TY - JOUR
AU - Koriem, Samir, M
AB - Abstract Area detection and measuring is one of the most important problems in wireless sensor network because it mainly relates to the continuity and functionality of most routing protocols applied to the region of interest (ROI). Electronics failure, random deployment of nodes, software errors or some phenomena such as fire spreading or water flood could lead to wide death of sensor nodes. The damage on ROI can be controlled by detecting and calculating the area of the holes, resulting from the damaged sensor networks. In this paper, a new mathematical algorithm, wireless sensor hole detection algorithm (WHD), is developed to detect and calculate the holes area in ROI where the sensor nodes are spread randomly. WHD is developed for achieving quality of service in terms of power consumption and average hole detection time. The dynamic behavior of the proposed WHD depends on executing the following steps. Firstly, WHD algorithm divides down the ROI into many cells using the advantage of the grid construction to physically partition the ROI into many small individual cells. Secondly, WHD algorithm works on each cell individually by allocating the nearest three sensor nodes to each of the cell’s coordinates by comparing their positions, WHD connects each cell’s coordinate points with the selected sensor nodes by lines that construct a group of triangles, then WHD calculates the area of upcoming triangles. Repeating the previous step on all the cells, WHD can calculate and locate each hole in the ROI. The performance evaluation depends on the NS-2 simulator as a simulation technique to study and analyze the performance of WHD algorithm. Results show that WHD outperforms, in terms of average energy consumption and average hole discovery time, path density algorithm, novel coverage hole discovery algorithm and distriputed coverage hole Detection. 1. Introduction Wireless sensor network (WSN) is one of the new technologies for detecting and monitoring life phenomenon. WSNs are composed of a massive number of sensor nodes operated by small batteries; sensor nodes are mostly deployed in open and unprotected environments. They have significant limitation in communication capabilities and battery power. Nowadays the nodes are spread in large scale due to rapid technological advances in micro-electronic industries and the newly developed routing protocols that save more communication and computation power [1]. Sensor nodes are spread in open environments to detect and collect information from the surrounded phenomenon [2]. Then the remote base station receives report messages from the sensor nodes [3]. Various applications are dependent on WSNs such as military field exploration, flood of water, border protection and forest fires [4, 5, 6]. WSNs have unique characteristics due to their physical design, such as unreliability of sensor nodes, undefined network topology, high computation and communication power consumption and lots of storage difficulties [7, 8], So many challenges are presented in the solutions design and applications development of WSNs. In the real life applications, sensor nodes are randomly scattered over the ROI that allow some uncovered areas (holes) to be present in the ROI, which significantly degrade the network performance. The hole can be defined as the amount of area within the ROI that is not covered by any living sensor. The holes can also be created by the dynamic operations of the sensor nodes. Sensor nodes usually vanish by impact of random deployment, overheat, movement of animals, vehicles and people accidents. Such failures occur due to the static nature and the random deployment of the sensor nodes [9, 10]. The failure within any part of the network directly affects the performance of the total network locally and globally. The presence of holes in ROI definitely affect the routing paths, may cause failure of the routing protocols or separation of the network to many individual small networks. For illustration, the area region that is uncovered by any sensor node is considered as a hole, in which events of interest cannot be accomplished. To overcome the holes problems, the location of the holes and their areas must be determined, also alternative sensors are used respectively to keep the sensor alive as much as possible [11]. Therefore, holes coverage and connectivity of the network are from the most important aspects in WSNs [12, 13]. In this paper, a new wireless sensor holes detection algorithm (WHD) is proposed, which enables the sensor nodes to detect all the holes areas within the ROI, and calculates the holes areas to help the routing protocol to change its routing paths or to put extra mobile nodes to cover the holes areas. The proposed WHD algorithm uses the advantage