Research Article  Open Access
Energy Aware Simple Ant Routing Algorithm for Wireless Sensor Networks
Abstract
Network lifetime is one of the most prominent barriers in deploying wireless sensor networks for largescale applications because these networks employ sensors with nonrenewable scarce energy resources. Sensor nodes dissipate most of their energy in complex routing mechanisms. To cope with limited energy problem, we present EASARA, an energy aware simple ant routing algorithm based on ant colony optimization. Unlike most algorithms, EASARA strives to avoid low energy routes and optimizes the routing process through selection of least hop count path with more energy. It consists of three phases, that is, route discovery, forwarding node, and route selection. We have improved the route discovery procedure and mainly concentrate on energy efficient forwarding node and route selection, so that the network lifetime can be prolonged. The four possible cases of forwarding node and route selection are presented. The performance of EASARA is validated through simulation. Simulation results demonstrate the performance supremacy of EASARA over contemporary scheme in terms of various metrics.
1. Introduction
Wireless sensor networks (WSNs) are captivating growing attention because of their merit and aptness for autonomous and unattended monitoring in largescale applications. Examples of such applications range from terrestrial to aquatic, from earth to cosmic, from agriculture to jungle, and from home and offices to international borders [1]. Most of these applications require deployment of thousands of lowcost batteryoperated tiny sensors which are expected to stay operational for a longer period. Despite considerable advancements in different aspects [2–4], network lifetime, as yet, is one of the most prominent barriers in deploying WSNs for largescale applications because sensor nodes dissipate most of their energy in complex routing mechanisms.
Routing has been widely investigated in the context of mobile ad hoc network (MANET); however, due to unique characteristics, it is still a challenging issue in WSNs [5]. Most of the existing routing protocols either employ conventional [6–8] or nonconventional approaches [9–11]. Nonconventional approaches are based on natural systems such as bioinspired computing that is the focus of this work. In bioinspired computing, ant colony optimization (ACO) has been widely used to develop metaheuristic algorithms for combinatorial optimization problems such as asymmetric travelling salesman [12], graph coloring problem [13], and vehicle routing problem [14]. Metaheuristic algorithms such as tabu search [15], simulated annealing [16], iterative local search [17], evolutionary computation [18], and ant colony optimization [19] provide computational methods for heuristic optimization of a problem. These algorithms must be customized according to the application to find the optimal solution. For example, authors in [20] have advocated the adoption of ACO for MANETs. WSNs share some common characteristics with MANET (e.g., infrastructure less and dynamic topology); however, differences such as high node density and severe energy constraints [21] instigated us to investigate adoption of ACO for WSNs.
In dynamic changing conditions, routing in WSNs aims to deliver data from sensors to the sink in reliable and energy efficient manner. The focus is on finding the optimal path between source and destination. In our previous work [22], we have developed an analogy between foraging behavior of ants and routing process in WSNs. Routing can be handled with the help of ACO utilizing the information of neighbor discovery, routing and data transmission, and route maintenance. ACO mimics the real ant colony’s foraging behavior to address the combinatorial optimization problem of finding the best path on a weighted graph. The resulting shortest route mapping determined by the artificial ants can be applied to the optimization problem of energy aware routing. Indirect communication used by social insects to coordinate their activities is called stigmergy, directing towards pheromone deposition. By exploiting the stigmergy approach (responds to new and modifies the existing environment), a number of successful algorithms in such diverse application fields of combinatorial optimization are introduced for routing in communication networks. Initially, ant agents (packet) move by following a random decision based on link and node position. Routing table is updated based on the stochastic decisions made for pheromone deposition after they reached the sink/destination node, and their decision is based on sampling the complete paths. Proactive path is selected by the ant agents so that it could affect the topology change of the nodes in a sensor network and similarly the routing. Reactive paths are then used once the routing table is updated for optimal routing of the shortest path. The resulting optimal route can be determined by the ant agents to obtain a solution of the optimization. Forward ants maintain memory of each visited node while moving towards the sink thus avoiding loops formation. The ant’s internal memory helps to construct the goodness of a solution, contribute in each move, and manage feasibility of solutions. The backward moving ants retrace the path towards the source node. Figure 1 elucidates the phenomena.
