Data aggregation in resource-limited wireless communication environments — Differences between theory and praxis
The primary objective of wireless sensor networks is the monitoring of a system or area by measuring different kinds of sensor data. Dependent on the application scenario, a specific measurement scheduling scheme is preferred. Based on this scheduling, an optimisation of the available energy resources in each sensor node and accordingly in the whole network topology is possible. At the same time, an application-specific synchronisation of the measurement points ensures a maximum quality for the merged sensor information. One essential requirement for any kind of scheduling scheme is a synchronised time base in the entire network topology. In order to define capable measurement schemes for any kind of available sensor node hardware, lightweight synchronisation concepts are necessary. In this paper the authors present a distributed scheduling concept for embedded, resource-limited sensor and Ad Hoc networks. The concept operates on the application layer and enables application-specific, energy-efficient measurement schemes. Therefore, the proposal uses a resource-optimised, flexible time synchronisation algorithm. For proof of concept, the approaches were implemented on a real-world sensornet hardware platform. Several test scenarios evaluate the feasibility and clarify an excellent usability in wireless sensor network environments.