Target coverage and continuous connection are the major recital factors for Wireless Sensor Network (WSN). Several previous research works studied various algorithms for target coverage difficulties; however they lacked to focus on improving the network’s life time in terms of energy. This research work mainly focuses on target coverage and area coverage problem in a heterogeneous WSN with increased network lifetime. The dynamic behavior of the target nodes is unpredictable, because the target nodes may move at any time in any direction of the network. Thus, target coverage becomes a major problem in WSN and its applications. To solve the issue, this research work is motivated to design and develop an intelligent model named Distributed Flexible Wheel Chain (DFWC) model for efficient target coverage and area coverage in WSN applications. More number of target nodes is covered by minimum number of sensor nodes that can improve energy efficiency. To be specific, DFWC motivated at obtaining lesser connected target coverage, where every target is available in the monitoring area is covered by a smaller number of sensor nodes. The simulation results show that the proposed DFWC model outperforms the existing models with improved performance.
One of the emerging networks which can configure automatically is the Wireless Sensor Network (WSN), have an undefined number of sensor nodes connected logically and physically with the help of various communication mediums. Various WSN applications have deployed different kinds of sensors for monitoring purposes in a deterministic manner. Sensors are the most important component of any programmed system. Sensor nodes are distributed in a defined region to be watched [
In target monitoring, most of the targets can be monitored/sensed only within a determined distance. The sensors are restricted with a sensing range as
Network coverage is quickly becoming one of the most critical variables influencing wsn service quality. Sensors can cover targets inside that coverage area when deployed, and if all sensors, target location, and coverage areas are known, the coverage relationship can be determined. The purpose of the coverage challenge is to find a subset of the sensor and extend its life while maintaining the original remaining resources’ life limit and overall power consumption in mind. The Voronoi diagram feature can help to simplify decision-making while enhancing target coverage.
In WSN, the target coverage, area coverage and connectivity are the major factors to be considered. This research work focused on target coverage and area coverage. Target coverage is monitoring a set of target nodes using a set of sensor nodes. If any node fails while monitoring, it disconnects the entire network communication. Applications need
Most of the target coverage applications require a centralized mechanism for forwarding the observed data to the destinations. It takes more energy consumption for communication and data transmission. But a distributed mechanism enables each node to decide its own task and can change its mode accordingly. Comparing with the centralized mechanism, the distributed mechanism consumes lesser energy and increase the performance as well. Hence this paper proposed the Distributed Flexible Wheel Chain (DFWC) model as a distributed algorithm, as it is highly suitable for large-scale surveillance environments. The main contribution of the paper is given as follows. Design and implement a distributed FWC algorithm for target coverage Illustrate multiple scenarios for solving the target coverage problem and to maintain DFWC continuously and constantly, so as to assure network reliability and accuracy. FWC is distributed so that it can be implemented for any large-scale heterogeneous network applications. Balancing the load, range of sensing and communication, DFWC can manage the sensing ability, energy and target coverage efficiently.
This paper provides the following scenarios towards illustrating the target coverage problem which has been solved. The overall flow of the proposed model is given in Single Static Target with Multi-Static Sensor Nodes ( Multi Static Target with Multi-Static Sensor Nodes ( Single-Dynamic Target with Multi-Dynamic Sensor Nodes ( Multi-Dynamic Target with Multi-Dynamic Sensor Nodes (
Target and area coverage, continuous connectivity is the major factors to be considered as the major and necessary problems in WSN applications. The target coverage problem either covers the target continuously without any break, cover the specific target area by deploying minimum sensor nodes and verify the area and target is covered well. Nowadays, target coverage and connectivity problems have paid high attention to the surveillance industry. To provide better and efficient methodology, several algorithms and methodologies have been proposed in the earlier research works. For example, the author in [
A fine analysis method is used for solving coverage problems using two kinds of nodes with various abilities. It also discussed the impacts of a heterogeneous range of sensing and communication of the nodes [
The target must first be identified, and then its movement around the detection region must be followed continually. The problem of detecting and tracking a target must be solved with greatest precision and minimal power consumption at each sensor node [
From the above discussion, it is identified that various earlier research works have been studied extensively about coverage and connectivity problems. The main objective of this problem is to prolong the network lifetime. However, this paper proposed a new distributed FWC algorithm for solving the target coverage problem with a prolonged network lifetime. It also involves the load balancing, sensing abilities, target coverage, and reducing the energy consumption by focusing on each sensor nodes used in the network monitoring application. Before entering into deep, let us define the problem first. The sensor nodes can do (
To obtain high accuracy in target coverage the following are the assumptions made: All the sensor nodes are beacon nodes and are non-stationary. Nodes can able to know their location by itself, where it helps to estimate the distance. The sensing area of a node is a circle with the radius Nodes are distributed randomly with a unique identity in a 2D-Euclidean plane. All the nodes can sense and communicate with one another by themselves within the adjustable radius A pre-determined offset distance Sensors can be replaced when their battery is very low ( Nodes are arranged in a chain model and interconnected with an offset value The static and dynamic node’s information is always updated in a routing table, which can be erased after one round of operation.