(a)
(b)
The local and global pheromone are sharing help to build up the optimal solution. As far as the local pheromone information is concerned it involves the problem specific heuristic information, which is knowledge embedded in the pheromone trails involved at the beginning of the stochastic search process (ants generation, activity in search of paths, and pheromone evaporation).
However, the global pheromone concentration information deals with the shared local pheromone value that ultimately effect the ant’s decision while construction of a good solution to deposit additional pheromone information to collect useful global information. Reactive approach develops the multipath between source and destination in order to avoid link failure and robustness, but it becomes difficult when mobility among nodes causes frequent topological changes. Whereas the proactive approach is reducing the routing, overhead becomes infeasible in disruptive situations. The hybrid approach is combining both the reactive and the proactive approaches and utilizing their advantages helps to develop an optimized energy aware routing algorithm based on ACO.
The remainder of this paper is organized as follows. Section 2 describes some of the closely related works to our work. The working details of EASARA are described in Section 3. Section 4 proposed solution is discussed in Section 3 with detailed elaboration of proposed improvements in route discovery part of the baseline algorithm, simple ant routing algorithm. Also, the proficiency of EASARA algorithm over SARA algorithm in four scenarios that may arise during the forwarding node selection which are discussed is detailed in this section. Results and discussion are explained in Section 4 proceeded by conclusion in Section 5.
2. Related Work
Energy aware routing based on ACO has been widely investigated in the context of MANET [23, 24]. For example, ARA [25] is a reactive algorithm based on ACO that strives to reduce routing overhead. ARA consists of following three phases, that is, route discovery, route maintenance, and route failure handling. In route discovery, new paths are discovered with the help of FANT (forward ant) and BANT (backward ant). Each FANT containing a unique sequence number passes through intermediate nodes while moving data packets from a source towards the destination. Routing table entries consist of destination address, next hop, and pheromone value which is maintained by each node independently. Choosing a route among multiple paths depends on the link quality and usability. Routing paths established should be maintained (i.e., route maintenance) with the evaporation of pheromone value that indicates the quality of the link. Missing successive number of acknowledgments indicates nonavailability of a path that might be caused due to several factors such as node failure or mobility. Route failure handling deals with such situations that may call for reroute discovery, for example, finding alternative path. To conserve energy, availability of pheromone is considered while reroute is discoverd. This significantly reduces the overhead as it does not require an exchange of routing tables. However, the criterion for choosing a neighbor is not clear during route selection. Moreover, triggering a route recovery procedure based on missing an ACK may unnecessarily call for path discovery.
Similarly, AntHocNet [26] is another routing mechanism that effectively deals with network dynamics. It is comprised of four parts, that is, reactive path setup, proactive path maintenance and explorations, stochastic data routing, and link failures. In reactive path setup, each receiving node further unicasts or broadcasts the forward ants depending upon the pheromone table maintained at each node. A node only considers the first arrived forward ant that keeps the information of visited nodes on the path which helps the backward ant in backtracking. Proactive path maintenance updates exist and look for better routes. Data routing is based on stochastic decisions for load balancing. Link failure mechanism is triggered when nodes are unable to exchange unicast or hello messages. Like AntHocNet, SARA [27] is another mechanism that optimizes the routing procedure to reduce overhead. SARA employs a new controlled neighbor broadcast (CNB) mechanism in the route discovery where one of the neighboring nodes only broadcasts the received FANT further. It further reduces overhead by employing data packets to refresh the routes of active sessions. To limit the scope of recovery (i.e., route repair phase), SARA strives to find a new link between the two end nodes of the failed path instead of establishing a new route from the source to destination. SARA pursues a broadcast search in case the recovery process remains successful. Unlike these mechanisms, the focus of our work is WSNs where node energy is a prime design consideration while making routing decisions.
Recently, a novel routing protocol based on ACO was proposed in the context of WSNs [28]. To increase network lifetime through load balancing, the proposed protocol employs a dynamic and adaptive mechanism that considers engagement of a node in data forwarding and residual energy besides data latency. The efficiency of the proposed protocol in terms of network lifetime improvement is also shown.
