International Conference on Information and Communication Technology and Its Applications (ICTA 2016) Federal University of Technology, Minna, Nigeria November 28 – 30, 2016 An Optimized Wireless Sensor Network Deployment Using weighted Artificial Fish Swarm (wAFSA) Optimization Algorithm A.T. Salawudeen1, A. O. Abdulrahman2, B. O. Sadiq3, and Z. A. Mukhtar4 Department of Electrical and Computer Engineering, Ahmadu Bello University, Zaria, Nigeria 1 tasalawudeen@abu.edu.ng; 2aaolaniyan@au.edu.ng; 3bosadiq@abu.edu.ng; 3zmabubakar@abu.edu.ng Abstract—This paper presents an improved method for The Artificial Fish Swarm Algorithm (AFSA) has been deployment of wireless sensor networks using weighted widely adopted in solving various optimization problems Artificial Fish Swarm Algorithm (wAFSA). In wireless sensor because of its numerous advantages such as; ease of networks, optimal deployment of sensor nodes is a very critical implementation, adaptive search ability, high convergence and important factor in obtaining efficient and reliable quality speed, ability to effectively acquire global solutions while of service. An improved approach called inertial weight is first avoiding local extreme solutions [11-13]. These advantages introduced into the standard artificial fish swarm algorithm to are due to a combination of chasing, preying and swarming adaptively select the visual and step sizes of the artificial fishes behaviour of the artificial swarm of fish while searching for before the preying, swarming and chasing behaviours of the global solution in multi-modal complex optimization wAFSA were used to randomly and optimally deploy a total of sixty sensor nodes in a network coverage area of 60 square domain problems. meters. The proposed method was evaluated using some performance metrics such as network coverage and mobile II. LITERATURE REVIEW node, network coverage and iteration. The method was also The AFSA has been used by several authors in the compared with results obtained using standard AFSA which deployment and tracking of WSN and a significant showed that wAFSA performed better and can be used to performance when compared with other algorithms have adequately improve the scalability of wireless sensor networks. been observed. The paper presented in [14], proposed the use of AFSA for wireless network deployment but did not Keywords-wireless sensor network; artificial fish swarm algorithm; inertial weight consider network energy efficiency based on node delivery time. In 2012, Ma and Pu [15] used niching Particle Swarm Optimization technique for energy distance aware clustering I. INTRODUCTION protocol with dual cluster heads in wireless sensor networks. The research in [16] used artificial fishes based AFSA for The deployment of sensor nodes is a very important cooperative search and rescue in underwater wireless sensor factor which affects the quality of wireless sensor networks nodes. However, effective coverage area utilization, while (WSN). Poor deployment of sensor modes in WSN could optimally deploying mobile nodes in WSN still remains a lead to low mode density which in turn causes major concern for researchers in the area of wireless sensor communication gap and high node density which causes networks. In this paper, we present a modified AFSA called message collision and retransmission, signal intrusion and weighted Artificial Fish Swarm Algorithm (wAFSA) for cramming, huge energy consumption, etc. thereby leading to deployment of sensor nodes in WSN. As presented in [11, challenges in scalability, stability, distributed architecture, 13] an inertial weight was first introduced into the AFSA to energy consumption, and autonomous operation of the WSN adaptively select its parameters before it is used for the [1]. The introduction and development of Artificial deployment of WSN so as to obtain a maximized value of Intelligence (AI) has been given significant research the objective function. The results obtained were then attention in engineering and related disciplines. These compared with that of standard Artificial Fish Swarm artificial intelligence (heuristic and metaheuristics) Algorithm. algorithms have shown robustness and strong ability in solving problems such as WSN deployment and target tracking as shown in [2-4] and [1]. These algorithms include III. WIRELESS SENSOR NETWORK MODEL Particle Swarm Optimization (PSO) [5], Artificial Bee Colony (ABC) [6], Ant Colony Optimization (ACO) [7], For the essence of simplicity and easy applicability, this Firefly Algorithm (FA) [8], Bacterial Foraging Optimization paper considers a wireless sensor network having fixed and (BFO) [9], Artificial Fish Swarm Algorithm (AFSA) [10] mobile sensor nodes. Sixty (60) random nodes were and so on. Several modifications of these algorithms have generated in a square coverage area of 60m  60m . This also been presented by different researchers in order to number of fixed and mobile sensor nodes are represented as improve their performance and address some of their N and D for easy adaptability into the optimization algorithm challenges when solving several optimization problems. that would be used for deployment of these sensor nodes. 203 International Conference on Information and Communication Technology and Its Applications (ICTA 2016) The significance and important of doing this is to create a IV. NETWORK COVERAGE AREA matrix search space which is equivalent to the network The Network Coverage Area (NCA) is one of the coverage area for the sensor nodes to be deployed. N important indices used to measure the strategy of the represents the population of sensor nodes to be deployed wireless sensor network deployment. The network coverage while D represents the dimension of deployment which can area we considered in this work was the ratio of the whole be termed as the search space usually in an optimization area that can be covered by the nodes in the total node-aware problem. Therefore, the combination of all nodes in their region and the total area of the monitoring region. Due to the respective location within the coverage area can be fact that monitoring the environment can be quite represented as: complicated, the probability measurement model described in equation (5) was adopted. In order to minimize energy N  {n1,1 , n1,2 ,...n i, j } (1) consumption, the following assumptions were made:  Network nodes are ideal (i.e. nodes in the network Where ni , j represent the ith node in the network, in the have the same communication radius RC and the jth location of the network. In this case, i has the same size same sensor radius R S while RC  RS because as N and j has the same size as D. where N is the number when the communication radius between the nodes of sensor nodes and D is the dimension of deployment. is greater than two times the sensor radius of nodes, Given a random node M with a position ( x, y ) in the then the current networks are connected).  Coverage is considered as a metric for the network coverage area and a specific node Q having a measurement of quality of service of a sensor position xi , y i within the monitored or specified area, the network. distance between the nodes can be calculated as [1]:  At the initial stage of network deployment, all nodes are randomly distributed in a square monitoring area whose length is N, while the coordinate range of the X ( M , Q)  Q  M  ( x i  x ) 2  ( y i  y ) 2 (2) monitoring area is from (0, 0) to (N, D) whereby the D is the dimension of the coverage. The detection probability of any node Q by another node  Every node obtains the coordinate position M can be calculated based on the probability measurement information of itself and its neighbouring nodes. model adopted from the work of [14, 1]. This can therefore  In order to reduce energy consumption, the be obtained as: movement of nodes is virtual when running the algorithm until after the end of the algorithm, when nodes move to the best location based on the 0 Rs  Re  X (M , Q) physical location for a single time period.     The model and physical structure of every node is Pp (Qi )  e Rs  Re  X (M , Q)  Rs  Re (3) the same and similar. 1 Rs  Re  X ( M , Q)  V. WEIGHTED ARTIFICIAL FISH SWARM ALGORITHM The Artificial Fish Swarm Algorithm (AFSA) was developed based on the intelligent random behaviours of Where R s is the perceived radius of various elements in school of fish which are preying, swarming and chasing the network, Rs is the uncertainty factor within the behaviours. It has been established that certain parameters of AFSA such as visual distance and step size have critical measurement range of the nodes for 0  Re  Rs ,  &  influence on the performance of the algorithm. For example, are measured parameters with respect to the physical devices when the visual distance is very large, the algorithm has a being