K-Means Cluster-Based Interference Alignment With Adam Optimizer in Convolutional Neural Networks

Abstract

In an interference channel, IA (interference alignment) yields exquisite channel state data and uncorrelated channel components and gains high DoF (degrees of freedom). This paper proposes the clustering predicated interference alignment with the neural network. Here Adam Optimizer utilized for signal optimization and K-means clustering in which it is utilized in clustering the minuscule cells with the base station and utilizer in the heterogeneous network based on MIMO mmWave. The neural network used here is a convolutional neural network (CNN) which is integrated with the Adam optimizer. The experimental results consider the parameters particularly DoF, spectral efficiency, energy efficiency, signal to interference noise ratio (SINR), and computational complexity. While considering energy efficiency, spectral efficiency, and maximum DoF, simulation results betoken proposed method procures better performance when compared to classical methodology.