µAMPS Research

Research - What is a sensor network?

In recent years, an exciting new type of networks has emerged, called sensor networks. Contrary to traditional computer networks, these sensor nets consist of a large number of autonomous devices. Each device, called a sensor node, is equipped with integrated sensors, data processing capabilities and wireless communications. Figure 1, shows our first version µAMPS-I node, implemented using COTS components, which served as our original testbed. The sensor net vision is to have tens, hundreds or even up to thousands or millions of nodes networked together sensing their surroundings and coordinating amongst each other autonomously. A whole spectrum of applications have been dreamt up, ranging from intruder tracking, machine failure detection, and wildlife monitoring, to biomedical applications.

Fig. 2: µAMPS-I sensor node

To counteract the limited resources of the individual nodes, collaboration is the mantra in sensor nets. By combining the simple sensor readings from multiple nodes, high-quality information can be obtained. This is not unlike individual tv-screens making up an incredible video wall. Each individual node is a mere cog in the sensing wheel, playing its part in the overall network functionality. One of the possible ways to architect the network is illustrated in Figure 2. In this 2-tier setup, simple low-performance nodes fulfil the functionality of trip wires. They do the initial target detection, possibly waking up the more powerful acquisition nodes. Information flows through the network wirelessly, where nodes might also serve as forwarding stations in addition to sensing entities.

Fig. 1: Sensor network deployment scenario

Obviously, to be effective as unobtrusive monitoring networks, the nodes need to be ultra-small, a promise delivered by dense integration, combined with MEMS technology. Exploring these avenues is part of stages 2 and 3 or our µAMPS project. However, a small-form factor also results in limited battery capacity and bounded energy scavenging potential. As a result, energy efficiency is of utmost importance, forming the dominant research focus that is percolating through all layers of the network and node design. In the end, extending the usefulness of the network beyond the limited energy resources, is truly one of extending the network lifetime.