An analysis of optimal capacity in cluster of CDMA Wireless Sensor Network

摘要

Interference is one of the factors influence the number of active sensors of the CDMA Wireless Sensor Network as it affects E_b/I₀ on the received signal. The higher the interference, the smaller the capacity is. One of the methods to minimize the interference is cluster formation in the CDMA Wireless Sensor Network. Using cluster, the interference can be minimized because transmission power to send data decreases as the sensor only transmits data to Cluster Head (CH), not directly to sink. There are two types of the interference on the CDMA Wireless Sensor Network, namely: intra-cluster interference (Multiple Acces Interference/MAI) and inter-cluster interference (Node Interference/NI). The intra-cluster interference is overcome by using power control so that the received power in the Cluster Head from every sensor is the same. However, the Cluster Head is not able to control sensor's transmission power from other clusters, so interference from other cluster cannot be managed. The research analyzes the factors influencing inter-cluster interference comprising of distance between clusters, cluster size, influence of sensors located near cluster's border and the network capacity. The research shows distance and position of sensors will impact to inteference. If the distance between clusters rises 1.5 times, then it will reduce the power of interference signal by 8.6 dB. And the sensors at the border generate higher interference than other sensors. This problem is overcome by using layering technique, in that high power sensors at the border will send data through intermediary sensor. As a result this methode can reduce transmission power of the sensors at the border. Therefore, applying layering technique is capable of reducing interference of the sensors at the border. The research shows that if the layering radius is 0.7 of cluster size, it will generate minimum inter-cluster interference. Finally utilizing the cluster formation will increase the network capacity. If the network applies the unequal number of sensors in the cluster, it will increase 3% capacity compared to the equal number of sensors in the cluster.