Nothing broad about it!

If wishes were horses, India’s broadband plan would have reached somewhere. What we see instead is that policy makers, telcos and ISPs are at loggerheads while businesses are still waiting for low-cost broadband connectivity, says SHIPRA ARORA

DR SETH states that unless the incumbent is willing, the broadband programme through DSL cannot succeed

INDIA’S bandwidth story is ironic. On one hand the country has seen flat Internet growth—which has at times gone into reverse—over the last three years, and broadband penetration of only 0.02 connections per 100 persons (as per the Telecom Regulatory Authority of India—TRAI). On the other hand there’s a bandwidth glut. Despite the surplus bandwidth availabile, most people are still stuck with narrowband connections and low access speeds.

Both CII (Confederation of Indian Industry) and TRAI try to answer this question in their respective reports on the Indian broadband sector (CII’s ‘India’s Broadband Economy: Vision 2010’, and TRAI’s ‘Broadband India: Recommendations on Accelerating Growth of Internet and Broadband Penetration’, which is based on CII’s report). The last mile issue finally

gets due consideration now. According to Sabyasachi Patra, consultant, CII, who was speaking from the broadband service provider’s perspective, lack of access to the customer in the last mile prohibits them from reaching out. Dr Devendra Seth, member, TRAI, however insists that this is just one of the core issues and that it has to be addressed in unison with other factors.

As broadband falters in the final lap of its journey—the last mile—recommendations made by the two authorities on unbundling the local loop together with those for promoting DTH (direct to home), VSATs (very small aperture terminals) and wireless for enabling last mile access, may well hold the answer to India’s ailing broadband market.

Under-utilised bandwidth

According to TRAI, the present capacity of international bandwidth is around 0.340 Tbps (including Tata, Bharti and Reliance), but utilisation is less than 0.010 Tbps. The approximate designed future capacity is estimated to be 16.5 Tbps. Service providers (including BSNL, MTNL, Reliance, Bharti, Tata and the various utilities) have laid around 670,000 route kilometres nationwide. The designed capacity supports an average of 1.2 Tbps on each route kilometre, considering an average 12 pair DWDM. The utilisation is, however, less than 0.003 Tbps on an average.

TRAI says that sufficient bandwidth will exist in the country in the near future. According to Dr Govind, director, department of information technology, Government of India, in the current situation the big pipe exists to carry bandwidth but the smaller pipes to carry it to end users are lacking.

What went wrong at the last mile?

Today’s half-baked DSL and cable solutions are not ubiquitous, says Shivaji Chatterjee, senior director, Sales and Marketing, Hughes Escorts Communications. Marred by monopolistic tendencies, lack of will, technology constraints and regulatory hurdles, the last mile still remains a bottleneck. The fact is that creating bandwidth isn’t enough; effective means for delivering it to users need to be put in place.

According to AMITABH SINGHAL, it is easy to surmise that our policy makers are reluctant to allow unbridled progress by removing all bottlenecks at one go

David Appasamy, chief communications officer, Sify, says that the limited last mile options available to new operators have led to a virtual monopoly for the incumbents. Lack of access to copper in the local loop to reach customers via DSL, and the high costs of duplicating this infrastructure, have been recognised by both CII and TRAI as the biggest roadblocks to last mile access. Most of the existing copper network (almost 95 percent) is controlled by incumbents BSNL and MTNL.

Yet another hurdle is that telcos have always been reluctant to work with other ISPs. A case in point is the effort by ISPs, over the last few years, to convince incumbent telcos to offer co-location to ISPs offering broadband. “They have never agreed to share the last mile to let other ISPs introduce DSL services,” Amitabh Singhal, secretary, Internet Service Providers Association of India, voices the ISP’s grievance. As a consequence, independent ISPs get busy stemming red ink rather than increasing their reach.

The price that incumbents charge from ISPs for wired last mile links (such as E1 lines) is a big barrier. “These tactics have left ISPs with little incentive to develop broadband services. So when it comes to DSL, incumbents and telcos are the only ones to effectively tap this potential,” adds Singhal. Furthermore, it is not just who owns the copper but also the quality of the copper that is in question. According to Patra, copper available with incumbents that is capable of handling DSL services ranges from 50-80 percent.

Beyond copper

A few bottlenecks have been addressed, such as delicencing the 2.4 GHz band and letting 802.11b be used on campus.