of dividing the ROI by using the grid theory [14] to divide ROI into many clusters, and it runs in two phases. In phase one: WHD divides the ROI into many equally partitions called cells by using the Grid algorithm, then it stores the exact location of the four edges of each cell to use them in calculating the holes area. In phase two: WHD algorithm works on each single cell individually by determining the coordinates of its four edge points also determines the coordinates of the nearest three (if possible) nodes to each cell's edge points, then WHD determines if the ranges of the selected sensor nodes cover the cell's edge point, if the sensor's range doesn't cover, this means there is a hole and WHD begins to calculate the hole area and its position, if the sensor's range covers this coordinate point so there is no hole in that region of the cell. Figure 1 shows how WHD determines the presence of a hole in a cell. FIGURE 1. Open in new tabDownload slide A hole in one cell. FIGURE 1. Open in new tabDownload slide A hole in one cell. The rest of this paper is organized as follows: Section 3 presents the related work. Section 4 presents the contributions of the research work. Section 5 describes the modeling assumptions and problem goals. The proposed WHD algorithm is described in detail in Section 6. Section 7 represents the performance evaluation and the simulation results. Section 7 presents conclusion and future work. 2. Related Work Detecting and measuring the holes area is one of the important problems in WSN. So many researchers focused on detecting and calculating the holes areas within the ROI [9]. Authors in [15] propose an algorithm for efficient detection of holes boundary, where each node sends a broadcast message to all its neighbors. If one of the neighbor nodes detects a hole it calculates the distance between itself and the hole, then replays a message to the originator node with the new data. The originator receives all messages from all nodes and builds the hole boundary region. The drawback of this method is authors use all the available nodes to detect the boundary holes, so the transmitting packets and dropped packets are maximum and the consumption energy due to transmitting and receiving data is very large. Authors in [16] propose a new algorithm to detect the holes area by forming a boundary region using the available sensor nodes in the region. After formation of the boundary region around the holes, all sensor nodes adjacent to that boundary region is considered as boundary nodes that gives alarm to the routing protocol to avoid this holes region. On the other hand, authors do not calculate the holes area and the proposed algorithm uses all the available nodes in the ROI that causes severe power consumption when transmitting data to the base station and when calculating the holes boundary. In [17] the authors propose two novel algorithms to detect the coverage holes in ROI. The holes’ borders and their adjacent nodes can be easily detected by the first algorithm, distributed sector cover scanning, while locating the coverage holes is done by the second algorithm, directional walk. In [18] the authors propose an algorithm based on the sentinel scheme to reduce the sleeping node detection density by defining a new deep sleeping technique. Network lifetime and power consumption are the factors to calculate the detection rate. Furthermore, the coverage holes is addressed by using a triangular coverage repair procedure to cover the coverage hole. Authors in [19] introduce a graphical method to detect the holes region in ROI. They developed an algorithm that divides the ROI into many regions; every region needs at least k number of sensor nodes to construct the boundary area. The proposed algorithm depends on communicating the k number of sensor nodes in a circular disc. But authors assume the coverage of a sensor nodes are usually uniform in all directions; furthermore, the number of participant sensor nodes to detect the holes region are very large that causes a great power consumption to fulfill the process. Authors in [20] introduce an algorithm to detect the boundary holes in ROI; the mechanism depends on the connection between each node and three of its neighbor nodes to determine the stuck node, which is the last node to forward the data message. The disadvantages for this method are the algorithm detects holes of fixed sizes only, the boundary detection is done by comparing one hop neighbor, message forwarding overhead could be very large and the boundary detection is not applicable for large density of nodes. Furthermore, the proposed algorithm does not calculate the holes area. Authors in [21] propose an algorithm for detecting the holes boundary in ROI. By detecting the intersection points of the adjacent