Apart from the above list of assumptions, it is also assumed that, for any target
The sensing ability (sensibility) of the
This paper aimed to provide a better model for solving the target coverage problem with continuous connectivity. The overall flow of the proposed model is given in
The main objective of the paper is to design and implement a novel metaheuristic approach for sensor node deployment to enhance the accuracy of target coverage, area coverage, and connectivity with increased QoS of the network. This paper used a novel flexible chain model for deploying and interconnecting the sensor nodes with the target nodes. Due to the restricted energy, recharging or replacing the sensor nodes is difficult during the target coverage problem. However, use all the nodes at all the time could deplete their energy level quickly. So, a partial node usage method is adopted in this work, so as to reduce the energy consumption and prolong the network lifetime. So, the nodes are divided into different subsets row-wise or column-wise and make use of them for target covering at a different time schedule. To formulate the target coverage problem, a network node deployment model is illustrated in
Initially, the number of sensor nodes is deployed based on the number of target nodes. The sensor node
This paper introduced a novel distributed flexible wheel chain model for deploying the nodes and creates a node connection for target coverage and monitoring. Various real-time applications like transport designing use a wheel-chain model for making the vehicle run continuously without any struggle in a road. The fundamental idea of the wheel system is followed and based on that our proposed FWC model is designed and implemented for deploying the sensor nodes for target coverage. Wheels are overlapped at a small distance
From
From
In the first scenario
The distance between T1 and the center of S1, S2, S3, and S4 are diagonally at equal distance. S1 is overlapping with S2 and S3. S2 is overlapping with S1 and S4 vice versa. The center point of the overlapping regions of the sensor node is perpendicular to the target node T1.
While node deployment Rs, Rtis verified based on the t value, arranged around the target (see
Once the time duration TD is over, then alternate nodes are activated as current nodes and other nodes have kept idle. Hence this
In the second scenario
In that case, T1 can be monitored by S1 and S2, T2 can be monitored by S2 and S4, T3 can be monitored by S4 and S3, and T4 can be monitored by S3 and S1. Target node T5 can be monitored by any sensors from S1 to S4, similar to
No. of target nodes | No. of sensors nodes (random analysis) | No. of energy balancing nodes (worst case analysis) | No. of energy balancing nodes (best case analysis) |
---|---|---|---|
1 | 7 | 2 | 2 |
2 | 7 | 3 | 2 |
3 | 7 | 4 | 2 |
4 | 7 | 5 | 2 |
5 | 7 | 6 | 2 |
6 | 7 | 7 | 2 |
and if the number of sensor nodes is 2
In the scenario of
The final scenario illustrates how the dynamic nature of the target node can be monitored using sensor nodes. The paper assumed that the nodes are moving. So, as the target moves, with the same velocity and direction, the sensor nodes are also moving. The target moving directions and the sensor following directions are illustrated in
One of the swarm intelligence algorithms that belong to Artificial Intelligence is Particle Swarm Optimization (PSO). AI techniques are used to simulate human behavior using computation methods. The set of all target nodes and sensor nodes information like node location, velocity and distance from the previous position to current positions are learned by the PSO from the routing table, maintained by the network routing protocol. PSO initializes the location of each node in the network (search space). The set of all sensor nodes are divided into various disjoint sets. Compute the distance and offset the value of the sensor nodes related to the target node. Assign the subset of the sensor nodes are particle group in a specific region. Compute the offset as the fitness value. If the location of the target node changes, then PSO will move the sensor nodes according to the target based on the distance and the offset value. Read the routing table at every interval of time. Calculate the variations in the target node and arrange the sensor nodes accordingly. In the case of target moving scenarios, the PSO algorithm is enabled for location updating to keep track of the target.
