Advances in wireless communication techniques as well as in low power electronics and processor technology, together with a growing interest in monitoring remote environments without any a priori communication infrastructure, have led to the proliferation of wireless sensor networks and have made the investigation of such networks an important area of study. Wireless sensor networks consist of a large number of small, low power devices, equipped with limited processing and communication capabilities, which measure a physical phenomenon in some region of interest. The goal is typically to reconstruct the measured physical field (e.g. temperature, pressure, light) at an intelligent central unit to within some prescribed accuracy, and this at the lowest possible cost on the communication link and power consumption by the sensors. Wireless sensor networks have some very specific characteristics that differentiate them from most classical communication systems. The data transmitted by a sensor network is not generated by independent sources but results from measuring a physical phenomenon, which provides the data with a particular spatio-temporal correlation structure. Owing to low power requirements and generally prohibitive inter-sensor communication costs, the major part of the data processing has to be performed in a distributed fashion and only very limited collaboration among sensors can be expected. The transmission channel is a wireless multiple access channel, which implies that the sensors can overhear each other’s transmissions and that the central unit has to deal with interfering transmissions.
We are investigating the role of physical laws in source coding problems appearing in typical sensor network applications with the aim of quantifying the potential gains in terms of rate-distortion behavior that exploiting these laws can provide and also of devising practical strategies to approach the theoretical bounds. As a first application, we have studied the heat diffusion process occurring in a metal ring when it is exposed to heat sources. Sensors, which are located on the ring, measure the local temperature values in the ring at regular time intervals, encode these measurements at a given bit rate and convey them to a base station, which attempts to estimate the spatio-temporal temperature profile in the ring to within some prescribed accuracy. For this scenario, we have determined the rate-distortion functions for different coding constraints and we have compared the performance of several distributed coding approaches.
Current and Future Research
Our current research efforts deal with extending our results to other geometrical structures where the resulting temperature processes are no longer space invariant, and, consequently, other analysis techniques have to be employed. Furthermore, we will study additional physical phenomena like electromagnetic waves, acoustic waves or gas diffusion.
Baltasar Beferull-Lozano, Martin Vetterli, Emre Telatar (EPFL – Information Theory Laboratory (LTHI)).
February 2003 – ongoing.
Swiss National Centre of Competence in Research for Mobile Information and Communication Systems.
B. Beferull-Lozano, R. L. Konsbruck and M. Vetterli, Rate-Distortion Problem for Physics Based Distributed Sensing, Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), May 2004, Montreal, Canada.
B. Beferull-Lozano, R. L. Konsbruck and M. Vetterli, Rate-Distortion Problem for Physics Based Distributed Sensing, Proceedings of the Third International Symposium on Information Processing in Sensor Networks (IPSN), April 2004, Berkeley, California, USA.