Table 1 summarizes the literature given in Section 2. To reduce energy consumption and prolong network lifetime, a multipath routing protocol (MRP) based on clustering and ACO is proposed in [29]. MRP picks a cluster head (CH) among nodes based on residual energy and employs ACO to determine multiple paths between a CH and sink. Several routing factors such as energy consumption are considered by the CH for data transfer while dynamically choosing a path. Most of these works do not consider WSNs dynamics such as decay in pheromone, link failures due to node mobility or dysfunction. Unlike these works, the focus of our work is different.

3. Energy Aware Simple Ant Routing Algorithm
As stated earlier, most applications of WSNs employ sensors with limited nonrenewable energy and routing is a major energy guzzler that shorten network lifetime. To cope with this problem, we present a reactive energy aware simple ant routing algorithm (EASARA) which is based on ACO. Unlike most algorithms, EASARA strives to avoid low energy routes and optimize the routing process through selection of the shortest path (number of hops) with more energy. EASARA mainly consists of three phases, that is, route discovery, forwarding node, and route selection. We have improved the route discovery procedure of SARA and mainly concentrate on energy efficient forwarding node selection (FNS) and route selection so that the network lifetime can be prolonged. We have described four possible cases of forwarding node and route selection. The following describes the details of EASARA.
3.1. Route Discovery in EASARA
Unlike traditional ACO based approaches, EASARA adopts a controlled neighbor broadcast mechanism [27] in route discovery to avoid flooding the network with FANTs. In CNB based route discovery, each sensor node broadcasts the FANT to its direct neighbors, and only one of them further rebroadcasts the FANT to its neighbors. Randomly choosing a neighbor for “rebroadcasting” may not be feasible. Therefore, EASARA adopts a probabilistic approach in picking a neighbor based on energy and link cost. The probability of a neighbor with more energy and lower link cost is higher in FANT rebroadcast. First, it determines a threshold energy level based on average remaining energy of neighbor nodes that is calculated as follows. Each node continuously updates its energy and exchanges heartbeat messages with neighbors to establish and maintain a neighbor list containing updated information such as ID and energy level [30]. Suppose there are nodes in WSNs; the average energy consumption of the network is the sum of the divided by aswhere is the difference between initial and final energy which can be calculated as Since data transmission is a major energy consumer, therefore the energy consumed by each node is the sum of transmission and reception :The energy threshold can be calculated using (4) which averages the remaining energy of all the neighboring nodes () of a source :The neighbors having remaining energy greater than the threshold will become the candidates for FANT rebroadcast. The rationale is to engage high energy nodes in route discovery and hence improve network lifetime. Afterwards, the candidate nodes determine the link cost based on the number of times a node was previously selected for rebroadcast. The selection time is inversely proportional to the probability of a node to be picked in the route selection procedure. The probability to pick a neighbor for forwarding FANT towards destination is given by (5), which is based on energy, link cost, and usability:where represents the link cost between node to its neighbor node towards destination . This cost is based on the number of times () the link was previously used and on , that is, the number of neighbors of node ; is the energy of node and is energy of node . This rebroadcast of the FANT packet continues in a similar fashion until it reaches the destination. Table 2 elucidates the FANT packet used by EASARA. The rest of the route discovery procedure is similar to SARA.

Most of the fields are selfexplanatory. FANT_ID is a unique identifier for each session and is generated by the source node. The addresses (source, destination, and node) represent the IDs of the source, destination, and address of the current node. The neighbor selected for rebroadcast is listed under Next. The next two fields indicate the number of hops traversed from source towards the destination on a particular path and path ID. Average energy threshold is the average remaining energy of neighbors.
