Incorporate ACO routing algorithm and mobile sink in wireless sensor networks

Today, science and technology is developing, particularly the internet of things (IoT), there is an increasing demand in the sensor field to serve the requirements of individuals within modern life. Wireless sensor networks (WSNs) was created to assist us to modernize our lives, saving labor, avoid dangers, and that bring high efficiency at work. There are many various routing protocols accustomed to increase the ability efficiency and network lifetime. However, network systems with one settled sink frequently endure from a hot spots issue since hubs close sinks take a lot of vitality to forward information amid the transmission method. In this paper, the authors proposed combining the colony optimization algorithm ant colony optimization (ACO) routing algorithm and mobile sink to deal with that drawback and extend the network life. The simulation results on MATLAB show that the proposed protocol has far better performance than studies within the same field.


INTRODUCTION
The WSN is divided into clusters, each cluster has a cluster head (CH). The sensors nodes within a cluster will transfer data to their CH. cluster heads (CHs) store the information then send the data to the sink (base station-BS) which can be either fixed or portable and can be recharged. However, the sensor nodes are limited in the ability to store and process the information [1], [2]. The further the sensor nodes are away from the sink, the more energy they will consume to forward data to the sink but the WSN still needs to maintain its life time. Therefore, to reduce the energy consumption, we will make the sink mobile to collect the data. Then, determine the path of the sink to the CHs [3]- [7]. The proposal presents an algorithm to balance the energy and optimize the life time of the WSNs by combining low energy adaptive clustering hierarchy cluster (LEACH-C) protocol with ACO algorithm that find the shortest path for mobile sink to CHs to collect information. The CHs will be selected based on the LEACH-C [8], [9]. After clustering, each cluster will have a CH (The coordinates of the sink and the CHs (xd, yd) are saved in an array). The simulation results show that the improved method extends the average life time of the sensor nodes thereby reduce the energy consumption of the system. The article includes the following sections: Section 1 introduction, section 2 related works, section 3 proposed algorithm, section 4 simulation and assessment of the performance of the proposed solution, section 5 conclusions.

LEACH-C
In order to overcome the disadvantages of traditional LEACH, LEACH-C [8], [9] was created and appreciated for its applicability. The selection of the cluster head at LEACH-C is done by the sink while LEACH chooses the cluster head randomly. With LEACH-C, the sink will rely on the information sent from nodes for clustering and cluster head selection. Moreover, the sink can add parameters, indicate the energy reserve of the node and use optimal algorithms to cluster and select the cluster head. The cluster is formed when the base station selects the appropriate cluster head which sends its ID to broadcast to nodes in the network. Each node will check if its ID is the same as the ID in the broadcast. If the same, the node will host the current cluster, otherwise will wait for broadcast information from the host. The clustering is more efficient, the energy is distributed evenly, the nodes in the network use the energy better. Through the location and information from the sensor nodes, cluster formation is more optimal. Collecting and transmitting data is performed continuously by controlling the number of clusters in the network. However, LEACH-C requires a large number of input parameters to select the cluster head, sending location information and energy status will consume a lot of energy if the distance from the nodes to the sink is quite large.

MSA
In WSNs, the sensor nodes are responsible for sensing, processing, and communicating. In particular, communication is the most energy consuming. To improve the life of the network, each protocol must have its own strategy, in which routing is one of the strategies that play an important role. Mobile sinks have been proposed with efficient extended simulations through the mobile sink assisted (MSA) algorithm [19]. Assuming that the sensor nodes are randomly allocated, the mobile sink can navigate the network to collect data and move to a fixed sink. Assuming a path for the portable sink, there are no obstacles and the communication range is high. Suppose that there is a fixed sink and four mobile sinks have infinite energy. For synchronized movement, the portable sinks move from the center to the four corners of the rectangular sensor area, where the mobile sink movements are determined by the fixed sink. For extended simulation results, the range of sensor buttons, initial energy, usable area, etc. can be changed. By using a mobile sink, the transmission distance is reduced, thereby significantly reducing the energy consumption on the line. However, the main drawback is that due to the probability CH selection, the energy level of the nodes when selecting the cluster is not considered. And this is the reason for the reduced performance of the MSA algorithm [20].

LEACH and a mobile sink in a fixed orbit
The sensor nodes are randomly deployed in the area. The portable sink starts at a fixed position and moves along a pre-scheduled static sink [21], [22]. How to choose a cluster head: the cluster head is selected based on the LEACH. The cluster head is responsible for collecting data from the cluster member nodes, saving data and transmitting data to the mobile sink within its range. How mobile sink moves: Based on the  [22], [23]. Let the mobile sink move in a clockwise or counterclockwise direction around the edge of the data field at a given velocity, fixed path, and predicted motion. Sink signals to inform all nodes of its current position with an explicit time and a radius from the middle of the sensor area R=100 m, 75 m, 50 m, 25 m, 15 m as shown in Figure 1.

LEACH-C and a mobile sink
The objective of this protocol is to propose a hierarchical protocol based on the LEACH-C protocol by combining a mobile sink with K-mean clustering methodology to collect data to balance the consumption energy. A number of experiments have been carried out, the simulation results show that the power consumption decreases significantly and the amount of data received at the base station increases [24]. Power consumption: The purpose of the LEACH-CM mode is to reduce energy consumption to extend the life of the network. As a result, energy consumption is an important parameter during the test period, especially the average energy consumption of nodes. Network life: The author assessed the lifetime of the WSN in a specific context. This context adheres to the first node died (FND) index, which is the time elapsed until the first sensor starts to stop. The quality of the data received by BS.

