The latest traveller by this road is Tim Cook, who, liberally scattering praise for India, hinted at iPhone manufacturing in the country.
These days most global roads to economic growth possibilities lead invariably to India. The latest traveller by this road is Tim Cook, who, liberally scattering praise for India, hinted at iPhone manufacturing in the country. Now we hear that Satya Nadella of Microsoft is coming to Delhi on May 30, where he is expected to push Microsoft’s White-Fi technology, using which he is ostensibly lobbying for free allotment of precious spectrum in the TV UHF bands 4 or 5 (470-590 and 585-806 MHz, respectively) for provision of broadband internet services in rural and remote areas. On the face of it, this is a noble and laudable mission, and Microsoft has been driving it for over three years in India. But if one undertakes a deeper technical analysis, one begins to see huge challenges and clearly worrisome developments.
In his typical frank and forthright manner, telecom minister Ravi Shankar Prasad forcefully upheld openness and transparency when, in January this year, he clearly stated that white space spectrum will only be provided through auction and no exceptions will be made. This came as a tremendous piece of reassurance to all that nothing would be supported which involves non-transparent allocation of a limited and extremely precious resource that, as per the highest court, is a public property.
In December 2015, this writer had highlighted the enormous potential of the so-called ‘white space spectrum’ (sought now for the White-Fi technology) in India. In fact, it had been shown that unlike Western countries, we actually don’t have the small gaps of white space spectrum amidst heavy occupancies of terrestrial broadcasting. On the contrary, we have the tremendous advantage of availability of UHF band 4 (470-590MHz) as most of them lie unused. Since we have only one terrestrial broadcaster—Doordarshan—IIT Bombay experts had pointed out as long ago as in 2013 that, on a very conservative basis, at least 70% of the band—viz. at least 84MHz—can be freed by reassignment of TV channel frequencies in India. This is 84×22=1848MHz, which is more than 90% of the total spectrum across six bands being put up for auction in September this year! If auctioned along with the lower UHF band, the face of Indian mobile communications and broadband could be dramatically transformed for the benefit of the consumers, the unconnected citizens and the economy, as well as increased inflow to the exchequer.
However, it is somewhat worrying that, despite the above facts and the minister’s unambiguous position in the matter, the dream of free spectrum for commercial use still appears to be relentlessly pursued by certain business interests. The first sign is the discard of the earlier-employed term, white space—which, anyway, has no relevance for India—and the use of the term White-Fi, positioning it as akin to Wi-Fi, which is generally accepted the world over as unlicensed free spectrum in the much higher frequency bands of 2400MHz and also 5700MHz. However, the basic concept in the White-Fi technology still remains that of utilisation of the gaps between usage frequencies of various terrestrial broadcast transmitters, deploying cognitive radio technology. This, therefore, continues to be inapplicable to India.
Moreover, the concept of a Wi-Fi operation in such low frequencies is itself very harmful for us. The UHF frequencies have excellent propagation characteristics, and even at low transmission powers, can travel for scores of kilometres. This, though much desired for a regular mobile application, would be deadly for a Wi-Fi operation, which has to be circumscribed within a relatively small area to avoid massive interference problems to other licensed users.
Unlicensed spectrum does not guarantee exclusivity and, therefore, there is no coordination between operators/users using this spectrum. Hence, managing interference from users of other operator’s network also assumes paramount importance. Usually, the operators are mandated to transmit at much lower power as compared to their licensed counterparts (4 watts ERP versus 20 watts ERP), so that multiple networks can coexist in the same frequency band very close to each other. However, uncoordinated use makes it impossible to eliminate the possibility of interference and, hence, the technologies using unlicensed spectrum use frequency hopping techniques and dynamically choose a block of spectrum (among multiple blocks) with least interference. Therefore, in order to enable reasonable data rates, an unlicensed operation needs a much larger spectrum as compared to its licensed counterpart (typically a block of 100/200MHz), of which only a fraction (20/40MHz) is used at a time. This is the reason why lower frequencies (less than 1GHz) are not conducive for unlicensed operations as these are highly spectrum-inefficient.
Importantly, it may be noted that the UHF spectrum is tremendously more powerful in propagation than even the coveted 900MHz spectrum. Thus, it is extremely valuable in urban and semi-urban areas and also for ubiquitous mobile broadband, including in-building coverage.
While White-Fi technology is being used for extending broadband service to rural areas, to extend 100Mbps of BharatNet/NOFN bandwidth at gram panchayats/optical fibre nodes, it will need at least 100MHz (at the rate of 1Mbps/MHz) of continuous spectrum—not counting spectrum required for redundancy and managing interference. This spectral efficiency is based on the results of factual trials carried out for TV white space technologies in India under controlled conditions in the test-bed and whose results are published. In real-life variations in place of controlled conditions, the spectral efficiencies could be even worse.
The trials done so far using TV white space have used it for fibre extension from the gram panchayat and the last-mile is connected using technology. So, to call it an access technology to connect 5 lakh villages is a misnomer. In fact, it makes TV white space technology a middle-mile technology for backhauling traffic from Wi-Fi clusters and not an access technology as is being projected. Given that trials have shown low spectral efficiency, if it is supported, eyebrows are bound to be raised since far more spectrum-efficient, cost-effective and proven middle-mile technologies exist which can provide at least a gigabit of guaranteed bandwidth with no interference whatsoever.
If any entity wishes to evaluate White-Fi technology in Wi-Fi application, it should be advised to do so in the regular Wi-Fi bands of 2.4 and 5.7GHz. In addition, other such high frequencies may be considered, but surely not an inappropriate UHF and other sub-1GHz frequencies.
The government is targeting a mega spectrum auction in September. To give away the entire or part of the more powerful regular spectrum to unlicensed operations could quite possibly lead to negative sentiment amongst prospective bidders, causing prices in the upcoming auctions to plunge southwards and cause a huge loss to the exchequer. This is akin to what happened in 2001, when, in the fourth cellular auctions, prices nosedived to R1,650 crore all-India, due to the prospect of free 800MHz CDMA spectrum which was a ground reality before the auctions.
Clearly, Digital India can ill-afford to fritter away precious regular UHF spectrum for incorrect and inefficient Wi-Fi usage.
The author is honorary fellow of the IET (London) and president of the Broadband India Forum.Views are personal