3.2. Four Scenarios of FNS for Comparison of SARA and EASARA Algorithms
The depletion of node energy that has been involved in various data and control sessions causes decrease in the lifetime of network. The algorithm EASARA helps to enhance the lifetime of the network by taking care of nodes energy. If we consider nodes energy then four scenarios occur where each shows a unique case in terms of energy utilized by a node and number of times a node is selected for discovering a route. The scenario is depicted clearly in Figure 2 where a node energy usage in terms of energy units and the number of times a node is selected earlier. Nodes communicate with each other through hello packets. The routes are discovered by calculating the threshold level of energy of the neighboring nodes to select the nodes for controlled neighbor broadcast procedure where the probability of neighboring nodes is calculated. The nodes which have greater energy level and higher probability are selected. The four scenarios which arise are as follows:(i)nodes with equal energy but different number of selection times,(ii)nodes with variable energy and variable used number of times with less number of nodes involved in controlled neighbor broadcast process,(iii)nodes with less energy and used less number of selection times,(iv)nodes with more energy and used less number of times.The algorithm details of EASARA along with the four scenarios are discussed below. Scenario depiction for the adaptation of EASARA is given in Figure 2.
3.2.1. Case 1: Nodes with Equal Energy but Different Number of Selection Times
Considering Table 7, source has two neighbors and and both are used with different number of selection times (). Previously in SARA algorithm the node with less number of selection times has a higher probability to be selected. Random decision is made if a node is used in equal number of times in SARA but in EASARA average energy is computed for node and node . Average energy calculation enables to decrease the processing for nodes which are below the threshold calculated from the average energy consumption. A threshold of 6.5 units is calculated and obtained from average energy of the neighboring nodes so the node having energy greater than 6.5 is selected to participate in CNB broadcasting FANT to calculate the link cost and probability. Link cost and probability are calculated from (5). For node , the link cost is 0.25 and link cost for node is 0.33 and so the probability for node is 0.43 and the probability to select node is 0.56. So node has probability decision greater than node ; node is selected to rebroadcast the FANT.
Table 3 shows a comparison calculation between the two algorithms.

3.2.2. Case 2: Nodes with Variable Energy and Variable Used Number of Times with Less Number of Nodes Involved in Controlled Neighbor Broadcast Process
Node is selected previously on the basis of higher probability and node energy. gets the energy information of its neighboring nodes; it calculates the energy threshold to be 6.5 with its three neighbors , and . So node and node can participate in CNB selected by the node . The link cost of the node is 0.5; for node the link cost is calculated to be 0.25 and probability will be 0.28 for node , and for node probability will be 0.71, so node is selected to rebroadcast the FANT among its neighbors, whereas when we compare SARA with proposed algorithm node calculates the link cost and probability for all its neighbors and only the neighbor with greatest probability will be selected to rebroadcast. EASARA makes the decision based on the average energy threshold value and calculates the probability of those nodes which are above a threshold energy level. Table 4 explains case 2.

3.2.3. Case 3: Nodes with Less Energy and Used Less Number of Selection Times
calculates the energy threshold to be 5.5; it selects node for calculating its link cost only. The link cost and probability is 0.2 and 1, respectively, whereas in SARA algorithm the nodes with less number of selection times are selected due to their higher probability among their neighbors irrespective of the fact that the node has less energy, and due to its number of less selection times it participates in the route discovery procedure without knowing the left energy level of the node. It may cause the network to have dead nodes which affects the network lifetime.
As the mobility in the network and four sessions are under consideration, this case occurs as a node is involved in a number of processes. Similarly in EASARA, only the node(s) in neighbor that will have energy above the threshold level is selected, so node only calculates the probability and link cost of that node whereas in SARA algorithm all neighbors link cost and probability will be calculated so processing energy will be consumed for each node. Table 5 gives an extended understanding to this scenario.

3.2.4. Case 4: Nodes with More Energy and Used Less Number of Times
This case resembles SARA algorithm initial session case when all the nodes have equal energy and link cost and probability will be calculated to have a load balance among the network nodes. Now node calculates the energy threshold to be 6 nodes for its two neighboring nodes and ; node is selected to calculate link cost to be 0.25 with probability 1, which ultimately reaches the destination; dissimilarly it reaches destination node. But in EASARA as shown in Table 6, only one node is selected based on energy threshold value for being further selected as the forwarding node as SARA calculates link cost and probability for each neighboring node.


