India lacks the bandwidth

The growth of broadband connections in India continues to be abysmally low. The initial push was based on landlines but soon utilisable landlines got exhausted. The cable television network is not robust enough to handle high-speed data, and fibre to home continues to be an expensive option. That leaves us largely with the option of wireless-based access.

The question is why, despite smartphones constituting a large proportion of handsets being sold today, India, unlike most countries in the world, continues to perform poorly on the broadband front both in terms of number of subscribers, and speed and quality.

The answer obviously lies in how well-prepared and organised our telecom infrastructure is for provisioning and running broadband services.

The wireless-based infrastructure needed covers spectrum and device availability. We have seven possible bands in which the two major broadband access technologies, namele HSPA (3G) or LTE (4G) can be deployed. These are 900 MHz and 2100 MHz bands for HSPA (3G) and 700 MHz, 800 MHz, 1800 MHz, 2300 MHz and 2500 MHz bands for LTE (4G).

The present situation is that there is no pan-India 3G operator and the maximum 3G spectrum any operator has is 5 MHz. In the case of LTE, there is only one operator with pan-India allocation of 20 MHz spectrum. In short, at present there is hardly any true competition in 3G services and when LTE (4G) service is launched by the single pan-India spectrum holding operator, there still will not be any real competition.

Our 2G experience shows growth of services is helped by competition through reduced tariffs. However, growing broadband is more complex than growing voice telephony. Increased network usage driven by cheaper voice does cause service deterioration, but seems acceptable to users. Therefore, limited overloading due to spectrum paucity in voice could be got away with. On the other hand, for broadband services, while price is very important, speed is also important.

As more data users are added, the network starts saturating and speed comes down rapidly and soon users become disillusioned and do not wish to use wireless broadband while continuing with voice services. In short, for a given available spectrum, the flexibility available with an operator to increase customer count is far more limited in broadband compared to voice telephony.

It is, therefore, evident that for wireless to act as access technology for broadband growth, adequate spectrum has to be made available. Another fact sometimes lost sight of is that the advanced capabilities of LTE (4G) such as higher speeds, can only be effective if adequate bandwidth (of the order of 20 MHz) is made available.

Experts believe it will be another four to five years before the eco-system matures to enable affordable devices to suit Indian pockets. In short, 700 MHz band eco-system is not yet mature enough to contribute to the growth of broadband in India over the next few years.

In the 800 MHz band, the spectrum available is much less than in 700 MHz and the eco-system much the same as in 700 MHz band with deployment in two small countries so far.

In the 1800 MHz band, the LTE eco-system is quite mature. However, with the extensive use of this band for 2G services in India, the band is so strongly fractured that very few contiguous slots of 5 MHz (minimum required and much less than optimum for LTE) are available.

The February 2014 auction of this band resulted in all contiguous frequency slots being taken up. However, the operator’s ability to deploy LTE will depend on the outcome of auction of 900 MHz band (reportedly slated for the end of 2014).

In the 2300 MHz band, auction and allocation of spectrum has taken place but due to the still developing eco-system for LTE in this band, hardly any service launch has happened in the four years since the auction.

The 2500 MHz band is much more attractive, but it is heavily in use by other services, leaving little for commercial deployment.

Of the other two remaining bands currently, the 900 MHz band is being used for 2G services by a few existing operators whose licences come up for renewal from 2015 onwards till around 2020.

It is understood that DoT plans to auction spectrum in this band in anticipation of the slots becoming available for licences which require renewal in 2015 and 2016. A detailed analysis of such licences shows that two slots of 2x5 MHz will be available in only eight circles, and one slot in the remaining circles.

The eco-system for HSPA (3G) is excellent in this band. If the spectrum is won by a new player (not having a service in this band), he may launch 3G service which is more efficient for even voice services than 2G. If on the other hand an existing player having installed 2G equipment working in this band wins spectrum, he may or may not launch 3G services unless he already has enough spectrum in 1800 MHz band to shift his 2G services.

The remaining band, namely the 2100 MHz band, is well harmonised internationally and therefore has an excellent eco-system for HSPA (3G) services. No spectrum is available due to the suboptimal sharing arrangement between defence and DoT, with DoT blocking spectrum in 1900 MHz band for which there is no demand.

Industry had suggested a swap (between 1900 and 2100 MHz) to generate three more slots of 5 MHz each. The armed forces whose clearance is necessary, have so far not agreed to this.

If this proposal does not work out, it will leave us with only the option of partial allocation on the lines adopted by DoT in the recent 1800 MHz spectrum auction.

It is, therefore, evident that to ensure the rapid growth of broadband in the next few years, the only clear possibility is creating more spectrum space in the 2100 MHz band. In the longer time frame, pursuing 700 MHz band will help.

The writer is a former member of the Telecom Regulatory Authority of India

COMMENT NOW