Modeling Data Gathering in Wireless Sensor Networks

The predominant protocol development methodology in the area of wireless sensor networks today can be characterized as being one of design-by-intuition/validate-through-simulation. However, a concurrent development of complementary theoretical models for these protocols is essential for the rapid advancement of this technology. This is because the severe resource constraints on energy, computation, and storage that are characteristic of these networks make it crucial to optimize protocols, in order to maximize network lifetime while providing an acceptable quality of sensed information.

We survey in this chapter some examples drawn from recent studies pertaining to data gathering in sensor networks. These examples demonstrate the importance of optimizing protocol parameters carefully to maximize performance.We emphasize the use of a first-order mathematical modeling approach. The optimizations are all based on simple expressions that are intended to capture the impact of essential environmental and protocol parameters. This approach is motivated by the desire to understand easily the key design trade-offs that are involved in each context. As we shall see, even such simple models can yield fundamental design insights to improve the performance of practical protocols.