Title: Smart Scheduling and Feedback Allocation over Non-stationary Wireless Channels

Abstract: It is well known that opportunistic scheduling algorithms are throughput optimal under dynamic channel and network conditions. However, these algorithms achieve a hypothetical rate region which does not take into account the overhead associated with channel probing and feedback required to obtain the full channel state information at every slot. In this work, we design a joint scheduling and channel probing algorithm by considering the overhead of obtaining the channel state information. We adopt a correlated and non-stationary channel model, which is more realistic than those used in the literature. We use concepts from learning and information theory to accurately track channel variations to minimize the number of channels probed at every slot, while scheduling users to maximize the achievable rate region of the network. More specifically, we employ Gaussian Process Regression that enable us to predict future channel state and also provide a way to measure uncertainty in the learning and prediction. Simulation results show that with the proposed algorithm, the network can carry higher user traffic. The proposed solution can be utilized in more complex and heterogeneous environments such as 5G and beyond networks where the model free approach is needed.