Improving India’s mobile phone service

India has notoriously bad cellular connectivity, especially in terms of quality—call drops, for instance. Sure, prices are about the lowest in the world, and the overall footprint is reasonable, but even urban areas, especially Delhi, have worse than average quality. A simple “solution” would be to build more cell towers. This is technologically easy, but (1) it’s expensive; and (2) zoning would remain a challenge—getting permissions to put up cell towers. Instead, why not offload much of the traffic to “femtocells”—tiny, licence-free (and sometimes individual-user) “cell towers” installed by end consumers?

Contrary to popular belief, we suffer poor connectivity not because existing towers are too far away (their reach is many kilometres)—but usually because they are overloaded. In a (say) kilometre radius, one may have hundreds of people whose signal is strong enough to connect, but the spectrum cannot handle that many simultaneous users owing to congestion.

The rising requirement for data (especially for video) needs far greater bandwidth than for voice calls. The only solution, short of enhancing our spectrum allocations or significant technology upgrades, is to make smaller but many more cells. But this would lead to economic and zoning problems.

Introducing femtocells and other disruptions: Unlike traditional cell towers, femtocells, are tiny. Being ultra-low power (hence the tiny cell sizes), these can be owned and installed by the end consumer, like a home Wi-Fi router. The analogy to a home router goes beyond size or rough cost as femtocells use the end user’s broadband or internet connection to back-haul traffic to the rest of the system. If you have a broadband connection, you could plug the ethernet cable into the femtocell and automatically use this for your cellphone instead of the overloaded neighbourhood cell tower.

The telecom company could offload traffic to the femtocell and the consumer would get better connectivity, especially in overloaded or hard-to-reach areas: a win-win situation. Consumers might pay for the femtocell, but, in return, they could save minutes on their mobile plan, or get other rebates from the telecom carrier. The femtocell could be configured for one user as well as a restricted set of users.

Another innovation we should embrace is voice-over Wi-Fi (VoWiFi).

In the same way that data traffic can use broadband instead of 3G/4G cellular, voice calls could directly use any available Wi-Fi signal with similar back-haul over broadband. Many advanced phones have VoWiFi built in (or it’s a software update away), but the carriers need to enable such features—none in India do, while all four major US carriers do.

This is superior to WhatsApp or Skype calls, which use third-party apps. VoWiFi is direct and seamless and uses regular phone numbers, just like iMessage in an iPhone uses data for messaging parallelly with carrier texts/SMSes, based on what is available.

Why we don’t have femtocells: Buying the required hardware is only half the challenge. Both femtocells and VoWiFi need carrier coordination and configuration to work, else your phone won’t know which solution to use. Carriers must enable such change, and the good news is these need not reduce revenue since they can bundle this with their calling plans. Femtocells also require regulatory approval so that they are licence- and restriction-free (from an end user’s perspective).

The actual power level is so low that it’s comparable to home Wi-Fi routers, or even less, and shouldn’t need zoning and city/municipality permissions. Thus, they are consumer-safe and should be allowed.

Weaker signals pose a bigger health risk than stronger ones. As long as a cell tower isn’t illegally transmitting signals (radiation) that are too high, and you’re not too near one, a modestly nearby cell tower results in lower radiation than a distant one. This is counter-intuitive. But we forget about the other end of the pair—our cellphone. Trying to link to a weaker signal tower means the phone has to use more energy. A strong enough signal is actually not just safer, but also saves battery life.

There are a few minor policy or regulatory tweaks required. First, broadband providers should not deem such usage as violating their terms of service. Second, telecom rules should allow such interfacing of telephony with IP (internet protocol)—at present,voice-over IP (VoIP) is disallowed in India.

Banning VoIP was a commercial choice, given that voice companies paid hefty licences for their services; we should separately revisit those rules. But femtocells deserve an exception since they’re reusing already licensed spectrum, as should VoWiFi, because it doesn’t bypass the licence-holding telephony carriers.

Do we have enough broadband connections for consumer-based femtocells or VoWiFi?

In urban areas, we might, and if enough people don’t put up femtocells, maybe cell companies could give away femtocells in congested areas (sometimes with bundled broadband). They could also encourage more shared Wi-Fi hotspots, further offloading data and even voice traffic.

Many solutions exist, we just need to be innovative in terms not just of technologies, but also business and regulatory models.

Rahul Tongia is a fellow at Brookings India, and was previously vice-chair of the UN ICT task force working group on low-cost connectivity. All views are personal.

Comments are welcome at theirview@livemint.com