Software radio is a path-breaking innovation in wireless communication. Vanu Bose, the man behind this innovation, says software radio is set to revolutionise rural telephony, both, in India and the United States.
Vanu did his software radio research at the Massachusetts Institute of Technology. Bose was named one of the 29 Technology Pioneers for 2005 in December by the World Economic Forum for this innovation. The works of these pioneers, according to WEF, are expected to have a positive impact on societies.
Vanu Bose, founder of Vanu Inc, is the son of Amar Gopal Bose who founded the Bose Corporation which produces Bose acoustics. Both father and son are innovators par excellence.
Vanu Bose was in Chennai after he chose iSoftTech to develop technology products for software radio. In an interview with Shobha Warrier, Vanu Bose talks about his big plans for India and the United States.
What is software radio?
It is an emerging technology, that allows network operators to simultaneously support multiple communications standards such as Global System for Mobile Communications (GSM), Code Division Multiple Access (CDMA), Wide-band CDMA (W-CDMA), third Generation protocol (3G) etc) on one network infrastructure without being bound by a particular standard.
Vanu Inc's software is the first Federal Communications Commission approved software radio in the market. Vanu Inc offers commercially available software radio solution that uses a single-PC-based hardware platform to support multiple wireless services and standards entirely in software.
Then came the strategic partnership with the Centre for Development of Telematics (C-Dot) to make use of the power of software radio base station for India's rural communications market.
What are your plans for India?
We have an existing relationship with Indian companies, like the pact with C-DoT. We are running a system for them. We are also setting up a development centre in Chennai. We are doing it for different reasons than people typically set up software development in India.
The typical reasons are people want to hire armies of testers and debuggers to work on their products.
They also want to outsource engineering products to save costs. I don't believe the cost savings is going to last long. Salaries and infrastructure costs are rising while in the US, it is not rising as quickly, so the gap is closing in. We plan to go slowly and try to hire the best talent.
We are very selective. We find that there are very talented graduates in electronics and computer science from IIT Madras, and we want to further our relationship. Our research came out from MIT, Boston and we have a good relationship with MIT. I would like to replicate the same thing with IIT Madras and create a small high quality team.
It is for this reason that we have decided to go with iSoftTech (Integrated Softech Solutions). We need people here to support our customers like C-Dot to build products for the Indian market and ultimately to build products for other places as well. We also want to firm up research collaboration with students and faculty at IIT Madras.
How significant is the Indian market for software radio?
We are specifically targeting wireless communications for rural India, with C-Dot. So, for the rural areas, the problem is very different. It is very demanding and traditional equipments do not have the plus points or the features and functionalities that you need in these markets.
The flexibility in the software radio technology can be used for that market. So it is a very significant market.
Is it just cost effective, or are there any other functionalities that would help rural India?
One of the key functionalities is, in India, you have the GSM and the CDMA. If you want to have nationwide coverage including the rural areas with the traditional way, you have to build both the GSM network and CDMA network.
With our software radio technology, you can use the same hardware infrastructure and one software for GSM channels and then add up with software when you have the CDMA channels.
In short, you can have one network and support both the systems. It is much more cost effective to build a nationwide network with our software radio technology.
Is your software radio processor intensive? For example, will it allow a PDA phone to tune to AM/FM radio and switch from GSM network in India to say CDMA in the USA?
We are not doing anything with the handset. We are only working on the infrastructure side which is the equipment at the tower. So, whatever handset you have, GSM or CDMA or PDA, we can have the software for the infrastructure to talk to both.
The CDMA and GSM handset can talk at the same time, which you cannot do today. That's is a big difference. And you cannot afford to have two networks in 600,000 villages. So, you need one network that supports both standards and other standards in the future. We are starting with GSM, we will have CDMA in future.
How do you plant to support two networks?
Traditionally, the equipment is dedicated to one standard. You buy CDMA equipment if you want to support that standard, and if you want to support GSM, you have to buy that equipment.
