A Link-Based Variable Probability Learning Approach for Partially Overlapping Channels Assignment on Multi-Radio Multi-Channel Wireless Mesh Information-Centric IoT Networks

Abstract

The performance of the wireless mesh information-centric Internet of Thing (IC-IoT) networks can be greatly enhanced by adopting multi-radio multi-channel (MRMC) and partially overlapping channels (POCs). However, the network interference and channel assignment in IC-IoT networks become more complicated while using both MRMC and POCs. In this paper, a logical link-based partially overlapping channels interference model is analyzed to mitigate the inter-channel interference, and a channel selection scheme is formulated as a potential game. Moreover, a variable probability learning algorithm is proposed by selecting a channel in the strategy space based on its probability. The channel usage probability can be changed by its link utility. The channel with a larger link utility is then with a bigger probability in the strategy space. The simulation results show that our proposed algorithm can achieve high system throughput with fast network convergence.

Keywords

• Multi-radio multi-channel (MRMC)
• partially overlapping channels (POCs)
• potential game
• variable probability learning algorithms