Auctions and iMesh based task assignment in wireless sensor and actuator networks

Abstract

In this article, auctions are used to improve the recent information mesh (iMesh) based distance sensitive service discovery in the wireless sensor and actuator networks (WSANs), thus improving overall energy efficiency and lifetime of the network. An event that needs a single actuator response is sensed by one of the sensors, which then performs the search for the closest actuator to respond by means of the iMesh cross-lookup. The basic iMesh protocol is reported to have 95% efficiency in finding the closest service. We confirmed that result, and improved the efficiency by using the localized auction aggregation protocol (k-SAAP). The actuator found by the iMesh lookup starts the auction to find if there is a closer actuator among the actuator neighbors up to k-hops away. In our simulations, we consider scenarios where the sensors are placed in a rectangular grid, while the actuators form a connected network, or have a random distribution. Simulation results show that introducing auctions gained average efficiency improvement of 4.3% compared to the basic iMesh, which brings us to 99.3% closest service hit rate. In other words, the closest service miss rate was reduced from 5% to about 0.7%, which is significant 86% reduction.