INTERNATIONAL JOURNAL OF COMPUTERS COMMUNICATIONS & CONTROL ISSN 1841-9836, 11(6):877-888, December 2016. Coverage Optimization Strategy for WSN based on Energy-aware L. Zhu, C. Fan, H. Wu, Z. Wen Li Zhu, Chunxiao Fan*, Zhigang Wen Beijing University of Posts and Telecommunications, Beijing Key Laboratory of Work Safety Intelligent Monitoring, School of Electronic Engineering, No.10 Xitucheng Road, Beijing, P.R.China, 100876 jolie.zhl@hotmail.com *Corresponding author: cxfan@bupt.edu.cn zwen@bupt.edu.cn Huarun Wu National Engineering Research Center for Information Technology in Agriculture, Key Laboratory for Information Technologies in Agriculture, Ministry of Agriculture, Room 316, Beijing Agriculture Science and Technology Building A, No. 11 Beijing Shuguang Garden Middle Road, Haidian District West Suburb, Beijing,China, 100097 wuhr@nercita.org.cn Abstract: In order to optimize the wireless sensor network coverage, this paper de- signs a coverage optimization strategy for wireless sensor network (EACS) based on energy-aware. Under the assumption that the geographic positions of sensor nodes are available, the proposed strategy consists of energy-aware and network coverage adjust- ment. It is restricted to conditions such as path loss, residual capacity and monitored area and according to awareness ability of sensors, it would adjust the monitored area, repair network hole and kick out the redundant coverage. The purpose is to balance the energy distribution of working nodes, reduce the number of “dead” nodes and balance network energy consumption. As a result, the network lifetime is expanded. Simulation results show that: EACS effectively reduces the number of working nodes, improves network coverage, lowers network energy consumption while ensuring the wireless sensor network coverage and connectivity, so as to balance network energy consumption. Keywords: WSN, coverage optimization, energy-aware, hole repair, sensing radius. 1 Introduction Wireless sensor network (WSN) consists of many sensor nodes which draw on self-organization to transmit data between nodes. These sensor nodes are small in size with limited energy and certain awareness, making them widely applied to many fields, such as military, transportation, environment protection, medical care, disaster relief and agriculture. Currently, sensor nodes are distributed manually or randomly. Manual distribution is executed in certain circumstances requires high on environment. So in most cases, random distribution [1–3] is the primary choice as it is easy to operate and able to reduce the interference on human activities. However, under circumstances of high density and large area, wireless sensor network is weak in expansion, which results in that the transmission signal is easy to be disturbed, the network energy is unevenly distributed and network become unstable. It is urgent to solve problems of increasing the effective coverage in monitored areas as much as possible, lowering network energy consumption, extending network lifetime, and improving network performance under the condition of limited energy and using the least sensor nodes. And network coverage is one of the most important indicators of network performance measurement. Copyright © 2006-2016 by CCC Publications 878 L. Zhu, C. Fan, H. Wu, Z. Wen Among existing studies, literature [4] proposes a method to calculate redundancy of probability node. With this method, nodes in the network can acquire information about their own, such as redundancy, without knowing the geographic position. It also proposes a node adjustment strategy without knowing node position. But this strategy neglects the awareness redundancy area within two-hop neighbor nodes, making many redundant working nodes exist during node adjustment. Literature [5] proposes a node distribution strategy based on wireless beacon self- adaption. This strategy repairs network coverage holes by adding wireless beacon to the network but fails to consider the cost of adding the beacon and the influence on monitored area. Literature [6] proposes to use a circle to solve node adjustment coverage. It calculates the angle between the awareness circle and the monitored area to get the minimum number of working nodes and reduces redundant working nodes. But under high coverage, the connectivity is weak. Literature [7] acquires information of null nodes and introduces mobile nodes. To be specific, it replaces null nodes with mobile nodes to fill the network holes. The network coverage is increased, but the energy of mobile nodes is limited and there are too many null nodes in the network. Literature [8] proposes a coverage strategy based on genetic algorithm. It selects the maximum solution set of coverage nodes. With genetic algorithm, it uses evaluation function to optimize