Body Area Networks

Changing the world, one crazy idea at a time

by Peggy Aycinena

Julien Ryckaert received his M.S. degree from the University of Brussels (ULB), Belgium, in 2000. He then joined IMEC in Leuven, Belgium, as an RF designer. He is currently doing his PhD research on an ultra-low power transceiver for low data rate ultra-wideband applications.

I had a chance to speak by phone with Julien several weeks ago. It was a bit challenging setting up our phone call as he was setting the appointment by GMT – Greenwich Mean Time – whereas I was thinking in terms of the hour of the day in California versus the hour of the day in Leuven, neither of which are the hour of the day GMT. When we finally did connect, it was a pretty darn interesting conversation. Please read on to see why.

************

IMEC is a technical company, and our group is in a particular division. There are about 1500 people in total at IMEC, but in this group we are only 150. We use the technology being developed at IMEC and apply it specifically to wireless applications. Our work is divided into two main programs:

1. The M4 Program is for multi-standard applications, like software radio. This program concentrates most of our efforts.
2. The Human++ Program. This is where our wideband pulsar comes in.

The Human++ Program is where we are trying to develop Body Area Networks. We want to develop a set of sensors that would capture all kinds of health information. Things like body temperature, the level of glucose in the blood, the heartbeat, an EKG.

All of this information would be centralized in a central node – kind of a PDA – which would transfer the information to a doctor. We're in the process of proving the enabling technology to make this work. We're not a huge company. We only want to identify the difficulties and where companies might have problems [going forward] in developing these systems.

For instance, we're developing sensors, and also some packaging technology. For instance, if you swallow a pill/sensor, you would need some kind of special packaging for the sensor. That's the area that I'm working on. Of course, we're also working on the wireless technology – how to develop a radio to transmit the information without a wire.

We are working with some doctors at a hospital here in Leuven who are working on epilepsy. We're also working with some universities. I'm not an expert in this area, but if you look at brain activity by putting several sensors around the head – the [researchers] are trying to capture from the brain activity the signal that shows a pending epileptic crisis. We are working on the software and circuit algorithms, so that patients would know in advance [of a crisis] to sit down and protect themselves.

We are also working on an electromyogram sensor, which would be important for surgeons so they can see responses. Sometimes it's important for a surgeon to check if a muscle is contracting during surgery when the patient is under. The sensor could be put on the skin and a color code would tell the doctor whether a muscle is contracting, and how much.

But these are just some examples of what we are doing. We are trying to demonstrate the enabling technologies to show that these things could work. How to sense the different kinds of information around the human body is the focus of our work.

Power is a problem. If the sensor is inside the body, the user needs not to have to replace the batteries every week. So, you also need the sensor system to be as low power as possible. There are two directions you can take to make this happen.

Make all of the functions inside the sensor work at as low power as possible – and it's a challenge to make a wireless radio that consumes so little power has that it has years of lifetime.

Second, you could want to know how to make an energy scavenging system. This system would scavenge excess energy from the body itself to operate on. A human body usually walks and moves in various ways. You could capture the mechanical energy [of that system] and harvest it to power the sensors inside the body.

With today's technology, it's difficult to reach the power level required by those sensors – the power constraints that the sensors require. We need to find a completely novel way to solve these problems, not just by using low-power radios.

So, we have to completely rethink the radio. Typically, radios try to get as much data rate as possible – a 50 Gbit wireless LAN, or a 100 Gbit wireless LAN. But if you only need 100Kbits of data per second, the picture is completely different. Optimization has to go in a completely different direction.

We believe that ultra-wideband is probably a very good candidate for this kind of problem. UWB is a hot topic today because people think it will solve all the problems in wireless [designs]. I think UWB will solve some problems, but not all the problems of wireless systems.

We've demonstrated a UWB pulsar using 802.15.3a. This standard is one aspect of wideband and an alternative to ZigBee, which is also being considered because of the low-data rate standard. According to our first simulation, we believe we'll be using from half, to an order of magnitude, less power than a ZigBee radio.

Actually, one of the most interesting aspects of comparing the power numbers is that ZigBee will consume 10 nanojules per bit, and that will always be the case. Traditional radio can't scale that number. Even if the environment is better, it will always consume 10 nanojoules per bit. But, UWB allows you to scale down to lower areas of power consumption, and you can play with the number of pulses you need for just one bit.

So, it really depends on what data you want to capture. To capture body temperature – if the information flow is every 1 to 10 second, even if it's wrong once but the rest of the time it's correct, that's okay. Small errors can be tolerated.

However, with heart attacks – if you want to detect a heart attack, you can't tolerate a mistake. The device on the heart shouldn't tell the doctor that there's a heart attack when there isn't. So, it really depends on the application [as to what data rate you need].

We're just at the beginning of our effort. For the moment, we're only trying to identify where all the problems will be, and [for now] we've identified the transmitter as the biggest problem. Later, we will evaluate the complete system – it's only when the complete system is built, that you can expect to see the larger solution.

The market is difficult right now, and unpredictable, particularly in wireless communication. It's so hard to know what direction the market will go in wireless. Even for large companies that have thousands of people working for them, it's difficult. The market's evolving so fast.

I think the companies should be close to the real market for [their current] products, but companies should also look at what's coming in the future. That's where research centers and universities come in, because they look at what's going to come next. They have all the crazy ideas, and the ability to consider what's going to be needed to change the world.

IMEC gives us the opportunity to study these things because it has a different kind of business model. IMEC gets its money from different sources. There's the government on one side – and partially some money from the EU – which is something like the DARPA programs in the U.S.

But, the most important part of our money is from bi-lateral or multi-lateral contracts with companies. That's the way we try to transfer these technologies [to the commercial sector]. We try to attract the main players in the world, who come here to IMEC depending on the things they want to develop or sell.

IMEC is supporting the creation of new companies, and the government supports IMEC by supporting these kinds of initiatives. But, you have to have a very good business model [to prove your plans], and you have to create products.

IMEC develops generic solutions, so it's not always possible [to create products from the work here]. But there are several spin-offs each year [nonetheless]. Companies like Intel and Motorola come to IMEC, and we don't compete with them. We're way too small to compete. [Instead], we want to be a kind of driver for them – giving them the enabling technology to show them how to make progress in the coming years.

****************************

Editor's Note: IMEC (Interuniversity MicroElectronics Center) describes itself as "Europe's leading independent research center in the field of microelectronics, nanotechnology, enabling design methods and technologies for ICT systems."

You can learn more about IMEC's Human++ Program by going to their website: www.imec.be.