A Graph Based Clustering and Preconditioning of V-MIMO Wireless Sensor Networks

Abstract

This paper presents a graph based methodology for increasing the channel capacity of Virtual-Multiple Input Multiple Output (V-MIMO) defined over a Wireless Sensor Network (WSN). A fully connected graph $\mathcal{G}(\mathcal{V},\ \mathcal{E},\ \mathcal{W})$ is defined for a WSN. Then, we propose a new clustering algorithm based on the Fiedler vector of the graph $\mathcal{G}$ which divides the sensor nodes $\mathcal{V}$ into twoclusters (transmitting and receiving antennas). The links between these two clusters results in V-MIMO network. Next, a Modified Maximum Spanning Tree Search algorithm (MMASTS) is proposed on V-MIMO to enhance the average channel capacity. Simulation performance of average channel capacity and uncoded Bit Error Rate (BER) are plotted using different precoding techniques like Zero Forcing (ZF) and Minimum Mean Square Error (MMSE). These are also used for comparing the performance of proposed Fiedler vector based clustering with $k$- means clustering.