live nodes the algorithm can draw a figure about present holes area, then the holes area are easily detected by an algebraic method using the intersection points and the sensing ranges of the neighbor nodes to the holes. The proposed algorithm has two weaknesses; firstly, it uses all the deployed sensor nodes on the ROI, which causes severe power consumption until the task is accomplished. Secondly, the proposed algorithm just defines the boundary region of the holes but does not calculate the holes area. Authors in [22] developed an algorithm to calculate the holes area by comparing the size of designed Voronoi cell with the adjacent node’s sensing range, and marks the border nodes of holes area by using simple geometric calculations. The proposed algorithm assumes that each Voronoi cell has sensor nodes and these sensor nodes are able to communicate with each other that causes massive power consumption. Furthermore, the proposed algorithm uses all the deployed sensor nodes on ROI to determine just the boundary region not the holes area. Authors in [23] developed an algorithm to determine the holes in ROI based on Delaunay triangulation, authors use the standard geometric tool called Delaunay triangulation to detect the coverage holes. The proposed algorithm spends a lot of communication messages between the deployed sensor nodes to construct the Delaunay triangulation. The proposed algorithm has two weaknesses: firstly, the communication overhead is relatively high due to the direct connection between sensor nodes together. Secondly, the proposed algorithm does not calculate the area of the detected holes but defines the holes region in ROI. Authors in [24] propose an algorithm that mainly consists of two phases, namely coverage holes detection (CHD) and coverage restoration. In CHD, each sensor node independently detects any hole by updating certain information with its neighbor nodes. To restore the coverage, a sensor node with relatively higher residual energy is given priority to cover the hole closer to it by increasing its sensing range up to a maximum limit. The proposed algorithm has two weaknesses: firstly, it depends on direct communication between each two neighbor sensor nodes that leads to (i) failing of detecting holes if one sensor has no neighbor node and (ii) communication over head is relatively high due to the direct connection between sensor nodes together. Secondly, the proposed algorithm does not calculate the area of the detected holes but defines the holes region in ROI. Authors in [25] aim to design a localization-free and energy-efficient hole bypassing technique for fault-tolerant sensor networks. The idea of the proposed algorithm is firstly, construct a cluster tree rooted at sink node where network is partitioned into multi-hop clusters. Secondly, the authors proposed an inter-cluster energy-efficient solution in the first step and an inter-cluster robust solution in the second step by applying these methods. So the proposed algorithm aims to avoid the cost of localization and network-wide topology recreation. The proposed algorithm has two weaknesses: firstly, it depends on all the deployed sensor nodes to determine the holes area that leads to massive communication overhead. Secondly, the proposed algorithm does not calculate the area of the detected holes but defines the holes region in ROI. 3. CONTRIBUTIONS OF THE PROPOSED RESEARCH WORK In this section, authors explain how to determine and calculate the total holes area and their breadth formed in the ROI due to random spreading of sensor nodes or damaged sensor nodes by any activity. This study works on enhancing the functions performance of the applied routing protocol on the network. The proposed WHD algorithm studies the ROI for ease of building small cells, which increases the network lifetime and eliminates the direct communication between the base station and any sensor node. Therefore, the main contributions of the research work can be illustrated as follows: Build a homogeneous ROI for ease and better dealing with randomly scattered sensor nodes as shown in Fig. 2. Build WSN reliable model to increase WSN lifetime by cutting down the ROI into many small pre-defined cells. WHD operates in each single cell individually, then advertises the collected data to the base station according to the selected routing protocol technique as shown in Fig. 3. Eliminate the direct communications between the base station and each sensor node, which consumes less power and raises the network lifetime. Develop an NS-2 simulator model, describing the performance evaluation of the proposed WHD algorithm. FIGURE 2. Open in new tabDownload slide Formation of grid. FIGURE 2. Open in new tabDownload slide Formation of grid. FIGURE 3. Open in new tabDownload slide A single cell. FIGURE 3. Open in new tabDownload slide A single cell. 4. PROBLEM FORMULATION 4.1. Modeling assumption In this paper, some assumptions are used regarding the WSNs: Nodes are distributed randomly among a 2D space e.g. (X, Y). All sensor nodes have the same design, hardware characteristics and exact power supply (battery power). All sensor nodes have the same initial energy power. Radio channel is equal, the amount of energy consumption for message transmission (sending and receiving) is the same. The sensing and the communication ranges are the same for all sensor nodes. Each single node can find its own position after deployment through GPS devices or other localization approaches. 