. We understand that BT are reviewing the trial and their plans, perhaps delaying commercial deployment until July. This indicates to us that assuming that BT will always get it right regarding what the market wants is unrealistic.

1. Service providers require access to basic telephony products so that they can provide their own bundled services to customers. They seek disaggregation and the ability to select their supplier.
2. The basic elements can themselves be broken down into speech plus ADSL: SPs want one or both.
3. Oftel must take care that the BT ADSL trial does not lead to the situation where BT ‘cherry picks’ the market before potential alternative suppliers have an opportunity to enter the market.
4. Spectrum management is an issue that needs to be and can be resolved - we suggest the establishment of an xDSL management authority.
5. All the technical issues are surmountable.
6. SPIG is concerned that Oftel appears to be seeking simple solutions, which will not best meet the needs of SPs, to enable them to deliver cost-effective innovative services to the benefit of consumers and the economy.
7. The UK must not be out of line with other countries. UK policy must be in line with best practice in the EU and the world.
8. SPIG supports Oftel’s option 1 with modifications, notably the addition of an independent xDSL management authority, and preservation of the obligation on the incumbent to supply ‘calls and access’ services to SPs.

If BT are allowed to proceed with their service launch plans, our expectation is that the Oftel consultative process won’t be complete (i.e. a determination won’t be issued) until well after BT go into commercial service.

Assuming Oftel decide to unbundle the local loop we envisage that is unlikely to take place before 2001.

This will result in BT either being given a monopoly of the supply of ADSL or a very substantial run-up to deployment in a monopoly situation. Either way, they stand the chance of being able to use their dominant position in voice telephony to dominate the high-speed data access market. In our experience and opinion this is unlikely to be beneficial to Service Providers or consumers.

Given that high-speed access to the Internet is likely to be one of the most significant developments in communications in the next few years SPIG submits to Oftel that BT should not be allowed to exercise a monopoly. Secondly, SPIG is concerned that under present rules, ADSL and Internet access will be in the BT’s SB, which will not result in transparent service costing and pricing.

SPIG suggests that BT’s SB should be required to sell "ADSL bitstream access" to anyone that wants it at wholesale prices, and that there should be no retail service in the SB; BT’s SSB should buy the same wholesale product and compete on level terms with other service providers.

In practical terms, only Service Providers will find an offering like this of value. They would buy the service, "add value" in some way (Video on Demand, Internet access, etc) and sell the resulting bundle of services at retail.

Service Providers believe that as part of the policy making process OFTEL needs to safeguard the continuing obligation for BT to provide local loop access to Service Providers in order to increase and sustain service competition in the residential sector. 85% of households use BT’s local loop for basic voice telephony. The latter will continue to constitute the bulk of the residential telecommunications market for the foreseeable future and is one of the least competitive sectors in the telecoms market. Service Providers are concerned that certain of the options described in the consultative document are likely to hinder the development of competition in the basic voice telephony sector as well as limit the future development of service provider competition in the broadband sector.

In particular, Service Providers believe that, despite its high pricing, BT’s Calls & Access product is a major step forward in the development of competition in that it enables the entire replacement of the BT bill with a service provider bill, and the customer relationship is with the service provider rather than BT. BT has tentatively indicated that in due course it intends to offer ISDN and other higher bandwidth products to service providers through Calls & Access. In order to sustain service competition, OFTEL needs to ensure that BT stands by this indication and offers higher bandwidth products to service providers as soon as is practicable.

Options 1 and 2 will result in control of the local loop being passed to another operator who has no regulatory obligation to offer a Calls & Access type product and will therefore substantially hinder the development of service provider competition to the residential sector. Options 3, 4 and 5 do not have this consequence. We analyse each option in turn below.

Option 1: Unbundled local loop
The new operator will provide narrow and broadband services to the customer and would have no regulatory obligation to offer a Service Provider product. BT, who does have this regulatory obligation, would no longer be able to provide a Service Provider product for that line. The net effect would be a reduction in choice for the customer for narrowband services, which constitute the bulk of the market.

Under Option 1, narrowband service providers will see their addressable market shrinking as broadband service providers effectively block them out of the market. This provides less incentive for service providers to invest in new products and the acquisition of new customers.

Customers, who require narrowband services only, are likely to remain in the majority for some considerable time to come, especially in the residential market. If Option 1 is adopted, and narrowband service providers are discouraged from operating, then a substantial segment of the market is likely to see less competition and therefore less choice.

Option 2: Partial Baseband Leased Circuit
Option 2 has the same impact on restricting choice for the customer as Option 1 although the contractual arrangements between BT and the new operator is different.

Option 3: Bitstream access
Option 3 is preferred to Options 1 & 2 as it does not restrict the customer’s choice for their basic telephony and narrowband services. However, as it stands, in order to compete effectively in the provision of broadband services, option 3 still requires service providers to invest in telecom infrastructure procurement and collocation. This is not consistent with the view that service providers add value by purchasing products off infrastructure providers and do not in general invest in telecom infrastructure themselves. For this option to be consistent with the development of service provider competition, OFTEL needs to ensure that BT offers all broadband services to service providers through Calls & Access.

