New energy enabled cell-free massive-mimo high energy efficiency access point selection and power allocation method

Abstract

The application discloses a new energy enabled large-scale MIMO high-energy-efficiency AP selection and power distribution method without cells, which comprises the following steps: equipping part of APs with new energy collection devices, and establishing a system model of new energy enabled large-scale MIMO without cells; under the premise of random EH mode and the same QoS constraint, establishing an electric energy minimization problem of joint optimization of AP selection and power distribution, and converting the problem into an MISOCP problem; replacing the binary AP selection variable by the L0 norm of the L1 norm of the power distribution coefficient correlation vector of the AP, and based on the non-convexity of L0 norm optimization, using a sparse enhancement iterative optimization algorithm based on reweighted L1 norm relaxation to solve the optimization problem; further, a DRL driven adaptive AP selection strategy is proposed, which dynamically coordinates AP selection under random EH mode and QoS change, and then the result is brought into the original optimization problem to avoid combination complexity. The application has the performance of approaching the optimum under the condition of greatly reducing the calculation overhead, significantly saves energy consumption, and meets the demand of future network for low power consumption.

Images