HS-Sift: a Hybrid Spatial Correlation-based MAC for Event-driven Wireless Sensor Networks

Abstract

Data transmitted in large scale dense wireless sensor networks (WSN) is often highly spatial correlated, and filtering out the spatial redundancy has been proved to be effective in energy conserving. At the same time, according to the results of Min Dong, different MAC mechanisms should be applied to different SNR regimes. But no existing MAC schemes design exploiting spatial correlation take this into consideration. This paper tried to seamlessly adapt the MAC behavior between the above two medium access schemes in the process of exploiting spatial correlation. The design, implementation and performance evaluation of a hybrid MAC protocol, called HS-Sift, is presented in this paper. HS-Sift divided the entire sensing region into three sub-areas and applied different MAC mechanisms to each part. Under the proposed scheme, sensor nodes around the sensing region border sleep most of the time, sensor nodes near the event site are carefully scheduled to achieve high channel utilization and sensor nodes lie between competing for the shared channel with different transmitting probability and backoff time. Extensive simulation has been done to verify the effectiveness of the proposed scheme in terms of energy consumption, transmission delay while ensuring a maximum reconstruction distortion requirement.