An Integrated Design Environment to Evaluate Power/Performance Tradeoffs for Sensor Network Applications

Abstract : Networks of inexpensive, low-power sensing nodes that can monitor the environment, perform limited processing on the samples, and detect events of interest in a collaborative fashion are fast becoming a reality. Examples of such monitoring and detection include target tracking based on acoustic signatures and line -of-bearing estimation, climate control, intrusion detection, etc. The advances in low-power radio technology are making wireless communication within sensor networks an attractive option. However, it is typically difficult or impossible to replenish energy resources available to a portable sensor node, once it is deployed. Maximizing the life of sensor nodes is an overriding priority, and different energy optimization techniques are being developed to address computation/communication tradeoffs. A large number of research efforts are focusing on different aspects of the general problem of designing efficient sensor network-based systems - where the metrics to measure efficiency vary from system to system. With technological advancements such as silicon-based radios expected to become a reality in a few years, designers of sensor network-based systems will be faced with an extremely large set of design decisions. Each choice will affect the overall system performance in ways that might not always be cleanly modeled. In addition to the research challenges in design and optimization, the practical aspects of designing real-world sensor networks will become equally important. For example, the ability of the design framework to allow rapid specification and evaluation of a particular network configuration is critical for a more exhaustive exploration of the design space. A design environment for future sensor networks should provide tools and formal methodologies that will allow designers to model, analyze, optimize, and simulate such systems.