TRAI Consultation Paper on Spectrum

Q.1. What method should be adopted for refarming of the 900 MHz band so that the TSPs whose licences are expiring in 2014 onwards get adequate spectrum in 900/1800 MHz band for continuity of services provided by them?

Comments on Spectrum Refarming

1. Arbitrary Policies & Their Consequences
The proposed manner of refarming the 900 MHz spectrum is perceived to be as arbitrary as, for instance, the tax claims against Vodafone after the courts upheld its refutation of these claims.  Such actions contribute to India’s very low rating on contracts (184 out of 185 countries in enforcing contracts in 2013: http://www.doingbusiness.org/data/exploreeconomies/india/), and for being a very difficult place to do business.

2. Legitimacy Of Terminating 900 MHz Holdings Starting 2014
One question is whether a refusal to renew existing spectrum holdings in the ordinary course is legitimate, or if it needs to be tested for breach of contract in the courts.  This proposed manner of withholding access to assigned spectrum is also contrary to prevalent practice, as well as to the logic of spectrum being essential to the delivery of services of a wireless operator as a going concern.

3. Must All Spectrum Be Auctioned?
a) Another question is whether the Supreme Court order requires that all spectrum must in fact be auctioned.  If this is so, the auction of all spectrum is necessary when it becomes available.

b) This is so damaging to the public interest, however, that all reasonable efforts must be made once again to inform the Supreme Court of the facts, i.e., the technological reasons against splintering bandwidth, and the financial reasons against extracting payments that would otherwise be invested in the essential infrastructure of broadband.  If the facts are presented clearly and persuasively, there may be a reconsideration of the ruling to auction all spectrum in the light of these facts, as against continuing with this ruling based on miscommunication or misinformation.

c) If there is no alternative to auctions, to succeed, the reserve price needs to be relatively low, and bidders in difficult financial circumstances must be convinced they have no better option.  Perhaps one way of ensuring this is to auction for a shorter period, e.g., five years, while simultaneously laying out the path for transitioning to shared spectrum.  This is because parallel developments in spectrum sharing for Authorized Shared Access and Licensed Shared Access that are being pursued in the US and the EU are likely to be deployed by then.

4. Net Benefits of Refarming
Given the stage of evolution and coverage of networks in India, their technological level and usage, refarming should be held in abeyance until our markets are in a position to benefit from them.  This is because the detrimental effects if the 900 MHz band is cleared in the proposed manner are likely to far outweigh the benefits, as explained below.

5. Purpose of Refarming
What is the purpose of refarming?  If the answer is the potential benefits of services from 4G technologies and products, consider the likely nature of these benefits in India.  The purpose of refarming in OECD countries is to use 900 MHz for 3G and LTE for high-speed data.  This is appropriate for developed economies that have large numbers of data users.  In India, high-end users comprise only a niche segment (15.09 million broadband users in April 2013, despite over 725 million active wireless subscribers). Developed economies have refarmed the 900 MHz band because 3G and 4G assume widespread use of data services in the entire network.  In other words, if India had a large base of high-data users, 4G networks would be required to deliver high-speed traffic. Also, such users would presumably be (a) willing and (b) able to pay [for the expensive equipment required] for these services.

6. No Economic Basis for Refarming in India
a) The reality is that there are insufficient data users with the willingness and ability to pay for the higher level of throughput.  The present state of the economy and its trajectory pose additional constraints.  More important, existing technologies are capable of delivering data services at lower cost.  The priority is for access networks at lower cost, e.g., wireless middle-mile and last-mile that will enable large numbers of users to access data services at a reasonable price (“reasonable” in the cost structure of India comparable to TV services, and not in cents/minute comparable with OECD countries).

b) That said, a possible consideration is whether and how certain advanced technologies, such as “supplemental downlink” or “carrier aggregation” for augmenting capacity, may be made usable in our circumstances, and whether if certain bands are earmarked for them, such solutions can be introduced here.

7. Need: Low-Cost Last-Mile & Aggregation/Backhaul Capacity
How can networks be built at reasonable cost that have the capacity to deliver data services more comprehensively in India?  By providing much more wireless access for the last-mile, and more middle-mile capacity (in combination with existing wired networks).  This is where policies can facilitate network build-out and service delivery at lower cost.  The nature of required reforms are: reduced front-end charges for wireless last mile access; reduced microwave charges (administered prices) for aggregation and backhaul; incentives for broadband delivery, and perhaps higher incentives for rural broadband delivery.  Also, a whole host of initiatives can be orchestrated, as in South Korea, for instance, or Sweden, which contribute to the development of broadband services and usage.

a) South Korea*
South Korea’s digital economy resulted from a combination of macroeconomic, supply-side and demand-side policies and programs, with the government stimulating broadband adoption, particularly in the early years.  For example, Korea’s response to the financial crisis of 1997-1998 was to increase the export strength of key sectors such as electronics.  There was also a thrust on consumer credit, facilitating the purchase of consumer goods and electronics-related services such as broadband.  The initiatives to push broadband deployment and adoption included tax incentives, rural deployment and R&D grants, building certification incentives, and applications support.  There were also mistakes, as in the government’s choice of WiBro technology.

