International Journal of Computer Applications (0975 – 8887) Volume 88 – No.7, February 2014 14 Cluster based Energy Efficient Sensing for Cognitive Radio Sensor Networks Usman Mansoor Department of Telecom Engineering ICT Islamabad affiliated with UET Peshawar Muhammad Khalil Shahid Associate Professor ICT PTCL Academy Islamabad ABSTRACT In this paper, a wireless cognitive radio sensor network is considered, where each sensor node is equipped with cognitive radio. As energy consumption is the main problem when using sensors therefore a new clustering algorithm is developed according to which group of nodes form cluster having a single cluster head. Each cluster has balanced energy which prolongs overall lifetime of CRSN. Cluster heads are rotated, depending on a threshold value, in such a way as to improve the lifetime of a cluster. As new cluster head is selected immediately whenever energy of old cluster head drops to certain threshold thus improves sensing results by CRSN nodes with minimum number of faulty decisions. Simulation results demonstrate working of schemes proposed and compares the pros and cons of each scheme. Keywords Balanced energy, clustering algorithm, cluster head and cluster head rotation. 1. INTRODUCTION Cognitive Radio Sensor Network has recently attracted a large amount of attention due to its advantage of dynamic spectrum access which provides reliability in terms of communication and enhances energy conservation potential in sensor network. Each node in a [1] [2] CSRN is typically equipped with a sensors having cognitive capability for communication purpose. Preserving the consumed energy of each node is an important goal that must be considered when developing any protocol for cognitive radio sensor networks. To satisfy these requirements, different solutions have been proposed in the past that exploit the tradeoffs among energy, accuracy, and latency. Among many solutions one of the solutions uses hierarchical architectures [3], where sensor nodes equipped with cognitive radios are clustered according to application- specific parameters. Sensors in cluster networks can then co- operate to sense and process a physical phenomenon. Clustering of sensors [4] helps in scalability of network and conserves energy to a large extent. Much of work in this regard has been done. In a centralized approach every sensor sends data to the base station and thus consumes large amount of energy as the transmission distance is much larger. Using clustering approach, firstly data is transmitted to cluster head of each cluster and then that data is further transferred to the base station. Therefore much of the work load of base station distributes between cluster heads. Other than scalability [5], clustering helps in reducing routing table information. An important feature of sensor network is energy efficiency and to extend the network’s lifetime, battery source of each sensor in a network should be intelligently used so that minimum amount of energy consumption is made. Fig.1 shows general architecture of clustered network. In this figure there are two clusters i.e. cluster 1 and cluster 2. In each cluster circles represent non cluster head nodes and those circles which are connected to the base station are the cluster head nodes. Cluster head nodes get data from their respective non cluster head nodes and then pass it to base station. Base station actually is the main centralized entity where all the data is present. Fig 1: Hierarchical Architecture The aim of this study is to investigate clustering algorithm which improves the energy balancing between the clusters and also the lifetime of each sensor node. Specifically initial clustering is done on the bases of distance criterion so that cluster head uniformity is maintained throughout the network. After the initial clustering, in second stage of the algorithm de attachment of some sensor nodes occur throughout the network. This second stage helps in energy balancing throughout the network. In the third stage of the algorithm cluster head selection is done on the bases of energy. Therefore paper is organized in a way that section 2 describes a general framework of CRSN. The optimal method of solving this problem i.e. proposed clustering algorithm is presented in complete detail in section 3. In the section 4 results of some preliminary experiments are presented and section 5 concludes the paper.