Research On Adaptive Optimization Strategy Of MAC Protocol For Wireless Sensor Networks

As a favorite protocol in wireless sensor networks, IEEE 802.15.4 has been extensively studied. However, its reliability and energy efficiency would become poor when the network load is high, due to its MAC parameters are fixed. In order to address this problem, this dissertation proposes two adaptive parameters tuning approaches for different types of IEEE 802.15.4 network.First, we evaluate the impact of the MAC parameters on the reliability and the energy efficiency of the network. This study provides a few guidelines for the design of adaptive strategies for reliable and energy-efficient operations.Second, we introduce the well-known adaptive strategy:ADAPT, and analyze its drawbacks. Then a new tuning approach is proposed for effectively meeting the application-specific reliability in the ACK mode under a wide range of operating conditions. Compared with ADAPT, the simulation results show that FAD APT enjoys a better convergence and stability.In order to make the tuning valid without acknowledgement, we propose another parameters tuning approach:B-FADAPT. The key method here is estimating the colliding probability from the observation of CCA. We show that B-FADAPT is effective, in the sense that it can satisfy a target reliability constraint while consuming low energy.