Counter propagation ANN based massive adaptive array antenna beamforming for mmWave communication system

Next-generation cellular standard will use millimeter wave (mmWave) frequencies to provide higher throughput. The base station will use massive array antenna to ensure better transmission to multiple users. We propose a power and an adaptive beamforming algorithm based on Counter Propagation Artificial Neural Networks (CP-ANNs). The hidden layer is a Kohonen network which categorizes the input pattern and the output layer reproduces the correct output pattern for the category. An optimization problem has been formed that adapts the angle of arrival of array antenna to maximize the throughput by saving node energy. With the partial channel state information (CSI), the proposed antenna allocate different beam widths for multiple users for reducing the interference between users and thereby improve user capacity. Simulation results show the efficiency of the proposed scheme.