18 Dec

TRAI’s big budget plan for National Broadband – my analysis

Recently TRAI i.e Telecom Regulatory Authority of India came with National Broadband Plan for India. I tried looking into it, and it’s just not making much sense at all.

The Plan…

TRAI’s plan is to put massive infrastructure – basically over 12 billion Km’s of Fiber Optic cable connecting whole India. Plan covers 63 large cities, 4315 small cities & around 3,75,552 villages which have a population of over 500. Intention is to have a heavy backbone across country, which will bring more competition in Indian broadband industry.

Do we really need it?

Simple answer – NO!

While making this recommendation, it seems like TRAI has almost forgotten that Govt. owns a State operator named BSNL which presently owns the biggest domestic backbone of India, with fibers connecting around 400, 000 of villages. In recommendation, they also mentioned that this building a new fiber network will help in having a scalable backbone, and aims to provide speeds of 10Mbps to users in cities while at least 2Mbps to users in villages. This again makes no sense at all!

In reality, the fact is – over 70-80% of existing fiber pairs of BSNL (call it Govt.) are yet not lit, and those dark fibres have a tremendous bandwidth. In fact if we add the total cables by all private players in India, it reaches hardly 1million route km of fiber cables, while BSNL itself owns over half of million route Km of fiber cables. With that, we can clearly say that we have enough bandwidth on domestic backbone of country. It is NOT in hold by private players, but is not in use to due super poor management work of BSNL, Political factors and lobbying by Private carriers. The latest part is decision by BSNL to auction off the capacity on those line!

One interesting point is in most of links, BSNL is hardly using 10-20% of capacity of fibre, still BSNL MPLS seems to be collapsing. Reason remains political and other non-technical stuff. Apart from that, this is just about the (domestic) backbone capacity. We need to see the other 3 parts also, which are real bottleneck.

Firstly, last mile in India remains an issue. Over 80% of copper is still in hands of BSNL, and quality is poor. The hosting problem in India continues with over ~80% sites not hosted in India. Which is causing super high usage of International bandwidth which is mostly responsible for super poor capping in broadband connections. The doT seems performing even more poor in peering policies.Govt. actually started with niXi i.e National Internet Exchange of India to promote domestic peering. It was a very good initiative, but seems like it didn’t worked due to poor tariff policy.

My recommendations to fix broadband problems with *existing* technologies:

  1. Take BSNL seriously. DO NOT FORGET it’s govt. company and thus Govt. owns biggest telco infrastructure in country. All actions regarding build of infrastructure SHOULD be taken ONLY via BSNL.
    It makes no sense at all to bring more bodies/companies/departments into this. Seems funny? Seriously, that’s real!
  2. Be serious for last mile connectivity. Put more pressure on BSNL to fix the last mile. Even a small budget can fix most of existing copper infrastructure which can be used for ~20Mbps broadband connectivity via DSL.
  3. Based on Mr Sam Pitroda’s recommendation – unbundle the last mile copper pairs of country and make those available to private players who will help in removing the monopoly of State owned telco.
  4. Stop jokes about spectrum and 3G. The 5Mhz spectrum given in last auction was more or less just a joke. Based on low ARPU’s private players are all looking into wireless connectivity. 3G can help in improving broadband penetration, while technologies like WiMax can really make BIG difference. Be serious, and give more spectrum to improve last mile link.
  5. Fix the peering policies of niXi. Make sure charging is based on port speed rather the amount of data transfered. The per GB cost via niXi still seems very high.
  6. Put a good amount of money on building big data centers in India. Probably 4 – North, South, East & West, and have offer very cheap web hosting services including collocation, shared hosting, dedicated servers, vps etc via BSNL. In fact BSNL does already offers that, but offering is all outdated, and they don’t have good data centers yet.
    Apart from that, Govt. should give high incentives to private companies coming in data center business to bring more and more content in India, which will drastically reduce costs. Remember, International bandwidth costs ~50times more then domestic bandwidth and fact remains that over 80% of sites an average Indian visits uses that expensive bandwidth!

Govt. should ask this question itself  – Is broadband a one of basic needs of Indian citizens, if yes, then put $$$ in fixing it in good way, if not then keep on making good $$$ from crazy auctions and poor policies.

With hope that next time you will be on better connection, thanks for your visit.