random samples and reach a balance of network coverage. But the algorithm is relatively complicated and requires much calculate. To some extent, it increases network energy consumption. In order to improve effective coverage, this paper proposes a coverage optimization strategy for WSN based on energy aware, establishes a network coverage model. Network nodes are distributed randomly; a network coverage model is established the relationship between node residual capacity and sensing range are examined to set reasonable sensing range of nodes, bal- ance network energy consumption, repair network coverage holes, and extend network lifetime. Finally, a simulation experiment is conducted to prove the relationship between network energy consumption and coverage of the proposed algorithm under different circumstances. The pro- posed algorithm aims at balancing network energy consumption, improving network coverage, and minimizing node energy consumption. This paper establishes a non-linear model of coverage optimization strategy for wireless sensor network (EACS) and exercise stricter constraints to turn it to linear-constraints and get the second-best solution. 2 Network model and problem description 2.1 Analysis model of energy-aware node Suppose the wireless sensor network consists of N randomly distributed nodes, during node awareness, the energy of physical signal changes inversely with the distance between the signal and the awareness target. Such tendency is mainly due to path attenuation during transmission, Select any source node si within the scope of monitoring, when target node sj is anywhere in the plane, the awareness intensity Ψi(j) of souse node to target node is expressed as: Ψi(j) = { 0 Rs < d(i,j) λe−kd(i,j) 0 < d(i,j) ≤ Rs(1) (1) where k is the indicator of signal attenuation. Rs is the maximum effective sensing radius of the node. d(i,j) is the Euclidean distance between node i and j. λ is a constant value. The awareness intensity Φ(j) of node j is expressed as: Φ(j) = 1 − (1 − Ψ1(j))(1 − Ψ2(j)) · · ·(1 − Ψi(j)) · · ·(1 − Ψn(j)) = 1 − i=n∏ i=1 (1 − Ψ1(j)) (2) Coverage Optimization Strategy for WSN based on Energy-aware 879 When Φ(j) > �, target node j is sensed. When Φ(j) < �, target node j is not sensed, and at this time, j is the coverage blind-point (ε target node j is not sensed, and at this time, j is the coverage blind-point H0 : g(i) = ϑi i = 1, 2, . . . ,N H1 : g(i) = ϑi + ψi(j) i = 1, 2, . . . ,N (3) Where ϑi is the background noise signal following the normal distribution ϑi ∼ N(µ,σ2), and ϑi(j) is useful signal. The signal detected by sensor nodes is g(i). The target that exits is expressed by H0 and the target that does not exist is expressed by H1. 2.2 Network model N non-overlapped wireless sensor nodes, S = {si|1 < i < n}, whose sensing radius is Rs are randomly distributed in a two-dimensional monitored area A. Suppose the network has the following features: (1) The node position is permanent. There is no inherent relationship between the commu- nication radius Rc and sensing radius Rs. (2) All nodes have isomorphism. The wireless self-organized network is constructed in the monitored area. (3) All nodes adopt the node probability awareness model. (4) The sensing radius of sensor node can be adjusted according to node residual capacity. (5) The initial energy of sensor nodes is W , and they have synchronous clock. (6) Sensor nodes use the location technology in literature to acquire their own locations. At the same time nodes can acquire the information of neighbor nodes, such as residual capacity, node position within effective communication distance. 2.3 Problem description A number of sensor nodes are distributed randomly within the monitored area A. Probabilis- tic sensing model is deployed, and network nodes are adjusted according to data from 2.2network model; a wireless sensor network is established to balance network energy consumption and net- work coverage, adjust sensing ranges based on residual capacities, reduce the possibility of "death node" and redundant network coverage, and extend network lifetime. This paper expects to reach a balance between residual capacity of any sensor node and sensing range in the monitored area (where Wi is the residual capacity, =i is the sensing range and A is the monitored area), Wi < Wj && =i < =j (4) A = =1 ⋃ =2 · · · ⋃ =i · · · ⋃ =n = i=n⋃ i=1 =i (5) In the monitored area A, nodes should satisfy (3)(4) to reach the requirement of coverage. This paper also turns energy consumption and coverage balance strategy to a non-linear and 880 L. Zhu, C. Fan, H. Wu, Z. Wen multi-target optimization problem:   0 < Covmax = f( i=n∑ i=1 covi) ≤ A 0 ≤ Wmin = ϕ( i=n∑ i=1 ∆wi) ≤ nW0 Tmax = max(min(T1,T2 · · ·Tn)) ζmin = Θ( i=n∑ i=1 ζi) (6) Given that path loss, node residual capacity [9] and monitored area are key factors [10–12] to realize network performance optimization and extend network lifetime, the constraint conditions are dealt with according to mathematical programming so that the optimized network has high accuracy. Construct the optimized model with maximizing the network lifetime as the target. The optimized model is restricted to constraint conditions. Adjust the monitored area, and optimize the network lifetime and the overall network overhead. In the non-linear and multi- target optimization, optimized factors are not in the optimal state, but the network overhead is almost the minimum and the convergence has high accuracy. 