Energy-Efficient Resource Allocation For MU-MIMO Systems With Limited Feedback

Multi-user multiple–input multiple-output(MU-MIMO)system is a promising technology for the next generation wireless communication networks because of its high system throughput.By employing limited feedback techniques at the transceivers,the receiver knows the channel state information(CSI)of the transmitter,which makes resource allocation effective and thus improves the performance of the wireless systems.In recent times,energy consumption has increased rapidly.To conserve energy and meet the requirements for green communications in the next generation wireless networks,we emphasis on the energy-efficient resource allocation optimization for MU-MIMO systems.Our research objective is to optimize the system energy efficiency in MU-MIMO systems by considering limited feedback of CSI through resource allocation.We first investigate the energy-efficient unequal power allocation for downlink single cell MUMIMO systems with limited feedback.The energy-efficient resource allocation is formulated as a non-convex problem.A low-complexity optimal algorithm based on Lagrange dual decomposition method is proposed to assign power to the users.We consequently study energy-efficient joint user selection,power allocation,and number of antennas optimization for energy efficiency improvement for downlink single cell MU-MIMO systems with limited feedback.The formulated problem is mixed-integer nonlinear programming(MINLP)and non-convex and based on Lagrange dual decomposition and successive convex approximation SCA),a practical algorithm is proposed.Furthermore,we examine energy-efficient joint user selection,power allocation,and subcarrier allocation optimization with limited feedback for the MIMO-OFDMA systems.The formulated problem is NP-hard and non-convex.An energy-efficient iterative algorithm is proposed by utilizing modified Kuhn-Munkres algorithm and Lagrange dual decomposition to obtain optimal energy efficiency.Simulation results indicate that the proposed algorithms can converge with a few iterations and can attain higher energy efficiency as compared with the existing algorithms.