Throughput Maximization for Multi-Cluster WPCN: User Cooperation, Non-linearity, and Channel Uncertainty

In this paper, we study a general multi-cluster wireless powered communication network (WPCN) with user cooperation under harvest-then-transmit (HTT) protocol where the hybrid access point (HAP) as well as each user is equipped with multiple antennas. In the downlink phase of HTT, the HAP employs beamforming to transfer energy to the users. In the uplink phase, users in each cluster transmit their signals to the HAP and to their cluster heads (CHs). Afterward, the CHs first relay the signals of their cluster users and then transmit their own information signals to the HAP. The aim is to design the energy beamforming (EB) matrix, transmit covariance matrices of the users and time allocations among energy transfer and cooperation phases in order to optimize the max-min and sum throughputs of the network. The corresponding maximization problems are non-convex and NP-hard in general. We devise iterative algorithm based on alternating optimization (AO) and then the minorization-maximization (MM) technique is used to deal with the non-convex sub-problems with respect to (w.r.t.) the EB and covariance matrices in each iteration. We recast the resulting sub-problems as a convex second order cone programming (SOCP) and quadratic constraint quadratic programming (QCQP) for the max-min and sum throughput maximization problems, respectively. We also consider imperfect channel state information (CSI) at the HAP and CHs and non-linearity in energy harvesting (EH) circuits. Numerical examples show that the proposed cooperative method can effectively improve the achievable throughput in the multi-cluster wireless powered communication under various setups.