Deep-Learning Based Auction-Driven Beamforming for Wireless Information and Power Transfer

In this paper, we design a deep learning based resource allocation framework, in the form of an auction, for simultaneous information and power transfer from hybrid access point (AP) to devices energy harvesting devices, respectively. Using Myerson’s lemma concept virtual welfare maximization, develop optimal dominant-strategy incentive-compatible mechanism AP maximize its expected revenue, on devices’ bid profiles, valuation distributions, demand channel state information. so doing, formulate revenue maximization problem, which is mixed-integer non-linear program, propose efficient Branch-and-Bound (BnB) algorithm solve problem using semidefinite relaxation technique each branch. Since has exponential time complexity, BnB algorithms can be impractical real-time applications. To circumvent this, neural network (DNN) proposed, trained predict beamforming data towards We use offline populate training dataset. The proposed DNN architecture indeed multi-layer perceptron, well map heterogeneous input desired output with high accuracy. Furthermore, heuristic iterative solution whose accuracy performance comparable that DNN-based solution. polynomial complexity whereas linear complexity.