4.2. Holes problem Holes may occur in WSNs located in various environments having challenging conditions. A hole may stop the operation of the routing protocol completely or may partition the network into disjoint parts that may significantly reduce the event collection. To recover faults in the presence of holes, we propose to construct a WHD algorithm that aims to detect the boundaries of the holes and calculates the area of the holes on ROI. 4.3. Problem goals WSN is generally considered as an undefined topological environment due to the randomly deployment of static sensor nodes in the ROI, or damage of some sensor nodes due to any activity. So, there is some isolated nodes with no neighboring nodes to connect with. This is why the holes area is created. Our goal is to find the exact estimated amount of holes area in the ROI, which helps in determining the amount of damaged area in case of natural disasters, support the applied routing protocol to change the routing paths and helps in estimating the position of additional mobile nodes in appropriate places to cover the holes. 4.4. Definitions In this section, we define a few terms that are used throughout the paper to develop the proposed WHD algorithm. Let n be the number of chosen nodes by the application, Ci be the number of cells formed on ROI, Pi be the number of edge points in each cell on ROI, NHi be the head node (HN) of each cell, Ai be the area of one formed triangle, Aj be the area of K + 1 triangles formed at one edge point, Acell be the total holes area in one cell, Atotal be the total holes area in ROI, X0 be the point that has x = 0, Y0 be the point that has y = 0, (X0, Y0) = (0, 0), K be the number of horizontal lines, m be the number of vertical lines, Za be the length of the cluster and Zb be the width of the cluster. 5. THE PROPOSED WHD ALGORITHM In this section, the WHD algorithm is designed to detect and calculate the holes areas in the ROI. The WHD algorithm works on heterogeneous ROI that has no topology, WHD uses grid algorithm [14] to convert the ROI into homogeneous topology then divides the ROI into many predefined static similar-sized cells (according to the area to be served). WHD works on each cell individually, in each cell WHD chooses the nearest (n) number of sensor nodes by comparing the coordinates of the cell’s edges points with the position of sensor nodes laying within this cell and chooses the nearest (n) number of sensor nodes. After selecting the nearest (n) nodes within each cell, WHD elects the nearest node to each cell’s coordinate as a HN to collect the data from the other nodes and to calculate the holes area by using the triangulation method. WHD is made up of two phases: Grid formation phase Hole detection phase. 5.1. Grid formation phase Usually sensor nodes are scattered randomly on the ROI that is the reason of why holes are formed between the adjacent nodes. Also the ROI becomes heterogeneous because there is no defined shape of ROI. So that grid construction algorithm [14] is used to build a homogeneous ROI by building a grid of cells. Each cell is a rectangle of Za × Zb dimensions. Where Za and Zb are the length and width of the cell. In this phase the grid formation algorithm is used to divide the ROI into many predefined static similar sized cells, as shown in Fig. 2. $$\begin{equation} {\displaystyle \begin{array}{l}\mathrm{For}\ \mathrm{X}=0\ \mathrm{to}\ \mathrm{X}=\mathrm{k}\\ {}\{\\ {}\kern2em \mathrm{For}\ \mathrm{Y}=0\ \mathrm{to}\ \mathrm{Y}=\mathrm{m}\\ {}\kern2em \{\\ {}\kern3em \mathrm{f}\left(\mathrm{x},\mathrm{y}\right)=\left({\mathrm{X}}_0+\mathrm{X}\ast {\mathrm{Z}}_{\mathrm{a}},{\mathrm{Y}}_0+\mathrm{Y}\ast {\mathrm{Z}}_{\mathrm{b}}\right)\\ {}\kern2em \left\}\ \right\}\end{array}} \end{equation}$$ (1) 5.2. Hole detection phase This phase is responsible for detecting and calculating the holes area in the ROI. WHD works on each cell individually as shown in Fig. 3. WHD calculates the holes areas by choosing the nearest (n) number of sensor nodes to the coordinates of the cell, the choice is done by comparing the cell’s coordinates