Four cases have been discussed in it: firstly, if the node’s energy in neighbors is used equally but with variations in their link cost, that is, number of times a node is used in the network. Secondly, the nodes are used more but energy consumption is less as compared to neighboring nodes; thirdly the weaker neighbor cannot participate in calculating the link cost and probability to reach destination and the fourth case is the less the link cost is and the less the energy consumption will be. This will help to discover the paths where more energy efficient nodes update their energy level at each step to enhance the lifetime of the network.
3.3. Route Selection Criteria and Procedure
Once the routes are discovered, they are selected on the basis of a number of hops, energy of the node, and pheromone concentration. When multiple routes are discovered using a controlled neighbor broadcast then the route is selected on the basis of probability procedure. The route selection in EASARA differs from SARA algorithm. Two different metrics are used for the best shortest path selection. The only difference lies in the fact that now the routes are selected on the basis of nodes that have a particular energy threshold plus the number of time a node is selected. Number of hops from source to destination and the other pheromone level present in the link between nodes. SARA uses the hop distance, and the pheromone value as metrics whereas EASARA introduces another metric in route selection criteria which are node energy and number of hops:where the probability of the route selection between node and is dependent on the link cost where is the number of adjacent nodes. The link cost is calculated based on pheromone value between the nodes and number of hops. The pheromone value tells the fact that each node tracks the amount of pheromone on each neighbor link. A packet passes through the link between the nodes and ; the pheromone intensity ph increases by a factor . is the time when a packet crosses the link and increases pheromone level and is the time in which a packet crosses the link recently. is time when the pheromone decreases. is the pheromone decrease intensity. The and are used to optimize the link pheromone availability. The pheromone increasing intensity and decreasing intensity values are adjusted to tune the route availability based on the pheromone intensity. Table 7 gives the summary of this pheromone description.
4. Results and Discussion
For performance evaluation of our proposed algorithm, EASARA, with existing protocol SARA. We used NS2.31 simulator by considering simulation parameters as shown in Table 8. All the environmental parameters of both algorithms are same. The performance metrics which are used to evaluate the performance of both algorithms in route discovery and route recovery phenomena are as follows:(i)transmitted packets per node,(ii)received packets per node,(iii)total number of packets per node,(iv)energy consumption:(a)per node,(b)per selected path,(c)total network,(d)route recovery.Initial results during simulations for randomly selected paths for the proposed route discovery and route recovery procedure are presented. We have started the analysis based on the characteristics of typical sensor node architectures such as in receiving and transmitting the control packets during route discovery procedure. It helps to identify the various factors that affect system lifetime.

Figure 3 shows the packets transmitted per node where SARA has more value compared to EASARA. When any node in route discovery procedure communicates with its neighboring nodes for FNS, it generates packets called control packets. The average energy calculation of the neighboring node before controlled neighbor broadcast procedure helps to involve less number of nodes in link cost and probability calculations. This reduces the number of transmission packets among the nodes. The number of control packets that are transmitted during the route discovery procedure is analyzed in Figure 3. Also, the comparison of both competing algorithms in the number of received control packets during the route discovery procedure is depicted in Figure 4. That shows almost the same result of better performance of EASARA over SARA as shown in Figure 3.
Route discovery packets, route recovery, and route maintenance packets in the routing algorithm depict more overhead of the protocol. Overhead is exponentially related to the number of interacting nodes. Increase in control packets reflects in the form of decrease in the bandwidth available for the data packets to transfer among the nodes and to reach from the source to destination. This results in less reliability of data transmission. Same is the case that happened in the working of SARA which is using more control packets in route discovery, route recovery, and route maintenance processes. Hence, more energy is consumed per network node as is shown in the Figure 5.
Control packets are the primary source of energy consumption. Energy consumption by the node during the transmission of control packets is shown in Figure 6 which justifies the above facts that less number of packets transmitted between intermediate nodes for route discovery procedure is an effective way to save energy.
This is favored most in the situations where the nodes are mobile and distributed in the network area. SARA utilized more control packets for route discovery, so more energy is consumed per node as compared to EASARA by involving less number of nodes.
Similarly, Figure 7 shows performance of the node on the energy consumption while receiving control packets. The EASARA outperforms SARA algorithm, that is, the reflection of more entertainment of control packets by SARA compared to the proposed solution.
