Key points are not available for this paper at this time.
The demand for high traffic in the Fifth Generation (5G) and beyond mobile communication networks has resulted in intense competition for the limited licensed bands, prompting cellular networks to operate in the unlicensed band. This creates interference with WiFi the most popular radio access technology operating in the unlicensed band. However, the current solutions aimed at addressing this issue primarily concentrate on the LTE, which results in degradation of WiFi performance. To investigate this problem, we conducted experiments to analyze the performance degradation of a WiFi network in the presence of LTE-U. In our paper, we have proposed a mechanism wherein the WiFi Access Points (APs) are enabled with Q-Learning (QL) algorithm. The QL has transformed WiFi networks into intelligent systems, enabling each access point (AP) to autonomously choose an interference-free channel while also coordinating with LTE-U nodes. This advancement has made WiFi networks smarter and more efficient. The channel selection is based on detecting mid and high interference levels by WiFi APs. We then compare the performance of the proposed scheme w.r.t. the existing benchmark scheme and observe significant improvement in WiFi network throughput, without impacting LTE-U. This paper also demonstrates fairness of the proposed scheme on the competing WiFi users.
Mishra et al. (Sat,) studied this question.