Dynamic Power Allocation and Virtual Cell Formation for Throughput-Optimal Vehicular Edge Networks in Highway Transportation
In this paper, we address highly mobile vehicular networks from users' perspectives in highway transportation. Particularly, we consider a centralized software-defined environment in which centralized resources can be assigned, programmed, and controlled using the anchor nodes (ANs) of the edge servers. Unlike the legacy networks, where a typical user is served from only one access point (AP), in our proposed system model, a vehicle user (VU) is served from multiple APs simultaneously. While serving a VU from multiple APs increases the reliability and the spectral efficiency of the assisted users, an accurate power allocation has to be maintained for each of the transmission time slots. Therefore, it is essential to serve the users with the optimal power level of the APs. As such, we jointly formulate user association and power allocation problems to achieve enhanced reliability and weighted user sum rate. However, the formulated problem is a difficult combinatorial problem and thus, it is remarkably hard to solve. Therefore, we use fine-grained machine learning algorithms to efficiently optimize joint user associations and power allocations of the APs in a highly mobile vehicular network. We introduce a distributed single-agent reinforcement learning algorithm, namely SARL-MARL, which obtains nearly identical genie-aided optimal solutions within nominal number of training episodes than the baseline solution. Simulation results validate that our proposed RL approach outperforms existing schemes and can attain genie-aided optimal performances.
READ FULL TEXT