Here's a simple mathematical problem. If India now has 25 million phones and wants to take this to 100 million in 10 years at a cost of Rs 30,000 per line, how much will it cost? Answer: Over $52 billion at current rates. But there's a man in Chennai who insists he can do it for half that, or even one-third. No, this is no ordinary man. In fact, for some students at the Indian Institute of Technology (IIT) in Chennai, he's something of a demi-god.He's Ashok Jhunjhunwala, head of the electrical engineering department at the IIT. And he believes that a 100 million telephone lines (he recently upped the bar to 150 million lines) is one of India's most urgent requirements today. And what makes him think he can do it cheaper? CorDECT-based wireless in local loop (WLL), that's what. It is a technology jointly developed by the IIT and Midas Communications, a Chennai-based company incubated at the institute. The principle is simple -- eliminate costly last-mile phone connections made of copper and replace them with wireless.Jhunjhunwala says, "If you assume 15% for operational costs and another 15% to service the capital, each phone line will have to generate Rs 9,000 per year or Rs 750 per month for the telephone service to remain viable. But less than 3% of India's population can afford this. The solution therefore is to reduce the cost to Rs 10,000 per line. If this can be done the phone line can break even with a return of just Rs 250 per month or Rs 3,000 per year. This kind of costing would make the telephone affordable to 250 to 300 million people.'' Jhunjhunwala and his team have already managed to halve the cost to about Rs 15,000 (from Rs 30,000) using corDECT WLL. Now they're trying to cut it down to Rs 10,000. That's a staggering amount of money saved over ten years, without forgetting for a minute that all these are notional savings -- real only if licence holders and the Department of Telecommunications (DoT) go ahead and set up these lines. But that's the government's aim -- a total of 100 million phones in ten years.The professor has been putting the money where his mouth is. Some 120,000 lines are currently under various stages of implementation all over the world at a combined actual saving of $42 million (Rs 180 crore) over conventional technologies. "The target is to have at least one lakh corDECT WLL-based lines installed and running before the end of this year (in India),'' says Jhunjhunwala, adding, "From next year onwards, we are confident of seeing at least (another) 75,000 lines every quarter.'' That's 3 lakh lines a year and a saving of Rs 450 crore, but way behind the ten-year target. Jhunjhunwala though hopes the pace of deployment will accelerate down the line.Mahendra Nahata is one of the believers. The chairman of Himachal Futuristic Communications (HFCL) is implementing the solution in Punjab. "Jhunjhunwala's is a very realistic target if the cost of telecom equipment is brought down using indigenous technologies like IIT's WLL and fibre in local loop (FLL),'' he insists. Nahata should know. During the last 12 months or so, HFCL has commissioned over 4,000 lines using Jhunjhunwala's corDECT WLL solution in Delhi, Chandigarh and Ranipet, apart from Tunisia. Nahata also reveals that thanks to IIT's breakthrough technology, his cost per line has been as low as Rs 12,000 to Rs 13,000. Nahata has already decided that about 50,000 lines that HFCL is to install in Punjab (with Essar Commvision) in the next 12 months will all be based on corDECT WLL. Ask Jhunjhunwala for a progress report on the implementation of the technology he helped create and he will reel off the numbers. "DoT is implementing a 24-city, 25,000-line project with WLL and 15,000-line projects in Chennai and Bangalore. Shyam Telecom is planning 30,000 lines in Rajasthan.'' Both state-owned outfits, Mahanagar Telephone Nigam Ltd (MTNL) and the Department of Telecom Services (DTS, recently spun off from the DoT) have plans to deploy WLL on a large scale. But IIT-Madras and Midas will have to fight it out with other vendors which also offer similar WLL solutions. In any case, a start was made when MTNL chairman S. Rajagopalan inaugurated about 1,000 lines last fortnight at Mumbai's Marol exchange based on corDECT WLL. But not all at MTNL are quite convinced. Says a top-ranking MTNL official, "There are both advantages and disadvantages with IIT's technology. The advantage is that this is a locally developed solution and supports high traffic -- both voice and data. The disadvantage is that it does not eliminate copper fully. Only the distribution is wireless, while the primary cabling from the exchange is still copper.'' CorDECT WLL faced a lot of obstacles early on -- competition from global telecom majors, unfavourable duty structures, lack of vendor financing, bureaucratic delays etc. So by the time the technology was accepted in India, some 15,000 lines had already been commercially deployed in countries like Madagascar, Fiji, Tunisia, Nigeria, Yemen, Argentina and Brazil. It's been a struggle. Many years ago, when corDECT was still on the drawing board, Jhunjhunwala made a series of presentations to the private sector in an attempt to raise funds. One such pitch was made to the chairman of a telecom equipment maker. His response: "You have sold me a dream, not a technology.'' Jhunjhunwala recalls how the team and technology fell into place. An engineering graduate from IIT, Kanpur (class of 1975), Jhunjhunwala went on to an MS and a Ph.D (1979) from the University of Maine, US. He followed this up with a two-year stint as assistant professor in the electrical engineering department of Washington State University. In 1981, he returned to India and joined IIT-Madras. By the late 80s, his department had begun to undertake a number of research projects for industry. But Jhunjhunwala saw these one-off projects weren't going anywhere and started looking for the big opportunity. That was when the closely knit Telecommunications and Computer Networks (TeNet) group realised