Option 4: Permanent Virtual Circuit access
Option 4 has the same attractions as does option 3 in relation to the provision of service provider competition in narrowband services. For broadband access, Option 4 addresses the issue of duplicated infrastructure for the service provider, but under this option, bandwidth management between the exchange and the service provider premises remains with BT, thus enhancing and thus potentially sustaining BT’s market power. In particular, it is unclear who would bill the end user for the use of the bandwidth, as opposed to the use of the service (e.g. Internet access). For example, for narrow band Internet access, BT still bills for the use of the line and the service provider bills separately for the service. Under option 4 would BT still bill for the bandwidth? If so it is very important for OFTEL to ensure that BT offers high bandwidth products through Calls & Access.

Option 5: Indirect Access
Option 5 combined with carrier pre-selection offers customers comprehensive alternatives and therefore concurs with our aim of increasing competition in the residential market. We suggest that BT’s plans for introducing broadband to the public network be made public as soon as possible. This would mean that various service providers, and not just those selected by BT, can make adequate plans for development of products and services.

Oftel’s options 1 and 2, we believe, would inhibit the viability of narrowband service providers, and reduce the potential for service providers to offer broadband services. They would, therefore, significantly reduce competition in the residential marketplace for basic telephony. Options 3 and 4, we believe, represent workable alternatives, but we believe that it is extremely important that OFTEL ensures that BT offers broadband products to service providers via the Calls & Access product. Option 5 is workable. Implementation could be hampered by a lack of visibility of BT’s investment plans in relation to building a network where option 5 becomes viable.

SPIG’s suggested ‘modified option 1’ maintains the position of service providers using the Calls & Access product.

1. ISPA strongly supports the demand for unbundling the local loop as that is the only way to develop competition in the UK in existing and new services.

2. Through its control of the local loop, which it inherited, BT has a unique advantage over other businesses which offer Internet connectivity, typically Internet Service Providers (ISPs). This advantage is that ISPs are forced to include in their network infrastructure, an element over which, they have no control, the local loop. On the other hand, BT as a service provider is able to control ALL of its network, from end to end. Thus BT's current monopoly over access to copper means that they have an unfair advantage in providing bandwidth at a much lower cost to buildings with spare copper pairs than any other operator (who would have to dig up the road) could achieve.

3. ISPs have typically been able to offer real competition through the use of alternative networks and approaches to delivering new telecommunications products. The inability of ISPs to access the local loop reduces their ability to be innovative and competitive.

4. The significance of access to the local loop is now greatly increasing, as new technologies become available worldwide. Traditionally, large telecommunication companies have been slow to exploit and bring these technologies to market at competitive prices. ISDN is a good example. Whereas new technologies are more rapidly adopted and cost-effectively marketed through service companies, such as ISPs. Thus without unbundling the local loop, the UK would fall further behind in the exploitation of the Internet and telecommunications.

5. Access to the local loop is essential to allow the continuing exploitation of new technologies in the Internet market, for example xDSL modems. As faster access technologies like this come into use, it is essential to avoid a situation where one player, BT, has a monopoly and can control prices etc. This can only be avoided though unbundling the local loop. Given such unbundling, ISPs believe they could provide their own xDSL equipment at probably a much lower rate than they are currently charged for an equivalent leased circuit. This opening of the market to competing xDSL services would benefit all UK consumers, as the price of all "leased" bandwidth would drop.

6. Whilst recognising that there are technological issues in achieving direct access to local loop copper wire, and practical considerations which make ‘any-to-any’ copper wiring difficult, ‘any-to-exchange’, and access to the exchange should be possible. Even easier is ‘any-to-exchange’ and then interconnect trunk from the exchange (for instance access to ATM layer for ADSL customer connections, not just the IP layer). We know that enabling access to copper would be difficult and complicated, but that must be better in the long term for competition and development of new services than the effective monopoly BT currently enjoys.

7. It's worth noting there is no mandatory requirement to use the same copper pairs as carry telephone traffic. Most buildings have spare pairs already present. Houses are typically wired with two or three pair cables, for example, and an unused pair could be used to carry xDSL services. This might help smooth the path to unbundling by helping to prevent arguments about disturbing telephone service with DSL faults, and all the associated arguments about access to emergency services etc.

8. Alternative new fast access technologies may be offered by telecommunications companies other than BT, e.g. cable modems - which may not be opened to competitive carriers and ISPs. Oftel has no power to compel them to do this because the companies do not have significant market power. Access to the local loop may be the only way that ISPs can compete with these other technologies. Thus the survival of independent ISPs may depend on getting access to the local loop. ISPA firmly believe that if the Internet and UK users of the Internet will suffer if the innovative and dynamic smaller ISPs are driven from the market.

9. If Oftel truly want to be so successful that they are no longer needed, then the unbundling of the local loop is probably the first time that the opportunity has arisen. In fact once access to the local copper is allowed BT’s situation as having Significant Market Power could be expected to be rapidly eroded both through other carriers (e.g. Energis) managing calls and through ISPs carrying voice and other data as IP.