Bell Labs grapples with VOIP, open source

BANGALORE, INDIA – Voice over IP (VOIP) and open source technology hold great promise for cost savings, but also threaten traditional ways of doing business. Rather than shy away from the challenges that these disruptive technologies represent, Bell Laboratories, the renowned R&D (research and development) arm of Lucent Technologies in Murray Hill, New Jersey, is attempting to bring them into the mainstream.

The R&D unit is working on methodologies and processes required to use open source components without compromising on carrier grade capability, so that it can dramatically reduce the cost of developing new telecommunications platforms, according to Jeffrey Jaffe, president of research and advanced technologies at Bell Labs.

During a recent visit to Bangalore, India, where Bell Laboratories opened a research center in December, Jaffe spoke to the IDG News Service on some of the technologies Bell Laboratories is working on, the lab’s research strategy, its distributed model for R&D spanning four countries and other issues. Below is an edited version of the interview.

IDGNS: What are the potentially disruptive technologies that Bell Labs is working on ?

Jaffe: One of the areas that we are very focused on is what I call carrier-grade voice-over-IP. I am actually concerned that as the world moves there, some of the traditional things that we expect in our public switched networks, such as quality, reliability, performance and security, need to be preserved.

We have a collection of projects that basically have the intent to give you the best of the IP world but also the traditional carrier grade that you have in the telephony world, and that merger is I think absolutely disruptive, and will provide tremendous value to the future. Some of that work is going to be done in the Bell Labs research center in Bangalore, because a piece of that work is building operational support systems for the IP world. If you look at how the IP world developed, it was a little bit chaotic. The kind of operational support, network management, monitoring, performance understanding, fault detection — all that you have in the telephony network — never came to the IP network. So the fundamental topic of research is to put all those things into the IP network, to make it carrier grade.

IDGNS: What else is cooking in the labs?

Jaffe: Another thing that I think is very interesting and very important is not in the products that we develop, but in the methodology that we develop the products. Open source is very prevalent in the desktop, (and) Web servers, but for telecommunications systems, because of the carrier grade requirement, open source classically has been less important. We are exploring how do we get the carrier grade capability in open source, so that we can dramatically reduce the cost of developing new telecommunications platforms.

IDGNS: What are the challenges in using open source?

Jaffe: A lot of the challenges have to do with testing. The promise of open source is that much of the componentry comes from the industry rather than develop it yourself. But when code comes in from the industry, developed with unclear software technology processes, not clear whether it was rigorously tested, and then you stick it all together and just hope that it works, it is not clear that that is a big savings. So coming up with new processes and technologies, so that we can bring up open source to be able to be at a level of something that you had planned and done yourself, is a nontrivial problem in software methodology.

IDGNS: Bell Labs is also doing work in the area of nanotechnology. What are the potential applications of this research?

Jaffe: Nanotechnology is the technology of preference to make things small, light and cheap. We are for example focused on how to make better cell phones using nanotechnology, which gives us the ability to build more capability into the existing form factor. With MEMs (microelectromechanical systems) microphones, since MEMs are smaller, you can have multiple microphones, giving better directionality and improving voice quality. With our nano batteries that we are developing, you have longer battery life cycles. We have silicon antennas, which gives you more capabilities in terms of wireless communications at a lower cost and less intrusion.

Another dimension of nanotechnology is a technology that we call nanograss. We have looked at certain surfaces under the microscope and at the nano level they looked like two dimensional spokes, like pieces of grass that are sticking up. The result of that is that if you put a drop of water, it just floats on top of it, it is like there is no friction there, because these spokes are very close together. But with an electrical charge, then it sinks between the spokes, and then it becomes immobile. That allows us to direct water cooling to different parts of a chip. So if you had a processor board or a chip, and locally there was a lot of processing going on, so it is very hot while other parts weren’t hot, this could allow you to move the coolant to where the hot spots are. Telecommunications systems are extremely hot because of the hundreds of gigabits that are going through them simultaneously. So this is an example of how by understanding the nano properties of materials we are able to do better cooling and better packaging.

A third example of nanotechnology is we have used it for optical routing. We have built MEMs mirrors on an 8-inch wafer with a million MEMs mirrors on it, each individually moveable, that you can use for routing of optical paths.

IDGNS: Do you have a research strategy for markets in emerging economies? There is this perception that the requirement of these markets are quite different from those in the developing economies.

Jaffe: In my view, the best strategy for increasing teledensity is to get the costs down. What is interesting about reduced cost is that it helps everyone. In the last decade or so, there have been fantastic improvements in teledensity in India, for example, and I believe a lot of the reason for that was reduced cost, and a lot of those reduced costs were enjoyed by people the world over. If you look at things like convergence, that is the next generation of driving costs down. If you look at nanotechnology as an enabler to make things even smaller and cheaper, that is another enabler to drive cost down. From time to time, we might look to some extent at specific issues for developing countries, but I think that is not as powerful as just getting the cost down, and certainly we have seen evidence of this over the last decade.

IDGNS: One of the key attractions of Bell Labs for researchers was its multidisciplinary environment. Engineers for example working on Unix had the opportunity to interact informally with other engineers working on some thing as different as plastics for example. Do you think that you could lose some of this multidisciplinary environment by the distributed R&D model that you are setting up with centers in the U.S., Ireland, India and China?

Jaffe: This consideration guides how we expand. We are in four countries, and we are not in 24 or 104 countries. We only want to come to countries where we have a sufficient commitment to the country that will build a sufficient critical mass to have the interdisciplinary research that you are talking about. In my view you need a minimum of 25 (staff).

There is one other factor, and the other factor is that the world is getting smaller, and I do actually anticipate substantial collaboration amongst the different global research centers. One of the interesting projects that we have in Beijing is called “The Great Time Zone Gap Project,” and what that is trying to do is make people working across different time zones feel like they are really working together.