3 Energy coverage strategy Coverage strategy of the wireless sensor network optimize network energy consumption while improving network coverage [13–15]; in addition, they should also balance energy consumption of single node and overall network energy consumption. If nodes with lower energy are assigned the same work load as nodes with higher energy, the former will end in “premature death” and the transmission and reliability of the entire network will be undermined. To reach the balance of two optimization coverage strategies and solve the contradiction between individual nodes and the overall performance, this paper proposes a coverage optimization strategy for WSN based on energy-aware, namely Energy-aware Coverage Strategy (EACS). The strategy is mainly divided into two phases: one is the energy-aware phase in which the sensing field of each node is confirmed by probability according to residual capacity of sensor node in the monitored area. The second is the network coverage adjustment phase where the sensing radius of node is adjusted effectively according to the sensing field of each node and the overall network coverage to lower the redundant coverage, reduce redundant coverage and unnecessary energy consumption, so as to extend network lifetime. 3.1 Energy-aware phase In the wireless sensor network, sensor nodes in the monitored area transmit information through awareness coordination. Given that sensor nodes have limited energy, this strategy concerns not only about energy consumption of any node in the area, but also the equilibrium of the whole network energy consumption. As time goes by, survived nodes may suffer from breaking the equilibrium of energy consumption due to signal interference, resulting in the change of node residual capacity. Considering that node residual capacity is related to node sensing range, as shown in Fig.1, reasonable sensing range is set up for each node to balance the network energy consumption and finally, to extend network lifetime. Coverage Optimization Strategy for WSN based on Energy-aware 881 In the energy-aware phase, after time t of working, the relationship between the electric quantity Wi consumed by node si and the sensing radius Ri of sensing field Ai is (k is a constant value): Wi = kR 2 i (7) For any two neighboring nodes si and sj, their residual capacity is Qi and Qj respectively. After time t, their electric quantity is consumed simultaneously. The sensing radius between two nodes fits expression (2): Ri = d(i,j) · √ Qi√ Qi + √ Qj (8) Where d (i, j) is the Euclidean distance between nodes. According to expression (8), the sensing radius and residual capacity of si and sj have the following relationship: Ri : Rj = √ Qi : √ Qj (9) si sj d(i,j) Rs Ri d(i,j) Rj i(j) 0 sj d(i,j) sensor node Figure 1: Node sensing range Theory 1 The sensing filed of node Si is Ω, its awareness is radius Ri. Ω falls in the set γ(γ1,γ2, · · ·γi) and any γk satisfies d(si,γk) < Ri. If the probability of X(X ∈ Ω) sensed by any node in γ is p, there are Pxsensing fields in Ω on average that are sensed by γ. Suppose the sensed set in the sensing field Ω is Π{x1,x2 · · ·xn}, and these nodes are inde- pendent from each other following even distribution. If k(k < n) nodes in being sensed by γ is called event X, so X follows binomial distribution, namely X ∼ B(n,p). P{X = k} = Cknpk(1 −p)n−k where E[X] is the expectation of X. When n → +∞, the average awareness in Ω is: lim n→+∞ E[X] n = p So there are psensing fields on average that are sensed by γ in Ω. 3.2 Network coverage adjustment Select any node Si in the monitored area, according to the energy-aware strategy, acquire the energy-aware range and node residual capacity of last phase and send the information such as sensing radius and energy consumption to the neighboring node. According to awareness 882 L. Zhu, C. Fan, H. Wu, Z. Wen intensity, confirm the sensing range and monitored area of each node. When there produce holes in the monitored area, re-distribute the nodes. Given that the sensing field may overlap, when no new holes are produced, adjust the sensing radius based on node residual capacity to wipe out redundant