Search

Summary: Designing an energy efficient and durable wireless sensor networks (WSNs) is a key challenge as it personifies potential and reactive functionalities in harsh antagonistic environment at which wired system deployment is completely infeasible. Majority of the clustering mechanisms contributed to the literature concentrated on augmenting network lifetime and energy stability. However, energy consumption incurred by cluster heads (CHs) are high and thereby results in minimized network lifetime and frequent CHs selection. In this paper, a modified whale‐dragonfly optimization algorithm and self‐adaptive cuckoo search‐based clustering strategy (MWIDOA‐SACS) is proposed for sustaining energy stability and augment network lifetime. In specific, MWIDOA‐SACS is included for exploiting the fitness values that aids in determining two optimal nodes that are selected as optimal CH and cluster router (CR) nodes in the network. In MWIDOA, the search conduct of dragon flies is completely updated through whale optimization algorithm (WOA) for preventing load balancing at CHs. It minimized the overhead of CH by adopting CHs and CR for collecting information from cluster members and transmitting the aggregated data from CHs to the base station (BS). It included self‐adaptive cuckoo search (SACS) for achieving sink mobility using radius, energy stability, received signal strength, and throughput for achieving optimal data transmission process after partitioning the network into unequal clusters. Simulation experiments of the proposed MWIDOA‐SACS confirmed better performance in terms of total residual energy by 21.28% and network lifetime by 26.32%, compared to the competitive CH selection strategies. [ABSTRACT FROM AUTHOR]

Summary: For the past 5 years, localization has become an emerging area in the field of wireless sensor networks (WSNs), which can be applicable for diverse applications, such as outdoor environments and monitoring of objects located in indoor environments. The major constraint in localization is allocating a location for each node as diverse sensor nodes are employed to retrieve the information in WSN. Recent research works in node localization have generally moved by multihop range‐free localization approaches for achieving high accuracy through reducing the localization error among the actual and estimated position of nodes. The conventional approaches are aimed to improve the localization accuracy without considering efficiency concerned with algorithm's convergence time and energy costs. The main intent of this paper is to model the WSN localization model, which focuses on the localization of anchor nodes concerning the target nodes. The multiobjective function is developed by considering the distance function, received signal strength (RSS), and energy. Here, a hybrid metaheuristic algorithm named hybrid cuckoo–red deer algorithm (HC‐RDA) with the integration of red deer algorithm (RDA) and cuckoo search algorithm (CSA) is adopted for localization of unknown nodes. The fitness function of the developed model is the minimization of the derived multiobjective function. Both simulation studies and theoretical analysis demonstrate that the proposed approach can improve localization performance with a better convergence rate, thus ensuring high localization accuracy compared to conventional metaheuristic models. [ABSTRACT FROM AUTHOR]

Summary: Energy‐efficient clustering and routing are two well‐known optimization problems, mainly employed to achieve energy efficiency and maximum network lifetime in wireless sensor networks (WSNs). The clustering and routing processes can be considered as an NP‐hard problem, and metaheuristic algorithms can be applied to resolve it. In this paper, a dynamic clustering and process protocol based on multi‐objective particle swarm optimization with Levy distribution (MOPSO‐L) algorithm. Since the parameters in WSN are related to one another, multi‐objective parameters should be included in the process of cluster head selection and routing. The proposed MOPSO‐L technique is presented for organizing the clusters and CH chosen by merging consolidated and shared models. The MOPSO‐L algorithm incorporates the benefits of PSO algorithm along with the merits of Levy distribution to escape from trapping into local optima. The presented model undergoes comparison with existing techniques under three different scenarios based on the location of the BS with respect to average energy consumption, number of data transmission, and network lifetime. The experimental outcome reveals that the proposed model attains extended network lifetime as well as efficient energy over its comparatives. [ABSTRACT FROM AUTHOR]