position with the position of each sensor nodes laying within this cell. After choosing the desired number of sensor nodes WHD elects the nearest sensor node to the cell’s coordinate as a HN. The HN has two functions, firstly HN collects the data from the other nodes and cascades it according to the selected routing protocol. Secondly, HN calculates the holes area by using the triangulation method. Suppose P is a set of edge points of a cell, where P1, P2,…, Pi ∈ P. Then WHD algorithm compares the maximum transmission ranges of the (n) chosen sensor nodes with the position of each cell’s edge point, if one of the selected sensor node’s transmission ranges hits the grid’s edge then there is no hole, if not so definitely there is a hole in that cell, as shown in Fig. 4. FIGURE 4. Open in new tabDownload slide A hole in a cell. FIGURE 4. Open in new tabDownload slide A hole in a cell. Then WHD calculates the hole area by applying the triangulation method between the chosen sensor nodes and the cell’s edge point, by connecting the sensor nodes to the nearest cell’s coordinate point that has coordinates (Xn, Yn) with straight lines, then we get (n + 1) triangles created, as shown in Fig. 5. FIGURE 5. Open in new tabDownload slide Cutting down the hole into many triangles. FIGURE 5. Open in new tabDownload slide Cutting down the hole into many triangles. FIGURE 6. Open in new tabDownload slide Flow chart of WHD. FIGURE 6. Open in new tabDownload slide Flow chart of WHD. Depending on triangulation theory we can calculate the hole area in this grid by calculating the area of triangles A1, A2,…,An+1 according to the below formulas, as shown in Fig. 5. Where n is the total number of chosen sensor nodes, and r is the radius of the sensor. $$\begin{equation} {\mathrm{L}}_1=\left({\mathrm{Y}}_1-{\mathrm{Y}}_0-\mathrm{r}\right) \end{equation}$$ (2) $$\begin{equation} {\mathrm{L}}_{\mathrm{n}+2}=\left({\mathrm{X}}_{\mathrm{n}}-{\mathrm{X}}_0-\mathrm{r}\right) \end{equation}$$ (3) $$\begin{equation} {\displaystyle \begin{array}{l}\mathrm{For}\ \mathrm{i}=1\ \mathrm{to}\ \mathrm{n}\\ {}\{\\ {}{\mathrm{L}}_{\left(\mathrm{n}+1\right)}=\sqrt{{\left(Y(n)-Y0\right)}^2+{\left(X(n)-X0\right)}^2}-\mathrm{r}\end{array}} \end{equation}$$ (4) $$\begin{equation} {L}_{bn}=\sqrt{L{\left(n+1\right)}^2-{Ln}^2} \end{equation}$$ (5) $$\begin{equation} \mathrm{Area}\ {\mathrm{A}}_{\mathrm{n}}=\sqrt{S\left(S- Ln\right)\left(S-\mathrm{L}\left(\mathrm{n}+1\right)\ \right)\left(S- Lbn\right)} \end{equation}$$ (6) $$\begin{equation} \mathrm{Where}\ S=\frac{Ln+L\left(n+1\right)+ Lbn}{2} \end{equation}$$ (7) $$\begin{equation} {\mathrm{Lb}}_{\left(\mathrm{n}+1\right)}=\sqrt{L{\left(n+1\right)}^2-L{\left(n+2\right)}^2} \end{equation}$$ (8) $${\displaystyle \begin{array}{l}\ \mathrm{Area}\ {\mathrm{A}}_{\mathrm{n}+1}=\sqrt{S\left(S-L\left(n+1\right)\right)\left(S-L\left(n+2\right)\right)\left(S- Lb\left(n+1\right)\right)}\\ {}\}\end{array}}$$ So the area of the hole in one cell (Aj) can be determined by calculating and summing the amount of the individual triangle area by the equation $$\begin{equation} {\mathrm{A}}_{\mathrm{j}}={\sum}_1^{\mathrm{n}+1}\mathrm{Ai} \end{equation}$$ (9) where Ai is the area of the triangle i that is formed after cutting the hole to many triangles. By repeating the previous process of calculating the area of holes in one cell to the four edges of the same cell, WHD can discover any holes area in that cell by the equation $$\begin{equation} {\mathrm{A}}_{\mathrm{cell}}={\sum}_1^{k+1}{\mathrm{A}}_{\mathrm{i}} \end{equation}$$ (10) To calculate the total holes area in ROI, by repeating the previous process on all the cells in ROI, WHD can discover and calculate the total holes area in the ROI by the equation $$\begin{equation} {\mathrm{A}}_{\mathrm{total}}={\sum}_1^j{\mathrm{A}}_{\mathrm{cell}} \end{equation}$$ (11) For more declaration, in real WSN applications, each sensor network broadcasts a hello message to ensure there is other sensor nodes in its range of interest whether they lie inside or outside the desired cell. If the hello message replies positively so other sensor nodes are in the range of that node and there is no holes around that node, if the hello message replies negatively so there is no other sensor nodes around that node and might be a hole around that sensor. The proposed WHD algorithm (see next page). In the simulation process some standard simulation parameters are used [22] for evaluation and accurate comparison between the proposed WHD algorithm and both of VCHDA and PD algorithms. Table 1 shows the used simulation parameters. TABLE 1. NS-2 simulation parameters. Routing protocol CRP Shape of the monitored area Square Network size 500 m × 500 m Number of nodes 100 ~ 1500 Base station location (50, 175) Data packet size 40 Bytes