By way of illustration, a set of Korean initiatives are detailed below:

Table 5: Selected Korean Supply-Side Broadband Subsidy Programs

Infrastructure Deployment

• Tax benefits (credits, accelerated depreciation, exemptions, etc.) for broadband deployment

• Backbone provision or subsidy for broadband deployment (KII-Government program providing funding for operators to reach 40,000 govt. locations as well as rural districts)

Technology Support

• R&D grants and tax credits
• Applications support (KOREN—Korean Advanced Research Network or KII-Testbed) Building Certification & Codes
• Requiring or encouraging the pre-equipping of new buildings with fibre and/or broadband access points (e.g. DSLAM)
• Institution of certification programs for broadband readiness of MDUs (multi-dwelling units, based on three classes of transmission speed)

Source: Kalba International, Inc., 2012. Ovum Consulting, Broadband Policy and Development in the Republic of Korea

b) Sweden*
Sweden's regulator demonstrated a strong commitment to cover low-density areas, and one of the ways was to foster network sharing.  The government promotes a broadband strategy with incentives for all stakeholders.  State authorities are actively involved in the Digital Agenda for Sweden, a national initiative.  In education, for instance, about 25% of all students rely to some extent on distance education.  Open access policies and competition have had a significant impact on the development of broadband.

* Source: Digital Scotland 2020

Achieving World-Class Digital Infrastructure
21st December 2012
http://www.scotland.gov.uk/Resource/0041/00414982.pdf

8. Shared Facilities
One way that delivery costs can be reduced is if operators share networks, so that all operators can access these networks where they are licensed to do so.  This would be feasible if there were practical ways of structuring “common-carrier” or network-neutral access (as in roads, rail, flight paths and airports, ports, oil pipelines, etc.).  This would require a buy-in by service providers for radical changes in approach and policies, followed by radical changes in operating networks.  It is possible that open consultation with TSPs, other stakeholders, and specialists, done with the help of one or more expert facilitator/s, could yield such a solution.  If this were to happen, the process of organizing structures at (a) the wholesale (network services) level, and (b) the retail (user access level) could be addressed collectively.

9. Extend Voluntary Infrastructure Sharing to Mandatory Sharing
We already have consortiums for passive sharing of wireless towers.  This needs to be extended from voluntary commercial associations to mandatory, “common-carrier” access, after putting in place suitable commercial arrangements through negotiation.  Such commercial arrangements exist for oil pipelines and for oil exploration and production, and can be structured in like manner for facilities and spectrum.  They need the appropriate financial structuring with the help of financial specialists, in addition to the engineering solutions.

10. Shared Spectrum: Pool New Spectrum

a) An evolutionary step in this direction is to pool all available, unallocated [unassigned] spectrum, so that it can be shared by [existing] service providers.  This is being pioneered in the so-called TV White Space bands in the USA, the EU, the UK, and Singapore.  It can be extended to other bands here.  This could be a transitional step in evolving a shared facilities model.  Provided the stakeholders agree, and an equitable structure and process is devised, this will relieve the present constraints on spectrum availability by providing a common pool of spectrum.

b) Shared spectrum is an alternative that is technically feasible and economically far more viable than fragmenting available spectrum for the exclusive use of our many operators.  This also provides for complete transparency, as well as much lower capital and operating costs for society as a whole.  The implication is that broadband could be made available more widely at lower cost, leading to much better productivity and payoffs.

c) Many of the questions and associated problems would be resolved.  For instance, open access would allow for each operator to choose any technology that is compatible and that does not create interference.  Fees could be determined in the same manner as for taxes in inducing investment for manufacturing, as was done in South Korea.  It would need a whole range of supportive measures as in the case of South Korea, and if done right, could result in tremendous gains as an organizing force in society.

11. Some 900 MHz Access (Common-Carrier) For All TSPs
If a portion of the 900 MHz band is set aside for shared access, it may resolve one of the most contentious problems between the GSM and CDMA operators, of access to the highly advantageous 900 MHz for its low-cost equipment and ability to penetrate buildings, i.e., better delivery.  This step may create conditions that allow for stakeholder engagement for an overall resolution, including ultimately, shared infrastructure.

12. Revenue-Sharing & Consumer Surplus From Shared Spectrum/Networks
India’s experience with revenue-sharing after NTP-99 has shown that collections are far in excess of up-front revenues forgone.  Building a sound broadband service with a combination of incentives and forbearance will lead to much greater economic benefits overall, as well as much higher collections by the government over time.  The sector can once again prosper and be an engine of productivity.

Shyam Ponappa
Centre for Internet and Society
August 21, 2013