Abstract
Researchers have developed a range of methods and strategies to decrease wireless sensor network energy consumption. Mote clustering is one of the competent topological control approaches to boost the networks’ energy efficiency, scalability, and performance. Energy is dissipated during the cluster creation, cluster head selection, routing from the head and base stations of the cluster, and data aggregation—clustering and routing emphasis on the stability and the longevity of the network. This research work provides the optimization technique for the wireless sensor network to optimize the energy through NN-LEACH. The main goal is to extend network life and reduce power consumption by clustering and routing sensor nodes using the two-step NN-LEACH protocol, which is suggested. An additional goal might be to establish the appropriate course of action for the suggested approach for this network.
1. Introduction to Wireless Sensor Networks
Wireless sensor networks (WSNs) offer huge potential to improve people’s lives and study the environment using sensor nodes. The key obstacle in using these motes is independent and low-powered devices with small batteries. According to the NSF, new technologies lower motes and sensor arrays’ cost, size, and weight. Integration approaches reduce distance barriers by increasing performance and longevity while decreasing size [1]. WSNs are typically hidden and rely on small batteries for power. Changing batteries is not an option. Managing the power requirements of a wireless sensor network is difficult. Contrary to popular belief, energy consumption is not determining network efficiency [2, 3]. Figure 1 shows the example of the wireless sensor network [4].
Researchers focused on layer-based components, expecting that changes in one layer would immediately influence the whole system, but it did not work. Many energy-saving models consider sending and receiving data, ignoring other factors. Most contemporary energy reduction models ignore rest parameters [5].
2. Literature Review
The energy-efficient routing algorithm is categorized as communication architecture, network structure, reliable routing, and topology-based routing [6]. These protocols might be flat or hierarchical [7]. In the WSN, flat routing algorithms do comparable responsibilities. Smaller networks frequently use these networks. Algorithms such as SPIN and directed diffusion are examples of balanced routing algorithms. The cluster head rotates to balance the energy consumed. The hierarchical routing algorithm’s architecture is efficient and scalable. Each network mote receives the data via the SPIN protocol. The data of each node are comparable to those of its neighbor’s. This protocol disseminates information to all motes when the user does not require data to be transferred between motes.
In 2000, Heinzelman et al. [8] introduced the first hierarchical method, LEACH (“low energy adaptive clustering hierarchy”). A strategy for geographically segmenting a system into smaller cells is suggested by Naghibi and Barati [9]. Each cell can contain both single-hop and multihop cells. A novel EGPRM is being used to collect data from sensor nodes using two portable sinks. EECS [10], HEED [11], PEGASIS [12], TEEN [13], LEACH-IACO [14], and T-LEACH [15] are all based on LEACH. LEACH and its derivatives aim to improve energy efficiency through coverage, data aggregation, data protection minimal latency, resilience, and scalability, and the main goal of these algorithms is to save energy. Table 1 compares various LEACH variants.
3. Network Model
The nodes are chosen at random. The starting energy of a node is E0. The MS moves along the y-axis. We know that the base station has unlimited power and that sensor nodes and sinks are everywhere. Now, we have setup and steady phases.
3.1. Step-Up Phase
The setup phase is divided into three stages: the Task Presentation (TP), the Selection of cluster head (CH), and rendezvous node (RN). Clustering, i.e., creating clusters and election of the cluster head are carried out during the next part of the setup phase [17, 18]. After the cluster formation, the last stage is known as scheduling (S). In the entire scheduling phase, the message is disseminated from the cluster head to every cluster member. Every node itself arranges their organization in the period of transmission.
3.2. Task Ordination
In this phase, firstly, RNs are selected. Initially, all nodes are assumed to be normal nodes. Every hub themselves choose whether they meet RN condition or not. To become a rendezvous hub, the hubs must fulfil a requirement. The situation representing RN is specified below the following equation:
Following points are needed to be considered to get the cluster head using the method of HNN:(a)Prepare weights , where is element of class c exemplar(b)Put on input to the desired outputs (c)Recapitulate unless the system converges , where is hard restraint
3.3. Ant Colony Optimization (ACO)
WSN performance is mainly measured via routing. Routing is a method of moving data from one place to another. While routing, two ideas are defined [17]. The first is optimum routing, where the quickest path is found using various methods, and the second is internetwork, where packets are transmitted.
3.4. Proposed PSO (Particle Swarm Optimization)
Consider a situation where birds are randomly put in an area with only a piece of food to seek. The birds do not know where food is. They just know food distance. The easiest way to obtain food is to follow the bird to it. A particle is a bird. The particles have a fitness function and a velocity that can be computed. Every particle updates the “best” two values, pbest and gbest. Lesser particles update their locations and velocity using equations (2) and (3).
3.5. Nondominated Sorting Genetic Algorithm
In the last several years, EAs (evolutionary algorithms) t have evolved to solve mono-objective, multi-even-handed, and multitarget enhancement concerns in a precise order [18]. Despite explicit efforts to combine various mono-target developmental and nondevelopmental calculations, there are few studies that include all three types of enhancement concerns. The Pareto technique is used to segment the incoming population into subpopulations [19, 20]. Figure 2 shows the flowchart of hybrid ACO/PSO.
4. Simulations
Simulation is carried out using MATLAB for easy results computation. The limit limiting the framework’s display has been disregarded. For reproduction, we dissect findings by several nodes and then by region. Case 1: the proposed NN-LEACH is compared with existing RZ-LEACH and ACO-RZLEACH based on the node scalability by a varying number of motes and adjusting of the threshold values of various parameters such as alive nodes, dead nodes, the packet sent to BS, packet sent to CHs, and energy remained. Case 2: in this case, the adaptability issue is considered for the organization’s differing size for the consistent number of hubs (n = 100). The presentation of the present framework has been contrasted and existing one for the accompanying local sizes, for example, 50 m by 50 m, 100 m by 100 m, …, to 500 m by 500 m against the number of rounds.
5. Results Comparison of Leach and Optimized Leach
The proposed hybrid model is compared to the basic LEACH [8] and the optimized LEACH [17] with node scalability. Table 2 shows the simulation model’s initial settings. Motes placed randomly in the area of interest. As demonstrated in Figure 3, assuming that all of the nodes have the same initial energy level (0.3 J), the energy optimization of the enhanced LEACH is compared to LEACH and optimized LEACH. Packets sent from the cluster head to the base station and residual energy are observed after round numbers 100, 120, 140, …, 400 (0 : 3 J), and we compare their leftover energy in Figure 3. Table 3 shows the simulations results for energy remaining, data/communication packets and finally packets.
(a)
(b)
(c)
(d)
(e)
(f)
6. Conclusion
As shown in the graphs of Figure 3, LEACH has better outcomes in terms of initial, tenth, and whole nodes dead for the specified population. Proposed LEACH also has higher residual energy and sends more packets to the base station and cluster head. All of these factors lower energy usage and so extend network lifetime. The number of packets transmitted from the cluster head to the base station determines performance. The WSN lifetime can be balanced extended by the NN-LEACH algorithm that has been presented. In three different network settings, experimental results demonstrate that NN-LEACH surpasses its competitors GAECH, GCA, EAERP, and LEACH. Most real-time applications can benefit from the performance gain when the base station is removed from the network. In the best-case scenario, where the base station is not linked to the network, NN-LEACH exhibits a significantly longer lifetime than its equivalent. Additionally, we discovered that LEACH optimization reduces energy use by distributing it evenly among clusters by 5%.
Data Availability
The data used to support the findings of this study are available upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.