Fix Digital India’s cycle tracks first

Optic fibre is an obvious solution for connecting various mobile phone towers, but laying it continues to be an uphill task given the municipal permissions required.

Given the urgent need to augment wired or landline broadband capacity in the country, it is just as well that the government is planning to scrap licence fee on this completely; the 8% licence fee is reportedly to be replaced with just a one rupee annual fee. While mobile broadband speeds have risen considerably over the last year or two, this is still no substitute for wired broadband, more so when work-from-home is increasingly the norm; it is only when 5G technology is used, thanks to its particular characteristics, that wireline speed can actually be delivered wirelessly.

Of course, there is an equally urgent need to dramatically lower both licence fee and spectrum usage charges for mobile telephony since, while wired broadband is faster and more stable, mobile broadband is a lot faster to deploy. That is why, while India has around 2.2 crore fixed broadband connections, it has 64.2 crore mobile broadband connections. Apart from the costs involved in putting in individual copper/fibre lines to each house, getting permissions to dig, etc, is what makes sure delivering wired broadband will always be much slower than mobile broadband.

If the government is not doing this, it is because the bureaucracy has convinced it that revenues will suffer. Much of this, however, is lazy thinking; indeed, it is the same kind of thinking that, in the 70s, led to marginal rates of income tax rising to 97% or so; and just as high tax-incidence ensured there was poor growth in tax collections in those days, high licence fees and spectrum charges have ensured government revenues have, not just slowed, they have contracted, from Rs 70,241 crore in FY17 to Rs 32,065 crore the next year and to Rs 58,990 crore in FY20; indeed, in the last four years, there have been no fresh auctions as telcos have been too financially stressed.

In such a situation, it is critical to nurse the mobile industry back to health, and the way to do that is to scrap or substantially lower licence fees and spectrum usage charges. That way, even if revenues from these levies go down—the blinding competition which the government ignored played a very big role in revenues collapsing as well—the government will earn more out of spectrum auctions; a low annual levy can be considered to ensure revenues don’t fall since the usage will rise a lot. More important, since telecom is the bedrock of the digital revolution, revenue maximisation should, in any case, not be the objective; how can telemedicine really take off, to cite one sector, if there isn’t great connectivity in even the remotest part of the country?

To get faster mobile broadband speeds, though, the government will also need to move on two or three other fronts; getting telcos back in the black, of course, is critical as nothing can work till this is fixed. But, that apart, one of the main reasons for why mobile internet access is slow is because telcos can’t move data across their mobile towers efficiently as the microwave links between them are not fast enough; so the signal travels fast from your phone to the tower—and back—but it then slows down there. Optic fibre is an obvious solution for connecting various mobile phone towers, but laying it continues to be an uphill task given the municipal permissions required. Also, in congested cities, it may not even be possible to dig everywhere; as a result, just 20-25% of all mobile towers are connected via optic fibre today.

In such a situation, the E-band (70 GHz) is a great substitute as it allows a much higher amount of data to be hauled to/away from towers than conventional microwave links. And, given there is so much spectrum available in this band—around 5,000 MHz—there is no problem allocating this to telcos in place of their current microwave links. Yet, while talk on doing this started at least two years ago, there has been no movement on this since.

And, if the government wants to promote faster internet using wifi—including in hotspots like railway stations and airports—it needs to delicense the V-band and even the 6 GHz band; indeed, the 6 GHz band is better as this is a lower frequency than the V band. The signal attenuates fast, so it can’t be carried too far, but unlike the conventional wifi which can only give you a carrier size of around 20/40 MHz, the V-band allows a carrier size that is 50-100 times larger, making it ideal for truly amazing speeds on wifi within a building or a limited geographical area. The 6 GHz band has a different architecture, but is also a good way to increase wifi speeds significantly.

And, while Trai was incorrect in its belief—in the case of Vodafone Idea and Bharti Airtel—that higher-speed internet for those who paid more violated the principle of net neutrality, the regulator was right in the sense that, when there are enough subscribers to fill up the network, higher speeds for certain subscribers will lower speeds for the others. If that is to be fixed—the way to do it is to prescribe minimum speeds for different categories of subscribers—the government just needs to ensure telcos have adequate quantities of spectrum. Right now, however, vast quantities of spectrum remain unsold; all the spectrum is unsold in the 700 MHz band, it is around 50% in the 800 MHz band, 20-25% in 1,800 MHz, etc. Since Trai has, often enough, tried to ratchet up prices across auctions—irrespective of demand—it will be critical to get it to reconsider the way it goes about its business (read bit.ly/2BncVc2 to see how Trai distorts spectrum prices).

Unless prime minister Modi works on this urgently, including putting a fair and unbiased regulatory system in place, not just telecom, Digital India too will lag considerably. You can’t run a high-speed car on what are, essentially, cycle tracks.

sunil.jain@expressindia.com
Twitter: @thesuniljain