Modified DEEC: A varying power level based clustering technique for WSNs
Twinkle Tiwari, Nihar Ranjan Roy · 2015
Wireless sensor network (WSN) consists of large numbers of small battery powered sensing device called sensor node, having the ability to sense the data from environment of deployment. Limited energy is of prime concern for the WSNs, as source of energy is finite. Sensor nodes are used to periodically sense the environment, do limited processing before transmitting the sensed data to their cluster heads (CH) and periodically participate in the cluster head selection process. In the sensing field some member nodes are very far away from the CHs, which mean that they need more energy than the other member nodes in transmitting the sensed data. Long distance transmission and responsibility as a CH leads to early death of farther nodes there by reducing the network life. This paper presents energy efficient clustering scheme, where nodes do not transmit in full range instead have discrete power levels for transmissions. Thus transmitting nodes minimize energy wastage in transmission beyond the region of interest and avoid overhearing by most of the nodes. Intra-cluster communications are performed at lower power level and only those cluster heads are allowed to compete for cluster head selection, which have residual energy above a threshold level. We implemented this using DEEC protocol (Distributed Energy Efficient Clustering Algorithm) and called the new implementation as Modified DEEC. The proposed technique successfully conserves energy and prolongs the life time of the network in comparison to DEEC by 12.5%. There is also a substantial increase in the number of packets being transferred to base station, by a factor of 193%.