Developing A New Scheduling Algorithm for Wi-Fi 6 Technology Based on Machine Learning

Abstract

Wireless mobile communications have experienced tremendous growth in the number of users, data rate requirements, and coverage in recent years. As the data rate and system throughput requirements increase, researchers and system designers need to develop efficient methods to meet these requirements with reasonable effort and cost. In this paper, we discuss an efficient approach to deal with diminishing the overhead of Downlink (DL) and feedback using Machine Learning (ML) for 802.11ax. In particular, the antennas that are related to the router were divided into two gatherings. We used good samples of Channel State Information (CSI) that were taken from an open-access dataset and used it to train our linear regression model. The first group of antennas was used as input to our model and the second group was used as the output of the model. In the online mode, we need to estimate only one group of antennas and for the second group, we can predict it using the trained linear regression model by using the estimated CSI group as input to the model. Therefore, the output of the model using as the CSI for the second group. In this way, we can reduce the overhead of the DL in the router as shown in the result table so the router will work more efficiently compared to the existing systems. From the results table in the last section, the average sum rate has increased between 20% and 30%.

Energy-Saving Data Acquisition Model of Wireless Sensor Network Based on Nonlinear Algorithm

“…Four performance parameters, including the number of model adjustments, data acquisition, data transmission, and mean absolute error, were compared and analyzed through Matlab simulation experiments [9]. In this paper, a miniature wireless sensor network monitoring device is established to verify the prediction accuracy of the method for temperature sampling data with time series correlation.…”