Det-LB: A Load Balancing Approach in 802.11 Wireless Networks for Industrial Soft Real-Time Applications

Wireless control systems for industrial automation have been gaining popularity in recent years owing to the low cost of their components and their ease of deployment. Wireless local area network (WLAN) is a good candidate for industrial wireless applications, especially for soft real-time control systems. However, unbalanced traffic loads in areas that are covered by several access points (APs) could cause significant performance degradation, and the problem would be more serious in industrial contexts that have high requirements for timeliness and reliability. In this paper, we propose a deterministic load balancing algorithm (Det-LB) for WLAN, which is based on game theory, mainly intended for soft real-time control systems. The procedure of the proposed algorithm is described and analysed in detail. To validate the goodness of the proposed approach, we compare Det-LB with three other load balancing strategies (RLF, MMF, and DLBA) in a simulation. The results show that Det-LB can achieve better load balancing performance and improve network performance significantly. For instance, in a common scenario in which five APs are deployed, the RLF scheme has terrible performance and cannot be applied to industrial applications at all; Det-LB has 35% and 26% performance improvements over MMF in 500- and 250-byte payload size conditions, respectively, and has a 5% deadline miss performance boost and better packet loss performance compared with DLBA when the APs tend to be overloaded.