This is again demonstrated in Figures 8 and 9 in another way by figuring out the total energy consumption per node and of the overall network during the route discovery phenomenon, respectively.
Forwarding node energy consumption is calculated against five randomly selected paths during the simulation. From Figure 10 it is intuited that EASARA gives a better result than SARA. The reason of this improved performance is the reduction of energy consumption at each node during the set up phase as fewer transmissions occur in EASARA as compared to SARA, less transient nodes are involved in controlled neighbor broadcast and probability calculation, and less node to node delay from the overall packet energy from source to destination occurs.
In SARA, the route recovery process involves two phases. First the nodes recovery process is initiated locally by looking at the neighbor table entry. If it is not present then the node initiates the forward node selection. In another case if it fails then the source node initiates a route discovery procedure. SARA solves the link congestion problem already. We have taken the case of 5 randomly selected paths where a link break occurs and the node cannot find a node in its neighbor table and it calculates the probability of the next forwarding node to be selected. The path selection using the probability function that is based on the link cost and energy helps in reduction of overall energy consumption. That appears in the form of betterment in EASARA as compared to that of SARA. Our aim is to make the route discovered energyaware so that the lifetime of the network can be increased, whose achievement is depicted based on empirical results in Figure 11.
5. Conclusion
This paper has presented EASARA, an energy aware simple ant routing algorithm, to improve WSNs lifetime. EASARA is a reactive scheme based on ant colony optimization. To cope with limited energy problem, EASARA strives to avoid low energy routes and optimizes the routing process. EASARA mainly consists of three phases, that is, route discovery, forwarding node, and route selection. EASARA strives to avoid low energy nodes in route discovery and engages nodes with higher energy as forwarders. Moreover, it prefers to choose high energy routes with least hop count path to improve network lifetime and minimize delay. The four possible cases of forwarding node and route selection were also discussed. The performance of the EASARA was validated through simulation experiments. The experimental results validated the effectiveness and efficiency of EASARA compared to contemporary scheme in terms of various performance metrics. This scheme can also be used in M2M, in IoT, and in extension various cyber physical systems (CPS) [31–35], apart from the fact that a sensor based Big data will also benefit from such energy aware technique.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
Acknowledgments
The authors would like to extend their sincere appreciation to the Deanship of Scientific Research at King Saud University for funding this research through Research Group Project no. (RG#1435051).
References
 J. Yick, B. Mukherjee, and D. Ghosal, “Wireless sensor network survey,” Computer Networks, vol. 52, no. 12, pp. 2292–2330, 2008. View at: Publisher Site  Google Scholar
 A. A. Abbasi and M. Younis, “A survey on clustering algorithms for wireless sensor networks,” Computer Communications, vol. 30, no. 1415, pp. 2826–2841, 2007. View at: Publisher Site  Google Scholar
 M. Younis and K. Akkaya, “Strategies and techniques for node placement in wireless sensor networks: a survey,” Ad Hoc Networks, vol. 6, no. 4, pp. 621–655, 2008. View at: Publisher Site  Google Scholar
 C. Zhu, C. Zheng, L. Shu, and G. Han, “A survey on coverage and connectivity issues in wireless sensor networks,” Journal of Network and Computer Applications, vol. 35, no. 2, pp. 619–632, 2012. View at: Publisher Site  Google Scholar
 C. Li, H. Zhang, B. Hao, and J. Li, “A Survey on routing protocols for largescale wireless sensor networks,” Sensors, vol. 11, no. 4, pp. 3498–3526, 2011. View at: Publisher Site  Google Scholar
 S. Jabbar, A. A. Minhas, A. Paul, and S. Rho, “Multilayer cluster designing algorithm for lifetime improvement of wireless sensor networks,” Journal of Supercomputing, vol. 70, no. 1, pp. 104–132, 2014. View at: Publisher Site  Google Scholar
 A. A. Minhas, S. Jabbar, M. Zubair Aziz, and W. Mahmood, “QERT: querybased energyaware realtime routing for wireless sensor network,” in Proceedings of the IEEE 10th International Conference on Computer and Information Technology (CIT '10), pp. 2861–2866, Bradford, UK, JuneJuly 2010. View at: Publisher Site  Google Scholar
 M. Imran, A. Khan, and A. Abdullah, “Energy balancing through cluster head selection using Ktheorem in homogeneous wireless sensor networks,” in Proceedings of the 2nd Applications in Industry & Education (ICSTIE '08), Penang, Malaysia, 2008. View at: Google Scholar
 S. Jabbar, A. A. Minhas, T. Rashid, and S. Rho, Heuristic Approach for Stagnation Free Energy Aware Routing in Wireless Sensor Network, Ad Hoc& Sensor Wireless Networks, Old City Publishing, 2014.