At each iteration of the process, the PSO algorithm learns the routing table and updates the information such as the location of target, direction, and velocity. The information obtained by the PSO is broadcasted to the sensor nodes who are in activate-mode. Once the sensor nodes collect the target information, it automatically calculates the distance and offset and other information to move accordingly. PSO algorithm obtains the target node ID, the ID of the sensors currently available in the active mode. Then only the network nodes can understand the target and the sensors need to be activated. The four models of the node deployment process to do an accurate target coverage problem are given in the following Pseudocode. The pseudo-code involves setting the network, initializing the network parameters, enabling the deployment models based on the dynamic behavior of the target nodes. According to the target nodes’ behavior, the sensor nodes can be activated. The programming model of the four different scenarios can call the PSO when the target node’s behavior is moving. During the coverage problem, the sensors decide themselves go to an active state or idle state. In the beginning, all the sensor nodes are in the active state, after adjustment only a subset of the sensor nodes is assigned into an idle state. According to the implementation, the computational complexity is also calculated for the target coverage problem which only depends on the density of the target and sensor nodes deployed in the network, which is
The suggested DFWC model is implemented and the results are confirmed using MATLAB software, as indicated in
Parameter | Value |
---|---|
Monitoring area | 1000 × 1000 m2 |
Number of nodes | 50–200 |
Number of targets | 10–50 |
Sensing range | 50–60 m |
Transmission range | 100–120 m |
Initialization method | Random |
Idle energy es | 340 mW |
Receiving energy erx | 370 mW |
Transmit energy etx | 700 mW |
Based on the first experiment, connected subsets are constructed for a variable number of targets using a variable number of sensor nodes distributed in the network. Based on the number of targets, the number of sensor nodes required to generate the coverage sets is computed and confirmed in the experiment. The calculation of the number of cover sets obtained from the experiment before and after the deployment of the DFWC is shown in
Similarly, the number of cover sets from the installed sensor nodes is estimated from the experiment, and the result is shown in
Despite the fact that cover sets are developed and obtained, not all of them are used for target coverage or concurrent weather monitoring. To save energy, the cover set is divided into two parts: active and inactive. The active set is used for target coverage, while the rest of the nodes are inactive. Nodes in the active state will use a set amount of energy.
Another factor influencing the efficiency and performance of the suggested solution is its time complexity. Time complexity is computed by adding the time required to build and run a set of processes by the number of targets.
The power consumption of the sensor nodes is also calculated for performance verification. The energy is deducted from the total energy remaining in the sensor nodes based on node activity.
The main objective of this paper is to design and implement a novel algorithm for efficient target coverage problems with energy efficiency. To increase the target coverage efficiency, a Dynamic Flexible Wheel Chain model is proposed, where it interconnects the coverage nodes in a chain, without any break, but the nodes are alternatively changed from the active state into an idle state. The proposed model is implemented in and experimented over four scenarios where the target nodes are considered as static and dynamic. From the experiment, it is obtained that the DFWC uses very a smaller number of active nodes for target coverage and saves the energy, it leads to provide increased network lifetime. Regarding time complexity the DFWC has taken only less amount of computational time than the other conventional approaches. By comparing all the results, it is concluded that the proposed DFWC model is highly suitable for accurate target coverage in WSN applications with improved energy. The simulation indicated that the DFWC approach outperforms random deployment in terms of target coverage. Furthermore, comparing an Omni directional and Directional Sensors model with different types of networks and sizes could be an extension of the work presented in this research.