Improving the odds for 5G

Fixing the approach for networks and services is necessary for getting to 5G

Towards the end of 2021, the government acted admirably in the public interest, making difficult and courageous decisions to partially resolve legacy problems in telecom, such as redefining adjusted gross revenues and withdrawing retrospective tax demands. This gave rise to expectations that reforms would be less constraining, and more beneficial for the public interest. The announcement in June of 5G spectrum auctions and limited E-Band backhaul allocation, alas, falls far short. Why the disappointment, and how might the situation be improved?

Two areas need changes: Backhaul, and access spectrum. One aim is to remove self-imposed constraints. Another is to replicate the successful pattern of other sectors, where enterprises build their business, earn revenues, and then pay taxes. For some reason, this does not apply to communications despite it being a critical essential service.

Backhaul – E-band

A fundamental problem is the limited backhaul spectrum. Two bands of 250 MHz of E-band spectrum (70-80 GHz) are to be assigned to each telco that wins 5G spectrum. This is puzzling, as there are 10 GHz available (20 times 500 MHz). If press reports that this is temporary are correct because these frequencies are to be auctioned— and telcos must agree to pay the auction price at that time — the situation is disastrous, because 5G needs large backhaul capacity.

In other countries such as the US or the EU, policies are framed so as to enable usage of the full 10 GHz at minimal cost. This provides flexible capacity for much higher throughput, whereas we are creating a self-imposed constraint by restricting capacity. This means that the Organisation for Economic Co-operation and Development countries gain more from increased productivity than a developing country with the same resource. This can be remedied by adopting their regulations as appropriate.

The public interest is well served when policies enable telcos to use available resources to increase productivity and efficiency, instead of being obstructed in delivering services, or having to spend more to achieve comparable service levels. The same goes for the discriminatory allocation of spectrum to non-telcos (private companies), who will apparently be allotted spectrum on preferential terms compared with telcos.  India’s networks cannot compare with those of the OECD countries that have such policies, which is why such allocations would be too disruptive here.

Just as we are deprived of network capacity because spectrum is either not permitted for use, or is extortionately priced (both self-imposed conditions), limiting wireless backhaul or pricing it high will hamstring 5G and even 4G, because wireless backhaul will not be deployed extensively if the cost is prohibitive. These circumstances are aggravated by the threat of auctions for backhaul spectrum.
 

Illustration: Binay Sinha

Without extensive wireless backhaul, access spectrum from auctions cannot be fully utilised because of limited direct fibre connectivity.  Unlike countries that have good wireless backhaul— business hubs such as Mumbai’s Bandra Kurla Complex or Nariman Point, Gurugram’s DLF Cyber Hub, Connaught Place or the Dhaula Kuan, — the Airport stretch in Delhi, hospital and medical research complexes, engineering and manufacturing clusters, and so on, are denied vastly improved communications and access to data using gigabit wireless mesh networks. Even the spread of 4G small cells is constrained, reducing efficiency and productivity.

It would be useful to base India’s regulations on what others such as the US, the EU, and the UK are doing with E-band. They require non-exclusive nationwide licences, with mandatory coordination and link registration (usually through a geolocation database).  Our comparatively less developed networks make it inadvisable to adopt their policies wholesale, because it would disrupt equitable network development in India, accentuating the divide by skewing investment to the most profitable areas.

Our authorities need to focus on setting up and institutionalising processes such as non-exclusive licensing to telcos, and the mechanism of geolocation databases for backhaul spectrum for mandatory coordination and registration. Policies need to be framed so as to help build and grow gigabit wireless links to reinvigorate the sector, to recoup its stellar trajectory and contribution as in the past.

An instructive example is the approach taken by the City of London, or the “Square Mile”, in addressing communications for its 400,000 workers every day, 10 million visitors annually, and 9,000 residents. The mix of historical buildings and modern architecture poses a challenge for mobile network services. In 2017, the City initiated a project for providing free public gigabit Wi-Fi throughout the Square Mile (1).  The design incorporated 4G small cells for better connectivity, and the City offered 3,000 street assets (such as lampposts), stipulating that the system must be a neutral host open to all service providers.

The project was awarded to a joint venture of which one company is active in India. The backhaul uses a self-organising millimetre wave (mmWave) mesh as a “neutral host” that enables use by multiple operators. The mesh gives all service providers gigabit backhaul and access applications at 12Gbps with its 60GHz mmWave access and backhaul.

5G access spectrum

Assuming the objective is ubiquitous 5G and other services, India needs a different approach.  Spectrum auctions will not get us there. We achieved a level in mobile telephony by adopting a reasonable revenue share on licensing around 2003 after NTP-99. The same needs to be done for spectrum. 

One proposed approach (2) suggests that as there are only three serious telecom operators, spectrum can be allocated equitably to all three without auctions. This seems reasonable, as the funds diverted to auctions could then be invested in networks, and collections from revenue sharing are likely to far exceed collections from auctions, as they did after NTP-99 (3). The downside is that it would require the development of three networks, unless the operators share infrastructure. An alternative approach would be mandatory infrastructure sharing with one neutral host network, or two competing networks owned by different consortiums. 

If the authorities could take these points into consideration in improving the regulations, we are likely to have better outcomes in terms of networks, services, and the state of the sector.

shyamponappa@gmail.com

(1) The City of London Case Study: https://www.gsma.com/futurenetworks (2) Rajat Kathuria and Mansi Kedia: https://indianexpress.com/article/opinion/columns; (3) See Chart 2: Telecom Auction Fees Foregone vs Licence Fees + Spectrum Charges  at: https://organizing-india.blogspot.com/2020/08/configuring-indias-digital-ecosystem.html