Initial energy of nodes 100 J Communication range 10 m Data transfer ratio 250 Kbps Routing protocol CRP Shape of the monitored area Square Network size 500 m × 500 m Number of nodes 100 ~ 1500 Base station location (50, 175) Data packet size 40 Bytes Initial energy of nodes 100 J Communication range 10 m Data transfer ratio 250 Kbps Open in new tab TABLE 1. NS-2 simulation parameters. Routing protocol CRP Shape of the monitored area Square Network size 500 m × 500 m Number of nodes 100 ~ 1500 Base station location (50, 175) Data packet size 40 Bytes Initial energy of nodes 100 J Communication range 10 m Data transfer ratio 250 Kbps Routing protocol CRP Shape of the monitored area Square Network size 500 m × 500 m Number of nodes 100 ~ 1500 Base station location (50, 175) Data packet size 40 Bytes Initial energy of nodes 100 J Communication range 10 m Data transfer ratio 250 Kbps Open in new tab 6.2. Simulation result 6.2.1. Average energy consumption Energy consumption is one of the most important factors affecting the sensor nodes because sensor nodes are battery powered, routing protocol or any algorithm runs on the sensor node has to reduce the power consumption to give the sensor node the chance to live more time, otherwise network separation will happen. Average energy consumption can be defined as the total amount of energy (i.e. computation and communication power) consumed by only the participant sensor nodes, out of the total deployed sensor nodes, during the simulation time. This will continue until the WHD algorithm calculates the total holes area that is found in ROI. Figure 7 shows the simulation results of the average energy consumption by using all the participant sensor nodes when calculating the holes area in the ROI. This is done by using random number of holes and fixed number of sensor nodes. As it is known, the amount of energy consumption varies directly with the network density (number of deployed sensor nodes) in the ROI, so as the number of deployed sensor nodes increases, the average energy consumption increases and vice versa. Since the largest amount of consumed energy is mainly caused by communication, WHD uses only a defined number of sensor nodes efficiently to calculate the holes area. Thus, the minimal use of sensor nodes leads to many enhancements on the network. Firstly, the minimal communication cost for neighborhood discovery and control information exchange between the base station and any sensor node. Secondly, using minimum number of sensor nodes leads to the minimal amount of aggregated data by the HNs in each cell. Thirdly, by determining the holes area the working routing protocol will avoid that region of holes, which leads to reduce the total network delay. Fourthly, the dropped packets when communicating between nodes will be minimum so the minimum number of retransmission data between the nodes will be obtained. Accordingly, the total number of packets in the network will be the minimum. There is a direct relation between the transmitting packets in the network and the power consumption. So if the transmitting packets, dropped packets, network delay are reduced the average energy consumption decreases. It is clear from Fig. 7 that WHD algorithm outperforms PD algorithm in terms of average energy consumption by ~44% average, WHD outperforms VCDHA by ~21%, WHD also outperforms DCHD by ~15%. The enhancement of average energy consumption is mainly due to the participation of less sensor node. That leads to reduce the communication overhead between the base station and the sensor nodes. Since the WHD algorithm saves more power to calculate the holes area compared to other protocols, WHD enhances the network lifetime and reduces the communication overhead between the sensor nodes. 6.2.2. Average holes discovery time Holes discovery time can be defined as the total time needed for only participant sensor nodes to detect and calculate the total holes area in the ROI. Because of some hardware restriction, only the algorithms of low complexity can run on these sensor nodes. Figure 8 shows the simulation results of the average holes discovery time by using only the participant nodes, out of the total deployed sensor nodes, this is done by using random number of holes and fixed number of sensor nodes. As it is known, the average hole discovery time varies directly with the network density (number of deployed sensor nodes) in the ROI, so as the number of deployed sensor nodes increases the average hole discovery time increases and vice versa. Thus, the minimal use of sensor nodes leads to many enhancements on the network. Firstly, using minimum number of sensor nodes leads to the minimal amount of aggregated data by the HNs in each cell. Secondly, reduces the communication and computation power between