Open AccessModified threshold for cluster head selection in WSN using first and second order statistics
Author(s) -
Panda Sefali,
Behera Trupti Mayee,
Samal Umesh Chandra,
Mohapatra Sushanta Kumar
Publication year - 2020
Publication title -
iet wireless sensor systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.433
H-Index - 27
ISSN - 2043-6394
DOI - 10.1049/iet-wss.2020.0048
Subject(s) - cluster analysis , selection (genetic algorithm) , wireless sensor network , computer science , throughput , energy (signal processing) , statistics , cluster (spacecraft) , efficient energy use , base station , data mining , mathematics , engineering , wireless , artificial intelligence , computer network , telecommunications , electrical engineering
Wireless sensor network (WSN) comprises of numerous sensors deployed either directly or randomly in the region of interest. Due to the limited power of the sensors, these networks are energy‐constrained and thus need efficient power utilisation. Efficient clustering and cluster head (CH) selection ensures balanced energy distribution to the sensors within the WSN and hence prolong the network lifetime. This study proposes the method for evaluating the threshold for the CH selection after each round, which increases the network lifetime and throughput significantly. The threshold for CH selection is modified considering the normalised first‐order and second‐order statistical parameters, such as mean average low‐energy adaptive clustering hierarchy (AvgLEACH) and variance (VarLEACH) of the overall network energy. These methods have been formulated after studying the effect of the number of working nodes in each round on the threshold value selection. Apart from including energy parameter to the threshold equation, the methods of VarLEACH and AvgLEACH are augmented with a residual energy parameter that is local to the nodes and named as VarRLEACH and AvgRLEACH. The simulation results comparing all the methods suggest that the proposed method AvgRLEACH outperforms LEACH by a factor of 1.5 in delivering data to the base station and outlives the network driven by LEACH protocol by 30–40%.