used and  is the input parameters which can be strong ability in obtaining a near optimal global solution and defined as [1]: when the visual distance is small, the algorithm has a strong local solution searching ability. Furthermore, the bigger the step size, the faster the speed of convergence of the   X (M , P)  ( Rs  Re ) (4) algorithm and vice versa [11, 13]. This property of the artificial fish swarm algorithm necessitated us to employ the use of weighted Artificial Fish Swarm Algorithm (wAFSA) Therefore, the joint detection probability of multiple [11, 13] in which an inertia weight was introduced to sensor nodes when conducting simultaneous measurement adaptively select the visual distance and step size of the simultaneously can be obtained using [1]: AFSA to suite our problem formulation at every iteration The ultimate goal of every fish in water is to discover Pp (Q)  1   (1  Pp (Q, M )) available regions with more food either by vision or by sense (5) through intelligent behaviours such as preying, swarming wi W and chasing. This idea was used in modelling the behaviour of artificial fishes such that the environment where an artificial fish lives is mainly the solution search space of the optimization problem and this defines the state of the fishes. 204 International Conference on Information and Communication Technology and Its Applications (ICTA 2016) The basic idea of AFSA is to imitate the preying, swarming f  c   max[C( H ')] (8) and chasing behaviour of fishes. These individual behaviours of AFSA can be represented based on the following rules: Begin  Synergic rule: Here, a basic communication is maintained between each individual artificial fish. If Initialize Parameters an individual fish in the searching swarm receives a call sent by another individual, it moves forward to Start counter the calling individual’s position by a step with a Foraging certain probability Behavior  Reconnaissance move rule: If the individual artificial fish does not receive a call, it implements Apply Inertial Succeed? NO Weight reconnaissance (surveying) according to its own swarm historical experience. If it however finds a YES better goal, it then moves forward to that position by Preying a step using its respective probability. Behaviour  Stochastic rule: For this rule, if the searching NO individual fish does not receive a call or does not Succeed? find a better goal, it moves in steps randomly such YES Apply Inertial that if it then finds a better goal during the step Weight Swarming Apply Inertial Behaviour movement, it sends a call to the entire fish swarm. Weight Counter = Counter+1 NO For any optimization problem, the objective function Succeed? under consideration can be said to have a solution space YES Dimension-D and the swarm is initialized with P-population Swarming of artificial fishes, such that the state of one artificial fish can Behaviour be formulated accordingly. Detail description and relevant NO mathematical model for the implementations of wAFSA can Succeed? be found in [11, 13]. YES The flowchart of the weighted Artificial Fish Swarm Algorithm (wAFSA) as implemented in this paper is shown Update Bulletin in Figure 1. Condition NO Satisfied VI. PROBLEM FORMULATION YES The problem formulation was done based on the network Output area. Here, N.H and H ' represent a set of total nodes and End a set of active nodes respectively. These constitute the total number of nodes that can be represented as randomly generated artificial fishes within the solution search space. Figure 1. Flowchart implementation of wAFSA The inspection region of H and H’ are correspondingly G and G’ in which Gi is the inspection region of the ith Equation (8) represents a maximization objective function which was formulated as a function of network artificial fish (node). The network coverage area is region optimization. The best solution of this objection formulated as follows [17]: function returns the optimum value of the network coverage area in which the location of all the sensor nodes are H' (H ' )  (6) optimized in terms of deployment while consuming minimal N energy in the process. The quality of the network coverage area is then defined VII. SIMULATION APPROACH as: The wAFSA which was previously described was used to optimize the objective function shown in equation (8). Here, G i 1, 2 ,....N i mobile sensor nodes are represented as artificial fishes in a C(H ' )  (7) cluster form and the cluster head node travels based on the G nature of the objective function while following the inertial preying, swarming and chasing movement behaviour. The step by step procedure of the proposed method which is The objective function of the coverage quality which was represented in Figure 1 is as follows: formulated using equation (7) can be defined as:  Initialize all the parameters for wAFSA; 205 International Conference on Information and Communication Technology and Its Applications (ICTA 2016)  Initialize the wireless nodes and population of appears to have little or no significant effect on the wAFSA to be deployed within the coverage area’s performance of the algorithm. Also, for the standard AFSA, dimension; the network attains it maximum coverage of about 75.9%  Initialize iteration start point itr  0 ; when the number of iteration is 20, thereafter, increase in  Calculate the fitness based on the current locations iteration also, appears to have no significant effect on the of the nodes in the coverage area and update the network. it can be concluded that, even though the wAFSA score board with the best individual; performed much better in terms of coverage area, the  Select a new state randomly and evaluate its fitness standard AFSA attains it maximum much faster. The improvement of wAFSA over AFSA can be attributed to the  Evaluate the best fitness based on inertial preying, adaptive behaviour of the inertial weight which enables to swarming and chasing behaviour of wAFSA; explore larger solution space before it converges. The figure  Evaluate the current fitness of the population (nodes) showing the performance evaluation using the coverage area and compare with previous fitness and then update; and number of mobile nodes is given in Figure 3.  If the stopping criteria is met, output the current best From Figure 3, it can be observed that, the network fitness, else increase iteration by 1 and go back to coverage increases with increase in number of mobile nodes. step (5) until best solution is obtained. For the weighted AFSA, the network coverage attains a  Display best solution and configuration. maximum of 80.20% at mobile nodes of 50 and then, tends towards stability. Also, for the standard AFSA, a maximum The details of the parameters used for simulation is network coverage of about 79.9% was obtained at the same presented in Table 1. Since the coverage area under network nodes with similar network behaviours. The consideration is 60 X 60 square metres, 60 wireless sensor insignificant different in the network using both weighted nodes were also considered to be deployed randomly and AFSA and standard AFSA indicates the scalability of the thus, the problem dimension of 60 was used. network using both approach. TABLE I. PARAMETERS SETTINGS FOR NETWORK COVERAGE 0.78 SIMULATION USING WAFSA Weighted AFSA Standard AFSA S/N Parameters Description Values 0.76 1 Visual Visual Distance 20m 0.74 2 Step Step Size of AF 10m Coverage Area(%) 3 Crowd Crowd factor of 0.6 AF 0.72 4 N Number of Nodes 60 5 RS Sensor Radius 4m 6 RC Commutation 15m 0.7 Radius 7 R Uncertain Factor 2m 8 Try_num Number of trial 100 0.68 9 Kmin Cessation Value 40 10 K Current Value Iterative 11 KO Initial Value 200 0.66 0 5 10 15 20 25 30 35 40 45 50 12 R Attenuation 0.95 Number of Iteration Constant 13 Pm Measured 0.8 Figure 2. Network Performance Against Number of Iterations Probability 14 K Actuation 0.4 Constant VIII. RESULTS AND DISCUSION Simulations were carried out on MATLAB R2015b and the results are presented based on the performance of the model. The performance evaluations carried out were to check the relationship between network coverage certain parameters such as network node and number of iteration. Readers should note that, the best of every twenty (20) run of the model were presented. The Figure 2 shows the superimposed result obtained for the network coverage against the number of iteration for both the weighted wAFSA and the standard AFSA. It can be observed from Figure 2 that, the performance of the network coverage increases as the number of iteration is increased from 0 in a step of 5 to 50 iterations. It can be deduced that the wAFSA attained the highest network coverage of about 77.95% which occurs when the number of Figure 3. Network Coverage Performance Against Number of Nodes iterations is 25. At this point, increase in number of iteration 206 International Conference on Information and Communication Technology and Its Applications (ICTA 2016) IX. CONCLUSION AND FUTURE WORK [7] M. Dorigo, V. Maniezzo, and A. Colorni, "Ant system: optimization by a colony of cooperating agents," Systems, Man, and Cybernetics, This paper presents an optimized deployment of WSN Part B: Cybernetics, IEEE Transactions on, vol. 26, pp. 29-41, 1996. using a modified artificial fish swarm optimization algorithm [8] X.