nodes, this leads to the minimal amount of aggregated data by the HNs in each cell. Thirdly, by determining the holes area the working routing protocol will avoid that region of holes, which leads to reduce the total network delay. Accordingly, the average holes discovery time decreases. FIGURE 7. Open in new tabDownload slide Average energy consumption comparison. FIGURE 7. Open in new tabDownload slide Average energy consumption comparison. FIGURE 8. Open in new tabDownload slide Average hole discovery time comparison. FIGURE 8. Open in new tabDownload slide Average hole discovery time comparison. It is clear from Fig. 8 that WHD algorithm outperforms PD algorithm in terms of average holes discovery time by ~46% average, WHD also outperforms VCDHA by ~33% average and WHD also outperforms DCHD by ~21%, WHD also outperforms DCHD by ~21%. The enhancement of average discovery time is mainly due to the participation of less sensor node. That leads to reduce the communication and computation overheads between the base station and the sensor nodes. Since the WHD algorithm saves more power to calculate the holes area compared to other protocols, WHD enhances the network lifetime and reduces the communication overhead between the sensor nodes. 7. CONCLUSION AND FUTURE WORK Accurate detection and calculation of holes in WSNs do not only help the routing protocols to avoid the detected holes area and modify their routing paths to the destination, but also help in the field of movement tracking such as that of animals, people and vehicles, also some other phenomena such as fire spreading or water flood. This paper has presented the WHD algorithm to calculate the total holes area in the WSNs ROI, where the network nodes operate on limited battery energy. The reduction of the consumed communication time and computational power is considered the main challenge regarding this network. The WHD offers the advantage of small transmission distances for most of the sensor nodes. This concept leads to increasing the network lifetime and decreasing the communication and computation overheads as well as enhancing the quality of the network. The basic concept of the WHD is as follows: WHD divides the WSN into a number of similar cells by using the grid algorithm. WHD works on each cell individually by storing each cell’s coordinate positions and finds the nearest number of sensor nodes (depending on the application) to each coordinate. WHD begins to cut each cell into a number of triangles by drawing lines between each coordinate and the selected number of sensor nodes. WHD algorithm calculates the area of the formed triangles in each cell that represents the holes area in that cell. WHD sums the total holes areas in every cell to calculate the total holes area in the ROI. Since the energy consumption is mainly caused by communication, so less sensor nodes are used to calculate the holes area. This leads to the minimal communication cost for neighborhood discovery and control information exchange between base station node and the sensor nodes. Also, the amount of aggregated data by the HN are minimal due to the minimal use of sensor nodes in each cell. For the above achievements the average energy consumption, communication overhead and average holes discovery time decreases. Subsequently the network lifetime increases accordingly. In order to evaluate the performance of the proposed WHD algorithm, we compare its performance results with those obtained from PD and VCHDA algorithms in terms of average energy consumption and average holes discovery time. Based on these comparisons, it is obvious that the upcoming results are more efficient and increase the network lifetime in favor of WHD algorithm. In our future work, we shall try to recover these holes with less number of mobile nodes. References 1. Chen , Z. , Li , A. , Choi , Y. and Sekiya , H. ( 2017 ) Energy-Efficient Broadcasting Scheme for Smart Industrial Wireless Sensor Networks . Mob. Inform. Syst. , 2017 , 17 pages . Hindawi Publishing Corporation. Article ID 7538190 . WorldCat 2. Debnath , S. , Singh , A. and Hossain , A. ( 2016 ) A comprehensive survey of coverage problem and efficient sensor deployment strategies in wireless sensor networks . Indian J. Sci. Technol. , 9 . doi: https://doi.org/10.17485/ijst/2016/v9i45/99032 . WorldCat 3. Zhao , J. , Gan , O. and Gligor , V. ( Jan 2017 ) Environmental monitoring using wireless sensor networks based on IOT . Int. Res. J. Eng. Technol. (IRJET) , 4 , 1371 – 1378 . WorldCat 4. Wua , M. , Tana , L. and Xiong , N. ( 2016 ) Data prediction, comparison, and recovery in clustered wireless sensor networks for environmental monitoring applications . Inform. Sci. J. , 329 , 800 – 818 . Google Scholar Crossref Search ADS WorldCat 5. Ahadipour , A. and Haddad , A. ( 