will India leapfrog to 4G directly?

All the recent news in the aftermath of cancellation of 122 2G licenses, India's wireless telecom segment looks a bit uncertain. At least two foreign operators publicly stated that they are exiting India. However if one looks carefully, there are more positives than the negatives. The verdict makes every stakeholder more careful and focused on the long term. Also DoT's Telecom spectrum handling is expected to be more transparent and consistent across. The fact that India's mobile population is growing and has much younger user-base, makes it more attractive to the operators compared to other markets. Also it appears that India's mobile users are getting more interested to use their mobile phone for internet access. That will eventually make mobile data traffic way higher that voice traffic which opens a multiple avenues for new revenue generation compared to plain voice traffic.
  As per the September, 2011 data from TRAI, India's mobile subscription base stands at 907 million and most of them are on the voice network. A large chunk of this userbase is expected to upgrade themselves to smartphones and are expected to generate more data traffic than voice.
With 2G network still growing in India and are expected to continue the trend for another year or so, India's challenge is to allow the 2G operator continue generating revenue while preparing the network for high speed wireless broadband.
In 2010, India had the first auction of Broadband Wireless Access spectrum. The 4G Auction carried on the back of 3G Auction saw 5 winners: Infotel (22), Qualcomm (4), Bharti (4), Tikona (5), Aircel (8) and Augere (1) excluding state-owned BSNL and MTNL. It is interesting to note that top 3 winners are new operators. More interesting aspect is that almost all of them signed up for LTE technology. Now LTE is the most dominant 4G technology with an adoption slowly encompassing the whole world, as per the latest report from GSA (Global Mobile Supplier Association).
For record, GSA report from January, 2012 tells that 285 operators globally are investing in LTE. And DailyWireless reported last year that all the 4G auction winners [barring state-owned operators] have chosen LTE-TDD as the technology.
It is particularly important to mention that China too has chosen the LTE-TDD as the technology for their 4G network. With World's two most populous nation picking LTE-TDD, it is quite apparent that LTE-TDD will enjoy the dominance across world.
What is LTE-TDD
LTE-TDD or Long Term Evolution - Time Division Duplexing [for unpaired bands] was standardized in 3GPP Release 8 as the fourth generation cellular wireless technology. While the back-end is all-IP nework [unlike 3G], LTE comes in two different Radio access variants: Time Division and Frequency Division. TDD variant is asymmetric [between uplink and downlink speed] and therefore promises higher spectral efficiency [i.e. more bits per frequency channel]. Since it uses single carrier frequency, it is also relatively simpler to deploy. Recent trials from Ericsson, Huawei and others claim to reach downlink data rate of 100 MBPS, making it ideal for IP-TV type of broadcast service. With dynamic asymmetric configuration the downlink rate can be  4-9 times faster than uplink rate making it adaptable for different types of data services.
The architecture is made such that an operator can have both TD and FD network helping immensely to retain CAPEX investment made in either of the technologies. 3GPP identified 30+ frequency bands so that spectrum availability does not become a constraint. It can interwork with 3G technologies which means migration from 3GPP is smooth. Since LTE can be deployed over older GSM/GPRS bands, it can also become the upgrade technology for 2G. However that will require complete overhauling of network infrastructure.
There is one chink though, 3GPP did not define the voice technology for LTE. To address the gap, GSMA adopted the GSMA voice over LTE initiative that tries to define VoIP service over 3GPP standardized IMS [IP Multimedia Service] although that service is not part of LTE standard.

Salient points about LTE-TDD

• Technical Standards part of 3GPP release 8
• High theoretical peak user rates of 170 / 54 Mbps (downlink/uplink, 20 MHz)
• Asymmetric uplink/downlink rates to match traffic types for good efficiency
• Peak spectral efficiency as good as competitive standards e.g. WiMAX
• Network infrastructure compatible with LTE-FDD
• Voice over LTE standard is being developed by GSMA
• emerging as the most dominant 4G technology as per the operator adoption is concerned

Migration Path

Migration path from 2G to 4G can be through 3G or bypassing 3G depending on how much the operator is invested in 3G and how many subscribers have adopted 3G. Given that India is lagging behind other countries in 3G adoption, it probably could leapfrog to 4G directly since in India 4G band is allocated far outside the GSM/GPRS band. It is interesting to note that both China and India have allotted the 2.3 GHz band [3GPP band 40] for LTE-TDD service. It would not only save the operator CapEx of 3G infrastructure, it also will be operationally simpler with 4G deployement requiring new model of handset.

Incidentally China Telecom is the first large provider to commit itself to LTE-TDD. In India TRAI is recommending LTE-TDD for femtocells. Femtocells are very small cells, with a diameter of few meters, recommended for wireless broadband access in urban areas.
It's still far by at least one and half years
While the commitment to the technology is quite strong and trials in different parts of the world are very promising, it also should be noted that commercial availability of this technology [both network and handset] is still one and half years farther making it all the more attractive choice for leapfrogging especially for those operators who have a very small 3G footprint. In the meantime, if you are buying a smartphone try not to spend a lot of money since you may need to replace it soon!
Further Updates
There was one aspect that I intentionally did not touch upon. India's Telecom Policy 2011 draft [to be published this April] did not mention about DoT approach on 4G. As if to answer that yesterday Mr. Kapil sibal announced that further auction for 4G is to happen this year. It is to be noted that BSNL has surrendered their 4G spectrum [that was allotted to them in 2010] to Govt. couple of months back. Govt. also can refarm unallotted 2G and 3G spectrum for 4G which means that there are enough spectrum available to launch 4G service.
Update from March, 6, 2012:
Govt. decided to refarm 700 MHz spectrum for 4G service, pooling up more spectrum for 4G auction.[ Report: http://www.thehindu.com/todays-paper/tp-business/article2965040.ece]

From independent standpoint, right approach should be that Govt allot unified licenses to the operators and auction spectrum without association specific generation of technologies and let the operators decide which band they want to use for 2G or for 4G. To protect 2G investment of new operators, Govt is very likely to continue bar roaming in 3G and 4G for next couple years but eventually all technologies will merge to 4G with unrestricted roaming across circles and technologies [i.e. 2G/3G/4G]. That will take some time but handset manufacturers also need some time to build truly multimode [2G,3G,4G] handset. In summary, India's cellular wireless segment is going to be exciting for at least next four years.