-S. Yang, "Firefly algorithm, stochastic test functions and design (wAFSA). The model was simulated in MATLAB R2015a optimisation," International Journal of Bio-Inspired Computation, simulation environment and results show the superiority of vol. 2, pp. 78-84, 2010. the modified algorithm over the standard AFSA. The [9] K. M. Passino, "Biomimicry of bacterial foraging for distributed superiority of the modified algorithm is attributed to the optimization and control," Control Systems, IEEE, vol. 22, pp. 52-67, 2002. adaptive and dynamic behaviour of the modified algorithm [10] L. X. Lei., Z. J. Shao, and J. X. Qian, "Optimizing method based on in comparison the original AFSA. In our next research work autonomous animats: Fish-swarm Algorithm," Xitong Gongcheng the performance of the wAFSA will be evaluated on Lilun yu Shijian/System Engineering Theory and Practice, vol. 22, p. different target tracking and other optimization problems 32, 2002. such as data clustering, colour quantization, distributed [11] M. B. Mu'azu, A. T. Salawudeen, T. H. Sikiru, A. Muhammad, and generation etc. A. I. Abdu., "weighted Artificial Fish Swarm Algorithm with Adaptive Behaviour Based Linear Controller Design for Nonlinear Inverted Pendulum," Jounal of Engineering Research JER, vol. 20, pp. 1-12, March, 2015 2015. REFERENCES [12] A. T. Salawudeen, "Development of an Improved Cultural Artificial [1] S. A. Lawal, S. Garba, A. D. Usman, M. B. Mu'azu, and A. T. Fish Swarm Algorithm with Crossover," Masters Electrical and Salawudeen, "Optimal Deployment of Wireless Sensor Networks Computer Engineering, Ahmadu Bello University Zaria, 2015. (WSN) Based on Artificial Fish Swarm Optimization Algorithm," [13] A. T. Salawudeen and M. B. Mu'azu, "Stabilization of Inverted International Journal of Science and Engineering Investigations, vol. Pendulum System using Intelligent Linear Quadratic Regulator 4, pp. 45-51, 2015. Controller," in IEEE-Trans of 7th International Joint Conference on [2] T. Yang and T. Yong, "Short Life Artificial Fish Swarm Algorithm Computational Intelligence , Lisbon Portugal, 2015, pp. 325-333. for wireless sensor network," in Computational Problem-solving [14] W. Yiyue, L. Hongmei, and H. Hengyang, "Wireless sensor network (ICCP), 2013 International Conference on, 2013, pp. 378-381. deployment using an optimized artificial fish swarm algorithm," in [3] Z. Li, H. Zhang, J. Xu, and Q. Zhai, "Recognition and localization of Computer Science and Electronics Engineering (ICCSEE), 2012 harmful acoustic signals in wireless sensor network based on artificial International Conference on, 2012, pp. 90-94. fish swarm algorithm," Journal of Theoretical and Applied [15] D. Ma and P. Xu, "An Energy Distance Aware Clustering Protocol Information Technology, vol. 49, 2013. with Dual Cluster Heads Using Niching Particle Swarm Optimization [4] Y. Wang and L. J. Li, "An Improved Intelligent Algorithm based on for Wireless Sensor Networks," Journal of Control Science and the Group Search Algorithm and The Artificial Fish Swarm Engineering, vol. 2015, 2015. Algorithm," international Journal Optimization, Civil Engineering, [16] W. Zhao, Z. Tang, Y. Yang, L. Wang, and S. Lan, "Cooperative vol. 1, p. 15, 2015. search and rescue with artificial fishes based on fish-swarm algorithm [5] R. C. Eberhart and J. Kennedy, "A new optimizer using particle for underwater wireless sensor networks," The Scientific World swarm theory," in Proceedings of the sixth international symposium Journal, vol. 2014, 2014. on micro machine and human science, 1995, pp. 39-43. [17] H. A. Hashim, B. Ayinde, and M. Abido, "Optimal placement of relay [6] D. Karaboga, "An idea based on honey bee swarm for numerical nodes in wireless sensor network using artificial bee colony optimization," Technical report-tr06, Erciyes university, engineering algorithm," Journal of Network and Computer Applications, vol. 64, faculty, computer engineering department2005. pp. 239-248, 2016. 207