|
Response to "Proposals For The Licensing Of Radio Spectrum At 28 GHz" - Kingston Communications (Hull) |
Kingston notes the following proposals in respect of the 27.5GHz to 29.5GHz frequency band and is pleased to offer the following comment:
Kingston believes that this is a balanced way to offer the spectrum forward for auction, but notes that this leaves part of the spectrum of some 448MHz defined in the BFWAtg(00)11.rev1 document (27.8285-28.0525 GHz and 28.8365-29.0605 GHz). The RA will be required to give details to CEPT of its designation of this spectrum with respect to FS (Fixed Service) and FSS (Fixed Satellite Service). Notwithstanding any CEPT timetable, Kingston formally requests that the RA states it's designation i.e. primary FS or FSS in the Information Memorandum. The timing of and degree to which this spectrum will be available in the future for FS e.g. for a Time Division Duplex operator, must be made clear. The impact of any extra competition in the market via this spectrum needs to be taken into account by bidders for the duplex licences.
This is noted. The industry preference for a comparative selection process at 28GHz, has been over-ruled, and the prime argument of getting spectrum into use quickly is not accepted by the majority of telecommunications operators. It is still our belief that the RA should seek to exclude operators, on a primary basis, that already have access to a ubiquitous broadband local loop to encourage competition in broadband service delivery.
Given the greater complexity and both commercial and technological risk of the 40GHz frequency area, the RA should reconsider the proposal for an auction for this spectrum in favour of a comparative selection process. This would then allow weighting in the decision process, to promote operators without an established local loop, and those with clearly conformant multi media delivery services, rather than the auction that as set, is a simple financial exercise.
This principle will ensure a spread of 28GHz operators in a region, but will result in a very complex bid evaluation situation. Our comments on the auction proposals outlined by Joe Sonke at the 2nd May BFWACG Meeting, are detailed below.
This appears to be an unnecessary and unenforceable restriction unless it is incorporated into the use or loose criteria below. KPMG have identified that the 29GHz band will be a niche one for business services, and Kingston agrees with this. Current technology designed for use in this band is capable of supporting both variable rate and fixed rate services. For this condition to be guaranteed, the RA will be required to formally rule on the admissibility or otherwise of any point to point link that an operator wishes to establish within their spectrum. It is not believed that the RA would wish to be this closely involved in the deployments of any operator.
Kingston notes this, but comments that one of the prime arguments for an auction was that market forces would bring the spectrum into use more efficiently than a beauty contest. Use it or loose it criteria should therefore not be onerous in the first instance, and further, reflect an ability of an operator to stagger regional and multi-regional roll-outs. Finally, some technology solutions have yet to become available, and appropriate lead times must be allowed. The criteria need to be fair, and what is fair on a large operator could be regarded as extremely onerous or unfair on a small operator. Kingston has no real detailed advice on the establishment of such criteria but believes that the provision of network service to third party commercial customers demonstrates the commitment required.
Kingston is seeking for a clear description of all aspects of the auction, in the 28GHz Memorandum of Information. Quite specifically, the rules for allocation of spectrum packages within regions with respect to bidders ie, specified package bid or highest bidder first choice etc. The RA has clearly recognised the complexity of the technical co-ordination requirements and their potential to impact on bid prices. Clear simple rules on the timing of regional allocations ie at point of a clear result in a region or as a national exercise after bidding on all national regions is complete etc, must be available to bidders in the Memorandum of Information.
The RA has indicated a simultaneous auction of all regions, with the highest regional bidder barred in the next round. This appears appropriate, but the indicated speed of up to 10 rounds per day seems loaded against a fair assessment of the results of any round, before the next bid deadline. Further in such a rapid process, dissemination of the results of a bid round must be equally timed to each auction participant or there would be valid cause for complaint from individual bidders. Clearly an electronic dissemination system should be employed, but one that is not discriminative. Kingston believes that in fact, no more than 4 rounds should be attempted per day.
The proposed rule to limit bidders to a declared maximum number of regions but with transfer options between regions is a further factor that will require reasonable assessment time for bidders.
The decision to proceed by auction has now been taken for the 28GHz band. We have commented above on this as a precedent for the 40GHz band where the market and services are considered immature, and equipment is not readily available. Whilst this is still under study, Kingston proposes that there be a significant retention by the RA, rather than full disposal of the 40GHz band at this stage. The RA needs to ensure that 28GHz spectrum and 40GHz spectrum is completely de-coupled. The market for the latter is considered to include domestic subscribers, but that will not be possible before cost reduced (ie technology with volume production economics) equipment is available.
Kingston agrees that the optimum balance between the size of 28GHz licence areas and spectrum packages is three equal licences in a region, each of 2 x 112MHz. If however demand does not materialise for each licence in each region, then the balance of spectrum should be retained for a minimum of 1 year. This would enable successful licence bidders to establish service, and the commercial value of the un-bid spectrum be more readily determined at a time of future release.
The duration of the licences must be significant: a minimum of 15 years is proposed. Further, it is believed that operators need as free a hand as possible to change equipment and modulation schemes during the course of the licence.
Kingston notes the conflicting considerations in this area and supports the proposal for 12 English regions, plus Scotland, Wales and N. Ireland as single individual licence areas. The technical argument that the greater London area cannot be subdivided is accepted, as also the argument that the economic potential of each region cannot be equalised. Clearly, differential bidding will result to take account of this.
Kingston is generally pleased with the progress toward the 28GKz licencing, but believes that the RA has further critical detail to resolve and communicate about the auction process and arrangements. We will be concerned if "use or loose" criteria are too stringent or there are undue service restrictions placed on what has been identified as an evolving niche market for the business sector.
We have expressed concerns about the format of the 40GHz process and many of the issues, now virtually resolved for 29GHz, will be more difficult to deal with. It is our belief that to revert to a comparative selection process for the 40GHz spectrum, will be the best course of action.
This threat will be addressed through the imposition of a ‘use it or lose it’ (UIOLI) obligation on all successful applicants. This could be based on evidence that licensees had constructively built out services, without setting specific coverage targets. This would address the danger of incumbents strategically applying for spectrum, but it is also justifiable to place it on new entrants to encourage early and effective use of the spectrum resource. It would in addition guard against the possibility of speculative bidding for licences. The intention is therefore to design a suitable licence condition that guards against spectrum hoarding while avoiding unnecessary interference with legitimate commercial strategies.
 |