monitored area, so as to reduce unnecessary energy consumption. Detailed steps are described below: Step 1ŁşDivide the monitored area. For any node in the area, connect it with neighboring nodes and form the minimum triangle network as shown in 2(b). Subject each triangle to perpendicular bisector and connect the lines, as shown in 2(c). Finally, monitored area A consists of many regional polygons, as shown in 2(d). node distribution the minimum triangle network subject each triangle to perpendicular bisector Regional polygon Figure 2: Distribution of monitored area Step 2: detect irreparable holes and nodes filling. Firstly, detect holes on the edge of mon- itored area A. The polygon which cannot be fully covered by the minimum triangle network is called the edge monitoring polygon. Take the vertex of the polygon at the edge of the monitored area as the circle center and twice the maximum sensing radius of nodes Rmax as the radius to draw a circle. Places that cannot be covered by the circle are irreparable holes and need to be filled. New nodes should fill in the middle of the vertex and original nodes. For any side of the triangle in the network < S1,S2 >, if the side is twice more than the maximum sensing radius of nodes, namely L(S1,S2) > 2Rmax, it is also considered that there are irreparable holes. Thus, we should fill node in the middle of the triangle side. After nodes are filled, repeat step 1 to construct the new triangle network and the monitored polygon. Go on to step 3 until no new nodes should be filled. Step 3: detect regional holes. Confirm the initial sensing radius of nodes according to ex- pression (9). Construct the network model with the minimum triangle and detect the trian- gle network. To be specific, judge whether a minimum triangle S1S2S3 has sensing holes. If one side of the triangle < S1,S2 > is longer than the total sensing radius of two tip nodes, namely L(S1,S2) > R(S1) + R(S2), as shown in 3(a), then 4S1S2S3 has holes. If three sides fit L < Ri + Rj, in other word, two circles intersect, the premise for no holes is that any circle in- tersects with the overlapping part of the other two circles. That is to say, the intersection of any two circles should fall within the sensing radius of the third one. Otherwise, 4S1S2S3 has holes, Coverage Optimization Strategy for WSN based on Energy-aware 883 as shown in 3(b) and 3(c). When all minimum triangles fit this rule, a set with holes is formed. Calculate the number of nodes that have appeared, and form the hole-node set {Si, . . .,Sj}. S1 S3 S2 R1 R2L(S1,S2) S1 S3 S2 R1 R2L(S1,S2) S1 S3 S2 the total sensing radius < the triangle side the intersection circles with hole the intersection circles without hole Figure 3: Network hole detection Step 4: hole repair. Start repairing from the nodes that appeared the most times in the statistics in step 2. If two nodes appear the same times, the one with higher residual capacity is adjusted primarily. For example, Si is the first one to be filled. Increase the transmitting power of Si, by adding its sensing radius, until all triangles with Si as the tip node do not have holes. Repeat step 4 until all holes are filled, as shown in Fig.4 (a) WSN with coverage hole (b) WSN without coverage hole Figure 4: Coverage hole repair Step 5: wipe out the redundant holes. Get the complementary set of the triangle set with holes in step 2, namely the minimum triangle set without holes. As the initial sensing range of nodes is big, or two neighboring nodes are close to each other, there may be redundant coverage. Related nodes with relatively large sensing radius may consume unnecessary energy. Thus, it is necessary to adjust the sensing range of nodes in the complementary set. Firstly, calculate the number of nodes that appeared and rank them from the most appeared to the least appeared. Priority is given to nodes with less residual capacity. Reduce the sensing radius of nodes gradually until no new holes appear in the triangle area. Repeat step 5 to all nodes in the complementary set. 884 L. Zhu, C. Fan, H. Wu, Z. Wen 4 Analysis of algorithm performance 4.1 Analysis of coverage quality Conclusion 1 If the awareness intensity probability of j by node S in monitored area A is P{τ < Φ(j) < 1}, where Ss1,s2 . . .sn is the node set and τ is the threshold of awareness intensity. When and only when the expectation value of network coverage γ in A is bigger than γ0 (where γ0 is the coverage threshold), namely E[γ] > γ0, the coverage in A meets the requirement of the network coverage. Coverage connectivity is one of the important indicators to measure the service quality of wireless sensor network. According to the model, the probability of awareness intensity for node q in monitored area A is P{τ < Φ(j) < 1}. Elements in the node set S form a connected graph G(V,E), where V is the node set, and E is the side set. Nodes are independent from each other. The connectivity degree of G is Ξ(G) (n is the number of selected nodes). lim n→+∞ P{Γ(Sn,V, Ξ(G)) ≥ k} = 1 When n approaches infinity, the awareness probability of node in the connected graph in- creases. So does the connectivity degree. According to the definition, the expectation value of network coverage γ is: E[γ] = E[ ∫∫ A Cov(i)dA/‖A‖] = P{τ < Φ(i) < 1} > γ0 Select the sub-areas in the monitored area randomly. According to the probability event, when the connectivity degree has a high probability, the node sets in the area are all connectivity set and the network coverage would also meet the expectation. 