Pricing to licensing policy: India and the 5G network conundrum

The new thing in telecommunications is 5G wireless networks. Compared to 4G, 5G is capable of delivering much faster data transfers, much lower latency, more network capacity, etc. Actually, 5G is a grab-bag of multiple technologies. It can be used to create private Wi-Fi networks, and small cells can be used to create fast “mini-networks”. The extra speed and ca­p­acity means massive improvements in str­eaming video services as well as the po­ssibility of entirely new services, involving virtual/ augmented reality, scaling up of IoT, machine-machine interactions, etc.

But 5G isn’t all that new. It’s been rolling out globally for over two years. As of September 2021, 99 nations were said to have working commercial 5G networks, and it was available in over 15,000 cities with 5G subscribers edging close to the 1-billion mark.

India wasn’t one of those nations, although Indian operators have tested and developed their own versions of 5G technologies and services, and the government has encouraged this experimentation by setting up a test-bed at IIT Madras.

As with any wireless telecom network, running 5G involves using spectrum. Governments have to make policy decisions to auction and license spectrum, before operators can build networks and technologies around them.

In 2018, the Telecom Regulatory Auth­ority of India (TRAI) set reserve spectrum auction prices too high to be attractive to a cash-strapped and debt-laden industry. In January 2022, TRAI has asked for public comment and suggestions about 5G spectrum and a new 5G auction is due to be held in March 2022. Assuming telcos buy in, it will be late 2022 or early 2023 at the least before networks are rolling. So India will be three years behind the curve on 5G.

There are further potential problems with 5G spectrum licensing policy, apart from pricing. The networks can be run on many different frequencies. (Frequency is an inverse of wavelength — the shorter the wavelength, the higher the frequency).

Some of those frequencies (60GHz, for example) should ideally be reserved for small, localised private Wi-Fi, without being commercially licensed. It is feasible to use 5G to connect up a building or office with a private 60GHz network, which works like a super-efficient data-enabled intercom service. Some of those frequencies should be used for big public “macro” networks. And some of those frequencies should be reserved for satellite broadband services.

The ITU (the global organisation that assigns India a +91 number and ensures international interconnectivity) has non-binding recommendations about specific frequencies, which should be used for satellite broadband only. If the same frequencies are used for satellite communication and terrestrial 5G, there is interference. If the same frequencies are being auctioned in Indian 5G auctions and deployed in terrestrial 5G, there will be problems later with setting up satellite broadband, since existing satellite operators use those bandwidths.

An actively dangerous version of interference is happening in the US where one of the frequencies used by commercial aircraft/airports is very close to a frequency licensed out to telecom operators. Verizon and AT&T are deploying C-band spectrum for 5G in the 3.7-3.8 GHz range. Aircraft use radio altimeters broadcasting at 4.2-4.4 GHz, which is close enough for interference. If altimeters malfunction, an aircraft may not know how far off the ground it is. Europe doesn’t have this problem since its 5G network deployments are not in frequencies close to aviation bands.

Why does reserving certain frequencies (the millimetre [mm] wave band of 24.25-28.5 GHz) for satellite broadband matter? Wireless networks are hard to set up and maintain in difficult terrain — the Himalayas, the Northeast, the Ghats, the Nilgiris, afforested regions, where between 20 and 25 per cent of Indians live. Moreover, “beefing” up capacity on a terrestrial network — strengthening the backhaul as it’s called — can often be done more easily and cheaply by picking up a satellite signal than by laying miles of optic-fibre. The Satcom Industry Association, which is the relevant industry lobby, has been advocating reservation for certain bands, but it appears from the TRAI documentation asking for public comments that the government may offer these for dual use. This could lead to later technical complications and another round of endless controversy. One can only hope that the spectrum policy will be designed with these pitfalls in mind — drawing on the examples of all the countries that have already rolled out 5G.