Finding the shortest path for the sink movement
In this paper Ya-Qiong and Yun-Rui [25], the authors used the Dijkstra rule to seek out the shortest path from the CHs to the sink. This algorithm, reduces the formation of cluster heads, considers the distance to the BS base station, the distance to the cluster heads and also the energy of the nodes. At an equivalent time, presenting the shortest path between the cluster head and also the neighboring node, this ensures that this algorithm provides affordable communication and low power consumption owing to transmitted energy. Thus, the author has achieved the goal of reducing the energy consumption between sensor nodes in WSN.

Moving direction to save energy for sink mobility
In this paper [22]- [24], the author Ali et al. [26] studied to mix LEACH-C routing algorithm and Dijkstra algorithm to boost the energy consumption within the network. A mobile sink can move from the sink to CHs to retrieve information, the direction of the mobile sink's movement is set by the Dijkstra algorithm. The result's energy savings and increasing the network life time. The simulation results show that the proposed LEACH-CD is approximately the same as LEACH-C and superior to LEACH. But there is also a limitation when using the Dijkstra algorithm, which ignores some of the cluster head nodes and therefore some of the information collected is missing.

ACO
In 1991, Marco Dorigo introduced the ACO algorithm by observing, and drawing ideas from, the process of finding ant food sources. On the way to find food, they always leave signals to other ants to find the most effective way, the signal is the smell or pheromone. This odor may be lost or denser due to the choice of ant path. If the roads with the most ants pass, the odor concentration is highest, and the other roads the pheromones will evaporate. Finally, after a period of time they choose the path with the most pheromones are also the shortest way to move food to the nest [16], [27]- [31]. From the above behavior, the author has built artificial ant colonies, capable of leaving odors along the way, remembering the journey they have gone through and calculating the length of the path it chose. In addition to updating the smell, calculating, the ants can exchange information with each other. Thanks to this artificial colony, the author has built an ant system (AS) to solve the salesperson problem, compared to method of simulating natural simulated annealing (SA), genetic algorithm (GA), it effectively than through empirical and since then the author has developed, extensive application with common name ACO method [29], [32].

PROPOSED ALGORITHM
The author has presented some routing protocols such as LEACH, LEACH-C, MSA combined with mobile sink following the fixed orbital to improve the life time of WSNs and also point out the advantages and disadvantages of each method [24], [29], [33]- [38]. Then we proposed a new algorithm LEACH-CACO which is shown through the following main steps: − The sensor nodes are randomly deployed in an area − Clusters are formed from sensor nodes − CHs for clusters are selected based on LEACH-C algorithm: a. The sink receives the information about the remaining energy and current position of each node then calculates the average energy of WSN. b. Choose the CH, compare the average energy of the network with the remaining energy of the sensor node c. When the sensor nodes are within its communication limits, the CH is responsible for collecting and processing information from the cluster member nodes and transmitting data to the sink. − Find the shortest path between sink and CHs: a. After clustering, each cluster has a CH. b. Use the ACO algorithm to seek out the shortest path from the sink to the CHs c. The sink will move to the path to go over CHs and collect data The simulation results of the proposal will be compared with other algorithms. The drawback of the LEACH-C associated with the Dijkstra algorithm to find the shortest path is that it ignores some cluster head nodes. Our proposal will address this matter to reduce the energy consumption and increase the lifetime of the network. Figure 2 shows the flowchart of the proposal. The direction of the sink's movement varies based on the algorithm that finds the shortest path from the sink to the CHs. Mobile sink has a certain time to stop to collect data.

. Simulation settings
To test and analyze the proposed solution, some simulations were implemented by MATLAB2014a. The simulation environment is designed with parameters as shown in Table 1. The evaluation of WSN focuses mainly on the number of alive nodes, the remaining energy in the network, and the average power consumption, because there is no electricity, there is no work for the sensor inside the network.  Table 2 and Figure 3 shows the quantity of alive and dead nodes throughout the network lifespan. The primary node for LEACH, LEACH-C, LEACH-CD, and proposed protocol dies at 119, 224, 268, and 308 rounds, the proposed protocol is more than range of rounds that the primary node dies. And all nodes die at 433, 499, 519, and 531 rounds, that is above of the quantity of rounds that the last node dies. As is seen from Figure 3, proposed protocol performs higher than the other selected existing protocols in terms of network life, stability amount. Thus, the proposed protocol consumes less energy that leads to not solely prolong network lifespan. However additionally prolong stability period as compared to the other protocols remaining energy.  Table 3 and Figure 4 shows that the data sent to the BS is more for proposed protocol as compared to the others protocols. The throughput for for LEACH, LEACH-C, LEACH-CD, and Proposed Protocol are 245, 423, 436, and 459. Proposed protocol is higher than the number of packets. LEACH identifies the cluster head supported random chance whereas LEACH-C, LEACH-CD and LEACH-CACO supported the position and energy info of all nodes within the WSN sent to the sink. Particularly, LEACH-CACO features a sink that moves to the cluster heads on the trail outlined by the ACO algorithm. Simulation results show that the amount of nodes alive and the remaining energy of the LEACH-CACO, is beyond LEACH, LEACH-C and LEACH-CD. This implies that the LEACH-ACO helped improve the network lifetime [39], [40].

CONCLUSION
This paper presented a replacement algorithm LEACH-CACO to enhance the performance of WSNs. Through the proposed technique, it forms the sensor nodes into clusters by LEACH-C then combines with ACO algorithm to seek out the best path for the sink to maneuver over to gather information from the CHs. The simulation result shows that LEACH-CACO is better than LEACH, LEACH-C, and LEACH-CD in terms of the variety of nodes alive and also the quantity of energy consumed, therefore saving energy and enhance network lifespan. Within the future, we are able to attempt to add a lot of sinks or use different graph algorithm to seek out the simplest route for mobile sinks.