4.2 Analysis of network lifetime Conclusion 2 Distribute n sensor nodes in monitored area A. The node set is S = {s1,s2. . .si. . .sn} and si and sj are neighboring nodes. The neighboring node set of si is Ni. When the network parameters are set the same as that of the network model, the network lifetime is Max ti (ti is the lifetime of node, i = 1, 2......n). Node si can be any node in the network. Its neighboring node set is Ni = {sj,sk. . .. . .sm}. Select one neighboring node sj. At this moment, the communication distance of two nodes is within the effective sensing range of si. So the energy consumption between nodes is: Q = ∑ j∈Ni τijηij + ∑ j∈Ni νηji (10) where ηij is the information transmitted by si to sj. τij and νji are energy loss factor when node receiving and transmitting information. Under normal condition, the lifetime of si is: ti = $i Qi (11) where $i is the residual capacity of si. According to expression (10) and (11), the network lifetime Tlife is: Tlife = Max{t1...ti...tn} Coverage Optimization Strategy for WSN based on Energy-aware 885 5 Simulation experiment analysis Through simulation experiment, this paper makes a comparative analysis on algorithm per- formance. The setting is described as follows: place 120 sensor nodes in the monitored area 100 × 100m2. The sensing radius is 5-20m and the initial electric quantity is 200J. The ratio of energy consumption of working, leisure and dormant state is 24:5:0.02. EACS algorithm is compared with the distributed random algorithm and PAYY algorithm (Proposed Approach of yourim y) proposed by literature [5] to assess the performance of EACS algorithm. Network coverage, network overhead and network lifetime are important parameters in the comparison. Table 1: Experiment parameter Parameter Value M 100m*100m N 120 To 120s W 0.2J Node Sampling Frequency 1Hz R 10m Initial Energy 200J Minimum Energy Limit 0.02J Ro 5m Rm 20m In the initial state of network operation, the performance parameters of all nodes in the monitored area are set the same, as shown in Fig.5(a). After working for some time, there produce difference on nodes. Adjust sensing radius of nodes according to their energy consumption. Nodes adjust themselves according to residual capacity and relationship with neighboring nodes, as shown in Fig.5(b). In the whole operation, due to differences between nodes and the change of sensing range, there may be redundant areas or coverage holes. Hole detection is conducted followed by hole repair and redundancy elimination in other to enhance the coverage, as shown in Fig.5(c). (a) (b) (c) Figure 5: Network coverage Network coverage ratio refers to the ratio of the effective coverage area by sensor nodes against the overall monitored area. For Random algorithm, operate it for 50 times and get the average coverage. Fig.1 shows the network coverage under different network scales of different algorithm. From Fig.6, it is known that with the increasing of the number of sensor nodes, 886 L. Zhu, C. Fan, H. Wu, Z. Wen the network coverage also increases. Under the same number of working nodes, the coverage of Random algorithm is relatively low. When 100 nodes are started, EACS algorithm can reach 90% coverage, 15% higher than Random algorithm. EACS algorithm is better that the other two algorithms in terms of coverage, because such algorithm can adjust the sensing radius according to residual capacity and reduces redundant monitored area. When sensor nodes increase in number, the coverage would increase rapidly. Figure 6: Working nodes and network coverage Figure 7: Coverage changes over time Fig.7 shows network coverage of three algorithms changing over time. In the initial operation of network, the network coverage changes relatively slowly with little divergence. As time goes by, EACS algorithm witnesses the increase of network coverage, though the energy-aware and hole detection require some energy consumption. From Fig.7, it can be told that the coverage curve of EACS algorithm drops steadily, reflecting that the network energy consumption is distributed evenly. Fig.8 compares