 M. Qayyum, S. Jabbar, F. Azam, and S. Rizwan, “Stagnant free ant based energy aware heuristic routing in wireless sensor network,” in Proceedings of the International Conference on Collaboration Technologies and Systems (CTS '12), pp. 164–169, Denver, Colo, USA, May 2012. View at: Publisher Site  Google Scholar
 A. M. Zungeru, K. P. Seng, L.M. Ang, and W. Chong Chia, “Energy efficiency performance improvements for antbased routing algorithm in wireless sensor networks,” Journal of Sensors, vol. 2013, Article ID 759654, 17 pages, 2013. View at: Publisher Site  Google Scholar
 J. Bai, G.K. Yang, Y.W. Chen, L.S. Hu, and C.C. Pan, “A model induced maxmin ant colony optimization for asymmetric traveling salesman problem,” Applied Soft Computing Journal, vol. 13, no. 3, pp. 1365–1375, 2013. View at: Publisher Site  Google Scholar
 T. N. Bui, T. V. Nguyen, C. M. Patel, and K. T. Phan, “An antbased algorithm for coloring graphs,” Discrete Applied Mathematics, vol. 156, no. 2, pp. 190–200, 2008. View at: Publisher Site  Google Scholar  MathSciNet
 Y. Kao, M.H. Chen, and Y.T. Huang, “A hybrid algorithm based on ACO and PSO for capacitated vehicle routing problems,” Mathematical Problems in Engineering, vol. 2012, Article ID 726564, 17 pages, 2012. View at: Publisher Site  Google Scholar  MathSciNet
 F. Semchedine, L. BoualloucheMedjkoune, L. Bennacer, N. Aber, and D. Aïssani, “Routing protocol based on tabu search for wireless sensor networks,” Wireless Personal Communications, vol. 67, no. 2, pp. 105–112, 2012. View at: Publisher Site  Google Scholar
 M. Azami, M. Ranjbar, A. S. Rostami, and A. J. Amiri, “Increasing the network life time by simulated annealing algorithm in WSN with point coverage,” International Journal of Ad hoc, Sensor & Ubiquitous Computing, vol. 4, no. 2, pp. 31–45, 2013. View at: Publisher Site  Google Scholar
 Y. Xu and R. Qu, “An iterative local search approach based on fitness landscapes analysis for the delayconstrained multicast routing problem,” Computer Communications, vol. 35, no. 3, pp. 352–365, 2012. View at: Publisher Site  Google Scholar
 X.B. Hu and M. S. Leeson, “Evolutionary computation with spatial receding horizon control to minimize network coding resources,” The Scientific World Journal, vol. 2014, Article ID 268152, 23 pages, 2014. View at: Publisher Site  Google Scholar
 E. Amiri, H. Keshavarz, M. Alizadeh, M. Zamani, and T. Khodadadi, “Energy efficient routing in wireless sensor networks based on fuzzy ant colony optimization,” International Journal of Distributed Sensor Networks, vol. 2014, Article ID 768936, 17 pages, 2014. View at: Publisher Site  Google Scholar
 S. Mehfuz and M. N. Doja, “Swarm intelligent poweraware detection of unauthorized and compromised nodes in MANETs,” Journal of Artificial Evolution and Applications, vol. 2008, Article ID 236803, 16 pages, 2008. View at: Publisher Site  Google Scholar
 J. GarciaMacias and J. Gomez, “MANET versus WSN,” in Sensor Networks and Configuration, N. P. Mahalik, Ed., pp. 369–388, Springer, Berlin, Germany, 2007. View at: Publisher Site  Google Scholar
 S. Jabbar, R. Iram, A. A. Minhas, I. Shafi, S. Khalid, and M. Ahmad, “Intelligent optimization of wireless sensor networks through bioinspired computing: survey and future directions,” International Journal of Distributed Sensor Networks, vol. 2013, Article ID 421084, 13 pages, 2013. View at: Publisher Site  Google Scholar