2017 ) LPKP: location-based probabilistic key pre-distribution scheme for large-scale wireless sensor networks using graph coloring . Int. J. Inform. Security , 9 , 27 – 39 . WorldCat 6. Mahapatra , C. , Sheng Z. and Kamalinejad , P. ( 2017 ) Optimal power control in green wireless sensor networks with wireless energy harvesting , WakeUp Radio Transmission Control , 5 , 501 – 518 . WorldCat 7. Jan , M. , Nanda , P. and X. He ( 2017 ) PAWN: a payload-based mutual authentication scheme for wireless sensor networks . Concurrency Computat Pract Exper. , 29 , e3986. https://doi.org/10.1002/cpe.3986 . WorldCat 8. Kadu , R. and Jaini , P. ( 2015 ) Hole detection and healing for improving coverage in WSNs . Int. J. Recent Innov. Trends Comput. Commun. , 3 , 70 – 73 . Google Scholar Crossref Search ADS WorldCat 9. Kumar , R. and Dhingra , S. ( 2015 ) Coverage hole detection in wireless sensor networks . Int. J. Comput. Netw. Wirel. Commun. (IJCNWC) , 5 , 4 , 540 – 548 . WorldCat 10. Beghdad , R. and Lamraoui , A. ( 2016 ) Boundary and holes recognition in wireless sensor networks . J. Innov. Digit. Ecosyst. , 3 , 1 – 14 . Google Scholar Crossref Search ADS WorldCat 11. K oriem , S. , Bayoumi , M. and Nouh , S. ( 2016 ) Circles routing protocol for wireless sensor networks . Int. J. Intell. Comput.Inform. Sci. IJICIS , 16 , 21 – 36 . WorldCat 12. Zaidi A. and Rakrak , S. ( 2016 ) A comparative study of target tracking approach in wireless sensor networks . J. Sensors , 2 , 11 pages . Hindawi Publishing Corporation. Article ID 3270659 . WorldCat 13. Antil , P. , Malik , A. and Kumar , S. ( 2016 ) Neighbor adjacency-based hole detection protocol for wireless sensor networks. 7th Int. Conf. on Communication and Virtualization , pp. 866 – 874 . 14. Khan , A. , Abdullah , A. and Razzaque , M. ( 2015 ) VGDRA: a virtual grid-based dynamic routes adjustment scheme for mobile sink-based wireless sensor networks . IEEE Sensors J. , 15 , 526 – 534 . Google Scholar Crossref Search ADS WorldCat 15. Jewel , D. , Brundha P. and Swaminathan , G. ( 2016 ) Improved hole detection healing and replacing algorithm coverage in wireless sensor networks . Int. J. Sci. Res. Sci. Eng. Technol. (IJSRSET) , 2 , 724 – 731 . WorldCat 16. Zhao , L. , Liu , W. , Zhang , R. and Tan , Q. ( 2016 ) Detecting boundary nodes and coverage holes in wireless sensor networks . Mobile Inform. Syst. , 2016 , 16 pages . Hindawi Publishing Corporation. Article ID 8310296. WorldCat 17. Snega , R. and Senthil , K. ( 2015 ) Detection and healing of holes in wireless sensor networks . Int. J. Emerg. Technol. Comput. Sci. Electron. (IJETCSE) , 13 , 380 – 385 . WorldCat 18. Xu , Y. , Zeng , Z. and Deng , O. ( 2015 ) An energy efficient repair node scheduling algorithm for WSN. In Proc. of Springer, Science + Business Media , 8 December 2015 . New York . 19. Suku , D. , Philip , P. ( 2015 ) Hole detection and healing for a ROI in wireless sensor networks . Int. J. Eng. Res. Gen. Sci. , 3 , 234 – 240 . WorldCat 20. Febil A. and Loganathan , D. ( 2016 ) Sensor deployment algorithm for hole detection and healing by using energy based local healing . Int. J. Eng. Sci. Res. Technol. , 5 , 1423 . WorldCat 21. Prasan , S. , Ming , C. and Shih-Lin , W. ( 2016 ) An efficient distributed coverage hole detection protocol for wireless sensor networks . Sensors 2016 , 386 . doi: https://doi.org/10.3390/s16030386 , Vol. 16. WorldCat 22. Dai , G. , Liv , H. , Chen , L. and Zhou , B. ( 2016 ) A novel coverage holes discovery algorithm based on Voronoi diagram in wireless sensor networks ”, Int. J. Hybrid. Info. Technol. , 9 , 273 – 282 . Google Scholar Crossref Search ADS WorldCat 23. Soundarya , A. and Santhi , V. ( 2017 ) An efficient algorithm for coverage hole detection and healing in wireless sensor networks. 1st Int. Conf. on Electronics, Materials Engineering and Nano-Technology (IEMENTech) , pp. 1 – 5 . Kolkata . Google Preview WorldCat COPAC 24. Amgoth , T. and Jana , P. ( 2017 ) Coverage hole detection and restoration algorithm for wireless sensor networks . Peer Peer Netw. Appl. , 10 , 66 – 78 . Google Scholar Crossref Search ADS WorldCat 25. Yilmaz , O. , Dagdeviren , O. and Erciyes , K. ( 2014 ) Localization-free and energy-efficient hole bypassing techniques for fault-tolerant sensor networks . J. Netw. Comput. Appl. , 10 , 164 – 178 . Google Scholar Crossref Search ADS WorldCat © The British Computer Society 2019. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Improving the Mechanism of Detecting and Measuring Holes in Ad hoc Wireless Sensor Network
JO - The Computer Journal
DO - 10.1093/comjnl/bxz054
DA - 2019-09-01
UR - https://www.deepdyve.com/lp/oxford-university-press/improving-the-mechanism-of-detecting-and-measuring-holes-in-ad-hoc-MMhiu5PgGv
SP - 1505
VL - 62
IS - 10
DP - DeepDyve
ER -