nodes location of PAYY and EACS algorithm under the same initial settings and after the network operates for some time. Working nodes of EACS algorithm is more evenly distributed than those of PAYY algorithm. This is mainly because EACS algorithm adopts energy-aware model. The sensing radius of nodes is adjusted according to real situation of the network, so that the energy can be evenly distributed and the network lifetime can be extended. Figure 8: Working nodes distribution Coverage Optimization Strategy for WSN based on Energy-aware 887 Conclusion This paper proposes a coverage optimization strategy for WSN based on energy-aware, namely EACS algorithm. It can reach energy balance of nodes in the wireless sensor network. Re-assign tasks of awareness according to working nodes residual capacity, and adjust the sensing radius of nodes according to probability awareness in order to repair the network holes and redundant areas, improve network coverage and reduce redundant coverage and overall network consump- tion. Results show that the method proposed in this paper takes node residual capacity as an important factor of adjusting sensing field to reduce the burden of nodes with little energy and save them from being "null" too early. Otherwise, it may affect the connectivity and the net- work lifetime. At the same time, when the coverage holes are under repair, adjust the redundant coverage so that the balance of coverage reaches a reasonable range and the network coverage satisfies actual need. As a result, the unnecessary network energy consumption can be reduced effectively, a balance between the network coverage and energy consumption can be reached and the network lifetime can be improved. Acknowledgments The work presented in this paper was supported by the National Natural Science Foundation of China (Grants No. NSFC-61471067) and the National Natural Science Foundation of China (Grants No. NSFC-61271257). Fund for the Doctoral Program of Higher Education of China (Grants No.20120005110002).The work presented in this paper was supported by National Great Science Specific Project(Grants No. 2012ZX03005008) Bibliography [1] Mohamadi H., Ismail A.S., Salleh S. (2014); Solving target coverage problem using cover sets in wireless sensor networks based on leraning automata. Wireless Personal Communications, 75(1): 447-463 [2] He S.B. et al. (2013); Barrier coverage in wireless sensor networks: From lined-based to curve-based deployment, Marco A, ed. Proc. of the 32nd IEEE Int’l Conf. on Computer Communications , 10(9): 470-474 [3] Chen J.M., Li J.K., Lai T.H. (2013): Energy-Efficient intrusion detection with a barrier of probabilistic sensors: Global and local; IEEE Trans. on Wireless Communications, 12(9): 4742-4755 [4] Wu K., Gao Y., Li F., Xiao Y. (2005); Light weight deployment aware scheduling for wireless sensor network, ACM/Kluwer Mobile networks and Applications(MONET), 10(6):837-852. [5] Bulusu N., Heidemann J., Estrin D., Tran T. (2004); Self configuring Localization Systems: Design and Expermiental Evaluation, ACM Transac-tions on Embedded Computing Systems, 3(1): 24-60. [6] Khedr A.M., Osamy W. (2011); Minimum perimeter coverage ofquery regions in a hetero- geneous wireless sensor network, Information Sciences, 181(15):3130-3142. [7] La GuilingWang, Guohong Cao T.P.(2006); Movement-assisted SensorDeployment, IEEE Transactions onMobile Computing, 5(6): 640-652. 888 L. Zhu, C. Fan, H. Wu, Z. Wen [8] Y. Yoon (2013); An efficient genetic algorithm for maximum coverage deployment in wireless sensor network, IEEE Transactions on Cybernerics, 45(5):1473-1483. [9] Giuseppe Anastasi et al(2009); Energy Conservation in Wireless Sensor Networks: A Survey, Ad Hoc Networks, 7(3) : 537-568. [10] Martins F.V.C. et al. (2011). A hybrid multiobjective evolutionary approach for improv- ingthe performance of wireless sensor networks, IEEE Sensors Journal, 11(3): 545-554. [11] Wang D, Xie B, Agrawal DP (2008); Coverage and lifetime optimization of wireless sensor networks with Gaussian distribution, IEEE Trans. on Mobile Computing, 7(12): 1444-1458 [12] Amato G., Chessa S., Gennaro C., Vairo, C. (2011); Efficient detection of composite events in wireless sensor networks: Design and evaluation, Proc. of the IEEE Symp. on Computers and Communications, 10(11):821-823. [13] Yourim Y., Yong H.K. (2013); An efficient genetic algorithm for maximum coverage deploy- ment in wireless sensor network, IEEE Transactions on Cybernetics, 45(5): 1473-1483. [14] Hossain A., Chakrabarti S., Biswas P.K. (2012); Impact of sensing model on wireless network coverage, IET Wireless Sensor Systems, 2(3): 272-281. [15] Mini S., Udgata S.K., Sabat S.L. (2014); Sensor deployment and scheduling for target coverage problem in wireless sensor networks, IEEE Sensors Journal, 14(3): 636-644.