 Y. Kambayashi, “A review of routing protocols based on antlike mobile agents,” Algorithms, vol. 6, no. 3, pp. 442–456, 2013. View at: Publisher Site  Google Scholar
 B. Talwar and A. K. Gupta, “Ant colony based mobile ad hoc networks routing protocols: a review,” International Journal of Computer Applications, vol. 49, no. 21, pp. 36–42, 2012. View at: Publisher Site  Google Scholar
 M. Günes, U. Sorges, and I. Bouazizi, “ARA—the antcolony based routing algorithm for MANETs,” in Proceedings of the International Conference on Parallel Processing Workshops, pp. 79–85, British Columbia, Canada, August 2002. View at: Publisher Site  Google Scholar
 F. Ducatelle, G. A. Di Caro, and L. M. Gambardella, “Using ant agents to combine reactive and proactive strategies for routing in mobile Ad Hoc networks,” Tech. Rep. IDSIA28042004, 2004. View at: Google Scholar
 F. Correia and T. Vazão, “Simple ant routing algorithm strategies for a (Multipurpose) MANET model,” Ad Hoc Networks, vol. 8, no. 8, pp. 810–823, 2010. View at: Publisher Site  Google Scholar
 H. Xie, Z.G. Zhang, and X.G. Zhou, “A novel routing protocol in wireless sensor networks based on ant colony optimization,” in Proceedings of the International Conference on Environmental Science and Information Application Technology (ESIAT '09), vol. 2, pp. 646–649, Wuhan, China, July 2009. View at: Publisher Site  Google Scholar
 J. Yang, M. Xu, W. Zhao, and B. Xu, “A multipath routing protocol based on clustering and ant colony optimization for wireless sensor networks,” Sensors, vol. 10, no. 5, pp. 4521–4540, 2010. View at: Publisher Site  Google Scholar
 M. Imran, M. A. Alnuem, M. S. Fayed, and A. Alamri, “Localized algorithm for segregation of critical/noncritical nodes in mobile ad hoc and sensor networks,” Procedia Computer Science, vol. 19, pp. 1167–1172, 2013. View at: Publisher Site  Google Scholar
 A. Paul, “Graph based M2M optimization in an IoT environment,” in Proceedings of the Research in Adaptive and Convergent Systems (RACS '13), pp. 45–46, Montreal, Canada, October 2013. View at: Publisher Site  Google Scholar
 A. Paul, Y.C. Jiang, J.F. Wang, and J.F. Yang, “Parallel reconfigurable computingbased mapping algorithm for motion estimation in advanced video coding,” ACM Transaction on Embedded Computing Systems, vol. 11, no. 2, article 40, 2012. View at: Publisher Site  Google Scholar
 A. Ahmad, S. Jabbar, A. Paul, and S. Rho, “Mobility aware energy efficient congestion control in mobile wireless sensor network,” International Journal of Distributed Sensor Networks, vol. 2014, Article ID 530416, 13 pages, 2014. View at: Publisher Site  Google Scholar
 A. Paul, “Realtime power management for embedded m2m using intelligent methods,” ACM Transaction on Embedded Computing Systems, vol. 13, no. 5s, article 148, 2014. View at: Publisher Site  Google Scholar
 A. Paul, B.W. Chen, K. Bharanitharan, and J.F. Wang, “Video search and indexing with reinforcement agent for interactive multimedia services,” ACM Transactions on Embedded Computing Systems, vol. 12, no. 2, article 25, 2013. View at: Publisher Site  Google Scholar
Copyright
Copyright © 2015 Sohail Jabbar et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.