A Collaborative Filtering Recommendation-Based Scheme for WLANs With Differentiated Access Service

Abstract

As the deployment of WLANs becomes extensive, users commonly have to select one access point (AP) for association among multiple available APs. However, the existing AP selection policies are not sufficient enough due to their drawbacks (e.g., imbalanced load and time consuming for association). In this paper, we present a collaborative filtering (CF) recommendation-based AP selection strategy in wireless local area networks (WLANs) with differentiated access service. At first, a differentiated AP selection strategy by utilizing analytic hierarchy process (AHP) is leveraged to optimize AP selection metric with multiple objectives for different users. Furthermore, an improved differentiated AP selection approach by CF recommendation is investigated. Based on our method, the new user can associate with the recommended AP promptly instead of channel scanning and selection metric computation. Meanwhile, load prebalancing module periodically excludes the AP considered to be overloaded from the recommendation list, while the pioneer user (PU) updates the corresponding module periodically with an adaptive method to replace the AP with low recommendation efficiency. Simulation results demonstrate that our presented strategies can provide differentiated quality-of-service (QoS) assurance to users, improve overall network throughput, balance network traffic, and degrade association time cost.