What our technology does is we make different software. With one set of generic hardware, we can support different software. That is the key. It can even support new standards in the future.
Today, if you want to update, you have to change the hardware?
Yes, today we have to replace the hardware. For example, in the United States, when AT&T Wireless converted to GSM, it cost them $10 billion to put new GSM equipment across the country.
If our technology had existed at that time, you could just change the software in some of the existing systems and support GSM which would have been faster and more cost effective.
So, what we have to do is increase the pace of innovation in wireless because technology in general is moving quite quickly; processors get faster, there are new software. But when you look at cellular telephone technologies, it has actually moved at the pace of one new standard every decade.
In the US, it was the analogue in the eighties, and digital in the nineties, and GSM this decade. It is a very slow process. The reason is not technology. Technology is moving fast. The reason is economics.
For example, when AT&T spends $10 billion to build new infrastructure, they need to have that equipment operating for 7-10 years to get a reasonable return on the investment.
This is what sets the pace of innovation. We want to change that hardware upgrade to software download, and it will be much faster and more cost effective for carriers.
If you were to sell your technology to an Indian network provider who is setting up telecom infrastructure, how cost effective will it be compared to the conventional one?
There are two ways of saving capital expenditure and operational expenditure. The capex could be saved by about 25 per cent but operational expenditure is very significant.
Our system is smaller and requires less space, less power, and because we use the IP protocol, we can run multiple different kinds of back ends which is a significant cost. All that add up in the operational expense can be 40-60 per cent lower than conventional technology. That is an annual expense. So, capex is not the important part but the big thing is the ongoing saving the technology can enable.
If an existing telecom provider has to convert to software radio, how much do they have to spend?
It is similar to what they have to invest in a new technology. It will be a little bit less for the upfront capex but after that, they will be saving money every month because of low operation expenses.
You are talking about implementing your technology in rural India. Is it a social commitment or just business?
Social commitment has to be combined with economics. Only then, it will make sense. I think our technology truly gives an opportunity to do something that has great social value at an affordable cost.
It is exciting to bring wireless communication to people who have never had even landline connection. We can help do something for social good that makes economic sense.
You were born and brought up in the US, so what bring you to rural India?
I have spent a lot of time in India. Every summer I used to come to Chennai and be with my grandmother and cousins. Honestly, with all the software development and the economic growth, India is truly emerging. In ten years, it will be a significant economic power.
So, India is now an attractive destination for business. It is an emerging technology power in the world and we want to be a part of this and grow with it.
Is the presence of talented youngsters a major reason for India to become be a technology power?
Yes, that is why we are with iSoftTech. We have set very high hiring standards, in Boston and we also expect high quality from others. We want to continue doing that. For that, we have to tap the best talent pools in the world.
In the Boston area, we have the best talent pool with MIT and all the other universities there. But many of the students from the graduate schools in the US are graduates from IIT Madras. So, it makes sense to come here. I think the reason for high talent is not just because of the IITs.
Maths, science and English education at the high school level in India is excellent. English is also the technology language. This places India above all the other countries looking for software development.
Cordect WLL technology developed by Prof. Jhunjhunwala of IIT, Chennai is a homegrown technology. Do you have any plans to support it?
We have been talking to Midas Communications. We will be doing something. Cordect is an excellent example of designing technology to meet the needs of the rural market in India.
One of the things about designing for the markets in India, especially the rural markets, is that the constraints are very challenging. If we can solve those problems, we can achieve a lot of progress.
Can you explain what the constraints are?
The cost obviously is a big one. You have to design a system such that it works with unreliable power sources and unreliable communications network in remote areas. If you can solve them, you can use that for applications in other places.
We may actually see some applications for Cordect in the US, and are evaluating it.
GNU, the open source project, has a free, open source software radio. What are views on this?
GNU is actually a great story. GNU has an open source software radio product called GNU radio. One of the things we feel good about is, the original software for the GNU product came from our MIT research.
The software radio at MIT that we put in the public domain became the GNU code of the software radio. It has changed now significantly since then. I like the fact that we helped start that.
