Fuzzy Inference Systems for Load Balancing of Wireless Networks

Wireless networks offer many benefits such as cost-effectiveness, flexibility, and convenience. However, they have challenges that can impact their performance and security, one of which is scalability. This research focuses on the scalability challenge, specifically on the user-Access point (AP) association process and load balancing technique, which ensures that devices are associated with the best AP for their current location and network conditions. This paper presents the development of Fuzzy Inference Systems (FIS) to address the load balancing problem using Mamdani and Sugeno models to examine their impact on the candidacy connection of the AP. In addition, the use of different membership functions on Mamdani-type fuzzy inference system are investigated to evaluate the suitability for subscriber to connect to the AP (i.e., AP’s candidacy of connection).