Path Inference in Wireless Sensor Networks

Abstract

 Modern wireless sensor networks (WSNs) have become increasingly complex with increasing network bandwidth and dynamic nature of wireless communications. Many measurement and diagnostics methods rely on packet routing paths for each accurate and accurate analysis of complex network behaviors. In this research, we suggest iPath, a new path-biasing approach to rebuilding packet routing paths in dynamic networks and broadband networks. The basic idea of iPath is to exploit a great similarity in the path to long paths derived from short paths. IPath starts with a known set of paths and executes the path conclusion infrequently. IPath includes a new design for a light-weight haash function to verify generated paths. In order to further improve inference capability and efficient implementation, iPath includes a fast boot algorithm to re-create the initial set of paths. We also implement iPath and evaluate its performance using the effects of large-scale WSN deployments as well as comprehensive simulations. The results show that iPath achieves much higher re-creation percentages in the different network settings window compared to the newer methods.