How different is your software radio and what are the new functionalities that you offer?
It is different and, to be honest, I think GNU radio is an excellent form. The difference between what we did at MIT and what we are doing at the company is the robustness of the software, the integration of management of capabilities, things that will help the actual operation.
So we have different features in the commercial contracts. I don't think GNU radio is focused on that at the moment.
Once GNU radio starts supporting more radio forms, won't you be pressed to beat the free open source software?
We are not sure about that. We view the GNU radio as a positive thing because many universities have started to use GNU radio. So, the new graduating engineers are familiar with software radio and we can hire them later. There is more awareness about software radio because of the GNU radio.
So, we feel the publicity software radio gets is good for us. That is one of the reasons why we opened the software in the first place.
What exactly attracted you to software radio when you started your research? Now, university students are aware of this but it did not exist when you started research on this.
Yes, it did not exist then. It was the idea that wireless systems are built in hardware, which are not very flexible. Whether you are in an area with lots of traffic or little traffic, whether there is good signal quality or bad, you start using the same standard or same style of protocols that exist in the standard.
We thought we could do much better by building a software system that could dynamically adapt to whatever your needs are, the channel conditions are, the traffic requirements are, etc.
If you look at the PC, it replaced a whole lot of devices like typewriters, calculators, ledgers in an office. Our idea was to do the same thing for wireless -- make one general device where we can write software and run multiple different applications on it. That was the driving motivation at that point of time.
Fortunately because technology has progressed so much in the last ten years, we can do a lot more today than we could do ten years ago.
Did it put a lot of pressure on you to come out with something very innovative and different from what your father had built?
No, not really. A lot of people ask that. There was no pressure or competition at all. When I was working at the MIT, it was very exciting and I started seeing the commercial potential of software radio.
If I had not started my company, I would have probably been working for my father and doing equally interesting work. I didn't feel that I had to start a company of my own. It just happened that way.
My father is very supportive of whatever I wanted to do. At one point, I wanted to be a professor. But I am glad that I started my company.
Is it a natural course for all innovators to be entrepreneurs? Is it necessary for all innovators to look at the business aspects of their innovation?
It is not necessary at all. It can go both ways. Some people who are innovators stay in the technology side and bring in other business people. Sometimes that works, and sometimes that doesn't. It can go either way.
In my case, I was interested in driving the rest of the business because I wanted to make sure that the technology was realised in the market place. That drove me to be an entrepreneur. I find interesting challenges in other areas like sales, marketing as well as engineering. All aspects of business require creativity and innovation. It just takes different forms.
Did the trend of innovators turning into entrepreneurs start with globalisation?
I don't think it is due to globalisation. It has been happening in the US for a long time. My father started his company in 1964. I think you are starting to see globalisation effects in India.
You would not have started a company like iSoftTech ten years ago in India. I think you see more entrepreneurship in India and other places as a result of globalisation.
The general belief is that innovations are more accepted in developed countries. Do you think it is true?
I think in many countries, established businesses are viewed as safer choices and there isn't much value placed on innovation entrepreneurship. That is not the case in India.
Many startups in the US have been started by Indians. It is a part of the Indian cultural mentality to be innovative and entrepreneurial. I don't see that being a problem in India.
You see a lot of people preferring to join smaller companies in start-ups in India than the large established companies. People in India see the value in that.
Is software radio technology more useful in a country like India than a rich country like the US?
Yes, in certain applications, especially in believe the rural market, it is more useful. The rural market is our main focus.
What are your plans for the US and the Indian market?
We are focussed primarily on the rural market, and then moving up to the major carriers. Even with the major carriers we are working in the US, we are focussed on solving rural coverage problems because those are the areas where the conventional technology is not cost effective and does not have the right set of functionalities.
So, in a sense, the market plans for both India and the US are very similar. Our focus is entirely North America and India. We are not trying to be global at this point and going to different countries right now. India is the most compelling international market.
Photograph: Sreeram Selvaraj
