Networked MIMO with Fractional Joint Transmission in Energy Harvesting Systems

摘要

This paper considers two base stations (BSs) powered by renewable energyserving two users cooperatively. With different BS energy arrival rates, afractional joint transmission (JT) strategy is proposed, which divides eachtransmission frame into two subframes. In the first subframe, one BS keepssilent to store energy while the other transmits data, and then they performzero-forcing JT (ZF-JT) in the second subframe. We consider the averagesum-rate maximization problem by optimizing the energy allocation and the timefraction of ZF-JT in two steps. Firstly, the sum-rate maximization for givenenergy budget in each frame is analyzed. We prove that the optimal transmitpower can be derived in closed-form, and the optimal time fraction can be foundvia bi-section search. Secondly, approximate dynamic programming (DP) algorithmis introduced to determine the energy allocation among frames. We adopt alinear approximation with the features associated with system states, anddetermine the weights of features by simulation. We also operate theapproximation several times with random initial policy, named as policyexploration, to broaden the policy search range. Numerical results show thatthe proposed fractional JT greatly improves the performance. Also, appropriatepolicy exploration is shown to perform close to the optimal.