Evaluating the Impact on Application-Related Metrics for Sensor Data with Concurrent Voice Transmission in Mobile Wireless Mesh Scenarios

Abstract

The transmission of sensor data over wireless mesh networks in mobile scenarios is a strongly emerging application in the field of tactical and mission critical networking. In these scenarios, sensor data from the deployed units like GPSposition, vital and environmental data is transported over a wireless mesh network to a centralized command point. The inherently imprecise sensor measurements are processed at the command point using sensor data fusion algorithms in order to improve their precision and reliability. For evaluating the influence of node mobility and the effects of the wireless mesh network on the quality of the sensor data fusion result, it is not sufficient to only consider network-related metrics like throughput, delay or loss rate. Instead, application-related metrics must be considered that directly rate the precision of the processed data based on the fusion result. For this reason, we developed an integrated simulation and evaluation environment for sensor data fusion applications based on the popular network simulator ns-2 that allows for the inclusion and processing of real sensor data. In this paper, we present evaluation results for the transmission of GPS data in a typical tactical scenario. By doing so, we confirm that the quality of the fusion result can not be judged by evaluating network-related metrics only. Using the integrated simulation and evaluation environment, though, allows to analyze the effective impact on application level.