An energy aware modified PAM clustering with cluster head modulation in Mobile cloud sensor environment

Energy conservation is one of the major concerns in mobile cloud sensor environment, where mobile sensing devices are capable of acquiring the sensing information and send the collaborated data to cloud hosted applications. However, they are constrained by the limited computing power and the short battery life. Hence, sensor based applications need an effective scheme to collect the sensed information and offload it to the base station in an efficient manner. In this paper, an approach named Modified Partitioning Around Medoids clustering with Threshold based Cluster Head Replacement (MPAM-TCHR), is proposed that aims for better clustering and change of cluster heads for every round of data offload to reduce the energy utilization of the sensor nodes. The proposed work divides the process into four phases namely, initialization, clustering, Cluster Head (CH) formation and transmission phases. Initially, node initialization is performed, and then the modified PAM clustering algorithm accurately estimates the medoid points based on which optimal cluster set partitioning is done. Next, the four parameters namely residual energy, signal to noise ratio (SNR) of the sensors, average path loss between the node and the other nodes and the path loss between the sensor and the base station influence the CH selection. This involves a threshold based modulation measure to identify probability effective node to become CH, iteratively changing the CH at different time intervals. Ultimately, the CH offloads the collected information to the base station (BS) with low communication cost possible during a particular time interval. Experimental results show better performances in terms of metrics like network lifetime, network utilization, and the average number of alive nodes. Thus, the proposed method performs data transmission with low energy consumption and thereby able to prolong the network lifespan.