Channel Selection Relying on Probabilistic Adjacent Channel Interference Analysis and Pattern Recognition.

The number of Wi-Fi devices and their requirements for bandwidth keep on increasing, along with their hunger for spectrum. This fact is mostly noticeable in dense urban scenarios where neighbors fight for bandwidth and their networks struggle to deliver the requested information. While the inherent limitations of Wi-Fi technologies cannot be overcome, they can be mitigated through configuration. Optimally selecting a wireless channel is a critical aspect of access point (AP) configuration, and a challenging task due to a broad range of factors affecting the wireless connection performance. This paper addresses the channel selection problem by relying on a time-varying dynamic approach capable of modeling its surrounding wireless networks with respect to their usage patterns, channel utilization and adjacent channel interference. This contextual data is used in a channel selection model, which combines utilization patterns and statistics with a probabilistic mathematical model to accurately estimate the impact of adjacent channel interference on the signal to interference plus noise ratio, hence effectively selecting a wireless channel whose optimal performance is exhibited when the users' need it. The experimental results demonstrate that the proposed approach outperforms competing methods while closely tracking the simulation models, thus paving the way for smarter APs. [ABSTRACT FROM AUTHOR]