Developing RSSI-based clustering algorithm in wireless sensor networks

Wireless Sensor Network (WSN) can be defined as the spreading of a large number of sensor nodes (SN) in regular or irregular forms to the area and transmitting the data they perceive to the center by hopping over each other. Sensor nodes forming wireless sensor networks; They are units with low cost, limited energy and with a sensor unit. Since the transmission distance of data in wireless communication is directly proportional to the energy consumed, the distance must be shortened. For this reason, data transmission in WSN is preferred as multi-hop. In a multi-hop structure, SNs broadcast their data to the network to transmit their data to the base station (BS) and close nodes receive and broadcast this broadcast on them. In this way, the process continues until it reaches the base station. However, this situation brings with it many problems such as collision, overlap or overload of nodes. Minimizing these problems and transmitting data to the BS with the lowest energy cost is the main purpose of WSN applications. In this thesis, a new routing algorithm is proposed that increases the energy efficiency of the network and aims to eliminate unnecessary data load. The proposed Center Oriented Clustering Algorithm (CCA) is based on the logic of separating the nodes into clusters and selecting a leader and transferring their data to the BS through this leader. CCP was analyzed comparatively with similar algorithms in the literature in terms of energy efficiency, packet numbers, coverage and surviving node numbers. As a result, it was observed that MEKA achieved 30% improvement in the number of surviving nodes, 18% in the number of packets sent, 58.7% in terms of total energy remaining in the network, and 13% in terms of coverage area according to LEACH and approached the coverage area performance in the k-means algorithm.