Sandpile Model Based Load-Balancing Algorithm In Wireless Mesh Networks

Wireless Mesh Network as a new kind of wireless network, with its unique advantages to get the attention of academia and industry, has a wide range of applications. However, the network resources can’t adapt to the rapid increase in the number of users and large data services, leading to network performance degradation. The network performance should be guaranteed by building a reliable network connection. And this is the prerequisite based on which Wireless Mesh Network have been used widely. Load balancing can equivalently distribute data traffic and make reasonable allocation of network resources to improve network performance. Thus, load balancing in wireless networks is worth researching.Load balancing algorithms are designed with the aim of making network load evenly distributed to each node of the network. The traditional load balancing algorithms are based on the shortest path and make the load concentrated in a small number of nodes on the shortest path. Thus, some nodes are overloaded and will lead to the node congestion. Consequently, this will result in network performance degradation, especially the increase in packet loss rate.This paper analyzes the advantages and disadvantages of the existing load balancing algorithms, and summarizes the problems of existing algorithms. Then the similarity between the sandpile model and load balancing mechanism were analyzed. Based on the sand collapse mechanisms, the load balancing problems in Wireless Mesh network was modeled as sandpile model. Then, a load balancing algorithm based on the sandpile model was proposed. Finally, the performance of the algorithm was analyzed through simulation from three aspects:packet loss rate, average end to end delay and throughput. The results show that the algorithm improves the network performance to some extent.