A Deal That Rings False

IT'S touted as the latest technology in basic telephony. And, if the Department of Telecommunications' (DOT) plans work out, the wireless in local loop (WLL) will be introduced to India, or at least to the citizens of Delhi, by the end of the month.

There are many advantages to the technology. WLL uses a radio signal instead of a copper wire to connect the exchange to the subscriber's premises so that phone connections can be provided more easily, often in a matter of hours. More importantly, WLL does away with the need to dig up roads to lay cables which are often faulty, prevents cable thefts and provides wider coverage. In addition, it's a value-added service that would bring additional revenues into the DOT's coffers. The technology is being provided by San Diego, US-based Qualcomm Inc. The size of the deal, which envisages this service in 18 cities: Rs 400 crore.

Sounds great. But.

Five DOT officials signed the note recommending the procurement of three WLL systems from Qualcomm in 1995. One is the (in)famous Runu Ghosh, who was suspended as DOT's director (finance) in the wake of the Sukh Ram scandal last year. Another signatory was the then senior deputy director-general of DOT's Telecommunication Engineering Centre (TEC). He has since retired and is a consultant. To Qualcomm. Yet another is the deputy director-general (north), TEC. His son, a fresh engineering graduate, is employed at a salary of Rs 18,000 a month. At Qualcomm's Delhi office. His house, TU-22 Pitampura, New Delhi, is rented out to a company at a fat rent. The firm is Fib-com, a licensee for Qualcomm's code division multiple access technology (CDMA).

What happy coincidences! In fact, the Qualcomm WLL story is almost a fairy tale, with good fortune apparently favouring the company at every step.

It all began in 1995, in the heyday of Pandit Sukh Ram. DOT decides that to keep ahead of impending competition from private basic service providers, it needs to upgrade its network. It zeroes in on WLL as the technology in wireless fixed telephony. On September 8, 1995, DOT places the order for three systems each with 1,000 subscribers for New Delhi, Calcutta and Madurai with Qualcomm. No competitive bids are invited though there are other firms willing to provide the same technology. In March 1995, Qualcomm had given an undertaking that it would offer its system for technical testing by September '95. But since the DOT decision comes so late, the deadline cannot be met.

Mahanagar Telephone Nigam Ltd (MTNL) pays for the system at the rate of $899 per subscriber line, paying $899,000 (Rs 3.24 crore) for the 1,000 subscribers in New Delhi. The system is installed at Bhikaji Cama Place for MTNL. The TEC tests the equipment in April-May 1996 and flunks the system, saying it "does not meet full requirements".

Qualcomm asks for an extension for system testing till September 1996. This is granted. In the meantime, the Lok Sabha elections send the Congress out of power, and Sukh Ram is no longer minister. Qualcomm manages to replace all hardware and software by December 1996. The TEC carries out tests on this 'upgraded' system from January 10 to January 30, 1997. On February 5, the TEC writes to B.A. Majumdar, vice-president, India operations, Qualcomm, complaining that "it is observed during testing that a number of parameters have not been reoffered for testing, indicating that the system does not support these features and some of the parameters, though reoffered for testing, are not meeting the requirements...the features committed to by you are not implemented."

 So what happens? Qualcomm gets another extension. And for a third time, it fails to meet the mark. The technical report dated February 21, 1997, lists 18 "deviations observed during acceptance testing". For instance, speech quality: "Speech is not of uniformly good quality. Sometimes the call gets disconnected after first ring, sometimes an echo is heard, sometimes noises are heard and sometimes speech is muffled." The report lists 16 features that are not supported by the system, ranging from abbreviated dialling, call waiting facility, alarm call, hot-line and three-way calling. These features have all been committed to by Qualcomm. And many are exactly the points on which the system failed earlier tests.

Meanwhile, having coughed up money for the system, MTNL is eager to begin the WLL commercial service by its scheduled date of May 15. It writes to TEC and gets a go-ahead to begin commercial operations, pending technical approval. On April 9, a TEC director writes to the chief general manager, telecom, MTNL, that "the equipment has been found satisfactory for loading with commercial subscribers." He stated that "the equipment will be treated on field trial from the date of 75 per cent loading of the system," and "the pending points...shall be attended to by M/S Qualcomm." This is probably the only instance that DOT has okayed the commercial launch of a new system without technical clearance.

But then, a lot of exceptions have been made in the Qualcomm deal: 

No competitive bidding: There was no competitive bidding in the 1995 purchase order, allowing Qualcomm to charge $899 per subscriber line unilaterally. In the case of all equipment purchases made in the telecom sector so far, tenders were floated and bids invited after rigorous field trial. Why then in this case were field trials permitted to only Qualcomm? More so when there were companies like Ericsson and Hughes that had applied for testing of their system free of charge! Also, if the Qualcomm system was meant for trial only, why were three systems ordered for three cities each with 1,000 subscribers? Isn't one system good enough? 

Pricing anomaly:The Qualcomm rate of $899 per subscriber line compares unfavourably with the $600-$700 rate prevailing in most markets of the world! Even if this lower rate is because cost per subscriber line goes down as the capacity utilisation goes up, the question remains as to why there was no evaluation of what other firms were prepared to offer for the same technology. Moreover, had it not been for the political turmoil that brought down the Gowda government in April, DOT would have placed the Rs 400-crore order with Qualcomm despite the failure of the tests and despite adverse financial advice by experts who had evaluated the DOT proposal to install 76,000 lines of WLL equipment for 18 cities. The cost calculations have been made on the basis of Qualcomm quotations, said the experts, whereas the "project estimates may be prepared as per rules showing capital outlay, expenditure, expected revenue and rate of return for proper examination and evaluation of the scheme. " Why this special favour to Qualcomm?

Faltering technology: Code division  multiple access (CDMA), on which the Qualcomm WLL system is based, is acknowledged by international experts as being the latest technology but has no clear installed base and commercial services. Qualcomm is known more as a design and development company than as a supplier of equipment for commercial use. Even in the US, the technology was put to commercial application only in 1996. Almost the only other country that uses CDMA-based systems is Korea where surveys have revealed a high level of customer dissatisfaction with the quality of service. According to a survey by the Korean Consumer Protection Board, 57 per cent of respondents said they experienced disconnections on an average of once or twice per call. Why then should the DOT volunteer be the next guinea pig? Especially when the technology has been given repeated chances to improve and each time has fallen short of specifications.

Repeated extensions: According to informed sources in the DOT, this week a letter is being sent to Qualcomm, giving them an extension of another 11 months to fix the technical problems within their WLL system. This gives Qualcomm a total of three-and-a-half years for a field trial! And this for a technology that was supposed to be cost-effective for the DOT in the long run.

And the subscribers: MTNL is all geared up to offer the subscribers the WLL technology in the city of Delhi. Customers will pay a registration fee of Rs 25,000. And MTNL will charge a monthly rental of Rs 1,000. The DOT plans to introduce the service in 18 cities. When the system has not even passed technical tests, isn't the DOT taking customers for a ride? A final question: why should the gullible customer be offered a system which will not work as promised after he has paid up money for the service?