that the solution to India's poor telephone penetration was a system where the cost per line would be less than Rs 10,000. "Even then, we were clear that besides supporting voice, this system should also support a good bit rate for data as well,'' says Jhunjhunwala. After deliberations, the DECT standard was chosen because it was low-cost, but still supported a high data transfer rate. The problem was that it was a micro-cell system and macro-level deployment would pose challenges.It was at this time that Jhunjhunwala formulated the vision for a 100 million telephones in India along with a small group of professors and academicians who were all part of TeNet. But from the initial stages, Jhunjhunwala saw that instead of developing the technology as an academic project, it would be better to have a corporate entity jointly working on the project. He drew up a list of nine ex-students across the country and asked them if they'd like to be involved.Also roped in as advisor was US-based Analog Devices chairman Ray Stata (a great believer in India's research capabilities), who was even then working closely with IIT-Chennai. And soon Midas Communications was born. Stata's experience with Midas has only reinforced his faith in India. "Analog Devices' experience in working with the IIT and Midas teams as well as other design companies in India, confirms that Indian engineers are fully capable of designing and deploying world-class products for the Internet age,'' says Stata.Midas, headed by Shirish B. Purohit, was a unique experiment in combining academic research, entrepreneurial initiative and industry inputs. What was once a laidback R&D setup has been transformed into a beehive of activity. For Purohit and his team of engineer-entrepreneurs, the current financial year is going to be a crucial one. Midas plans to seek venture capital investments once its target of 100,000 telephone lines using corDECT WLL has been achieved. That could happen this year itself. In India, Midas has already licensed the WLL technology to four companies -- HFCL, Shyam Telecom, Crompton Greaves and Electronics Corporation of India. After delays and initial resistance (see BW 14 June 1999), deployment of WLL is picking up in India. Purohit is clear that Midas will continue to be an R&D-driven company, developing cutting-edge technologies and licensing them to outside manufacturers in return for royalties.By the late 90s, Jhunjhunwala and his team of professors including Bhaskar Ramamurthi and Timothy Gonsalves working along with a team of engineers of Midas Communications Technologies had come up with an indigenous solution to India's telecom tangle. After successful field trials by DoT in Chennai and by Telebras in Brazil in Sao Paulo in the mid 90s, the corDECT-based WLL technology was licensed to a few companies in India, Brazil, China and France for commercial manufacture. From the beginning, Jhunjhunwala was clear that any solution would be incomplete if it could not support reasonably high-speed Internet access as well. In fact, one of the main reasons why TeNet decided to go with the DECT standard was because it supported a high bit rate. Midas has already adapted corDECT WLL technology to handle Internet access as well. The result: 35 kbps-70 kbps wireless Internet access, even when existing telephone lines only support a maximum speed of 56 kbps.The low-cost technology developed by Jhunjhunwala and his team consists of three components -- the DECT interface unit (DIU), a compact base station (CBS) and a Wall Set IP (see graphic). The DIU is connected to the telephone network (PSTN) as also the Internet service providers' (ISP's) international gateway. All voice and Internet traffic pass through this, before being routed to the PSTN or gateway respectively. Each DIU can support 10 compact base stations. The DIU and CBS are connected by physical cable. Users within a 35 km radius around the CBS have a wireless voice and Internet connectivity through the wall-set IP. When a user makes an Internet/voice call, the wall-set IP establishes a wireless connection with the CBS and passes on the traffic. The CBS in turn relays the traffic to the DIU, which sends the voice to the PSTN and Internet traffic to the ISP gateway. The success of corDECT WLL has strengthened Jhunjhunwala's conviction that India needs homegrown telecom solutions. But there is still a long way to go before he can get any where near the one-million phones mark. Jhunjhunwala has to contend with problems that include competition from multinationals and lack of vendor financing. "It is imperative that government, through financial institutions, makes available long-term funding for projects based on local technologies like IIT's WLL. The government should encourage local R&D,'' says Nahata. One of the measures that he moots is a change in the revenue-sharing model so that operators using a higher level of locally developed technology will benefit. Other supporters call for more drastic measures to get the best out of the IIT technology. The government should go in for combined basic and cellular licences instead of having two separate licenses they argue, as this will bring down costs allowing the same network to be shared.All these problems are going to take some solving. However, Jhunjhunwala is holding on to that magical figure of 100 million lines in 10 years. "If it doesn't happen, we will raise the funds to make it happen,'' he declares. There are enough pessimists who feel a million lines is a mere pipe dream but Jhunjhunwala is unfazed. Meanwhile, even as they were busy with developing corDECT WLL, another new company Banyan Networks was being incubated by Jhunjhunwala and his team. This spin-off will concentrate on another technology with a focus on the last mile: digital subscriber lines or DSL. The company has a professional CEO at its helm, Vijay K. Raghavan, and has raised money from Intel and IL&FS Venture Corporation. That could mean Chennai will soon emerge as a power centre of telecom research.