The cost of using DTH and VSAT has been high, along with restrictions preventing them from being used to deliver broadband Internet services. Existing cable networks are not broadband-ready primarily due to the quality of cable deployed; also, the industry is marred by lack of organisation.

TRAI says, unbundling is the answer

JASJEET SAWHNEY points out that Britain has more than 103 ISPs offering ADSL-based broadband services using local loop copper owned by BT, NTL and Telewest

TRAI has recommended that the use of existing infrastructure on an incumbent’s copper to reach customers via DSL should be allowed, and that artificial costs in the operation of DTH and VSAT platforms should be removed.

It has also recommended that the local loop be unbundled. With around 40 million telephone lines (CII estimate), the country is sitting on huge untapped potential for delivering DSL broadband.

Three options have been suggested for enabling Local Loop Unbundling (LLU)— Full Local Loop Unbundling (full access), Shared Unbundling (line sharing) and Bit Stream Access. In Full Local Loop Unbundling competitors have access to both voice and data on the line. Under Shared Unbundling the competitive providers have access to either voice or data transmission on the line. In the case of Bit Stream Access, the incumbent installs high-speed access links to its customers, and allows competitive providers access to this link; the incumbent owns and maintains the said link.

For Singhal, shared unbundling would be first on the list of priorities of ISPs, followed by Bit Stream Access. This is simply because in shared unbundling the ISPs pay for co-location, put up the equipment, and are then free to choose the services they want to provide. At the same time, the telco continues with its telephone services.

Are incumbents ready?

While LLU has many points in its favour, its success depends on BSNL and MTNL, which are reportedly in discussions with DoT after the recommendations were floated by TRAI. According to Dr Seth, unless the incumbent is willing, this broadband programme through DSL cannot succeed. Seth justifies the adoption of LLU for incumbents: “It is common sense that they must push for it because there is a lot of money in it for them. There are enough revenues for the incumbents to go in for broadband, and enough revenues if they don’t do it themselves and franchise it to others.”

Jasjit Sawhney, CEO, Net4India, considers the recommended solution feasible, pointing out the case of Britain, which has over 103 ISPs offering ADSL-based broadband services using local loop copper owned by BT, NTL and Telewest. But Chatterjee does not see any significant results coming out of LLU. “Last mile quality and downtime is an issue for both BSNL and MTNL. Sharing the copper will not improve the service as it will be subject to the same outages.”

“It’s been over eight weeks and there has been no indication of any part of the recommendations being accepted,” muses Singhal. The TRAI report was prophetic: “Our consultations show that, except the incumbents BSNL and MTNL, all stakeholders agree that non-discriminatory local loop access is required.”

Broadband twins

DSL and cable are the two key technologies for broadband. According to Govind, “The rate of adoption of various technologies will depend on factors such as revenue models, cost, affordability, applications and content. To begin with, DSL will be more popular.” In the immediate future, it makes more sense to tap the existing copper network to reach 40 million homes. Dr Seth thinks that the immediate results should come out of DSL and cable.

There are over 22 million homes or SME connections in the country that are going to be the target for broadband services, and another six to seven million enterprises that can also be targeted for broadband connections. According to Singhal, with around 50-60 million cable TV homes, Internet over cable has the potential to be a huge market. “The market here is more price-sensitive than technology-biased. The largest group of network service providers in this country are cable operators,” says Sawhney. There are more homes with cable connections than with fixed-phone lines despite the fact that the cost of the access device in the case of cable is 10x that of a phone line; this makes it worth considering cable as a means of delivering broadband.

Loose ends

Sify feels that although TRAI had indeed recommended that ISPs should be free to choose the last mile (such as copper wire), the licence amendment permits use of underground copper only as an added option, besides those already in the licence. Net4India says that along with access to local loop, co-location of DSLAM and other equipment at the premises of the BSO need not be made mandatory, and the cost of co-location should not be predatory.

ISPAI says that TRAI’s recommendation for unbundling the last mile is a positive development. However, it was expected to go whole hog and lay down the terms and conditions under which ISPs could seek that last mile. Instead, it’s been left to the incumbents to decide. Worse, TRAI has recommended that only those links that are more than five years old be unbundled. Determining which links are more than five years old is going be another pain point.

In conclusion, it could be made out that the road to true broadband is still a long one, and India Inc is struggling with issues such as unbundling of the local loop. It looks like we have to work on providing mass access and then work on broadband penetration.

Meanwhile, the ISP-Telco tug-of-war continues.

shipra@expresscomputeronline.com