Ab (W im con nec col un ada dra me ove An opt the alg con bas inf the ove rea mi rou pro app pac obs loc opt 802 Qo are fun gat sin req sho dur Engineerin www.etasr Even A E Si s bstract—The m WSN) in disast prove quality nsumption. Lo cessary. With llected data is i ique form of o aptive routing awbacks of AC essages and l ercome by con n event-based timization (EB e performance gorithm is to i nsumption by sed on the b formation of no e sink (base sta erheads. EBC active routing nimizes energy uting algorithm oposed algo proximately 7 cket delivery r served in com calization bas timization (AL 2.11 standard i Keywords-AC oS; RSSI; WSN WSN consis e small in siz nctional node thering purpos nk node using quirement in ould perform ration. A trad ng, Technology r.com t Base Ant Co Enhan Shankar D Departmen inhgad Colleg Pune, Maha sdchavan27@ main challeng er situations i y of service ocation awaren hout knowing insignificant. A optimization m g and guarante CO are data f ong converge nsidering locat clustering loc BC_LEE-ACO) e of WSN. T improve QoS a cluster format biological insp odes. In cluster ation) through _LEE-ACO is algorithm wh y consummati ms like AODV rithm reduc %, in addition ratio and incre mparison with ed eligible e LEEP with AC in proposed wo CO; ALEEP; A N I. IN sts of a numb e and have lo s are deploye ses. The senso g multiple ho disaster situa the commun ditional routing y & Applied Sci Chava ed Clu olony cemen D. Chavan nt of E&TC ge of Engineeri arashtra, India @rediffmail.com ge of a wirele is to discover (QoS) and ness of nodes is the position Ant colony opti method, which i eed packet de flooding, huge nce time. Th tion informati alized energy ) algorithm is p The main focu and minimize tion and selecti pired routing- ring, data is ag h cluster head s a scalable a hich improves on of WSN as V, ACO, ACO ces energy n to improvem ease in packet d other algorith energetic Path CO) of the net ork increased p AODV; CH; clu NTRODUCTION er of sensor n ow power cap ed in a partic or nodes (SN) ops as shown ations is that nication witho g algorithm do ience Research an & Kulkarni: usterin Optim nt of W ing m ess sensor ne efficient routi to reduce e s also useful o n of sensor imization (ACO is highly suitab livery. The pr overhead of c hese drawback ion of sensor efficient ant c proposed to en us of the pro the network e ing the optima -ACO and lo ggregated and s (CH) which re and energy ef s QoS, lifetim s compared to O using RSSI consumption ment in throu drop which ha hms, i.e. autono h_with_Ant C twork. Use of packet drop. ustering; EBC; nodes. These pacity. These cular area for pass the data n in Figure 1 t the sensor out failure for oes not consid h V Event Based C ng Loc mizatio Wirele etwork ing, to energy r even nodes, O) is a ble for rimary control ks are nodes. colony nhance oposed energy al path ocation sent to educes fficient me and o other I. The n by ghput, as been omous Colony IEEE ; LEE; nodes multi- r data to the 1. The nodes r long der the loca whe algo solv QoS effi is p enh clus biol nod A. hyb con nod ene netw reac pro Vol. 8, No. 4, 20 Clustering Loca calized on for ess Sen Dr. ation informat ere stable com orithm is one ved [1-3]. The S and network Therefore, an icient ant colo proposed. The hance QoS, to ster formation logical inspire des. Classical Rou Classical rou brid [4, 5]. R nsumption beh des in the WS ergy efficiency work perform ctive protocol tocols due to 018, 3177-3183 alized Energy Ef d Ener Perfo nsor N Anju V D. Y. Patil In Pune, Mah anju_k64@ tion and thus mmunication is of the major p e main goals k lifetime and t Fig. 1. W n event base ony optimizati e main focus o minimize n n and to selec ed routing-AC II. REL uting Protocol uting protoco Routing proto havior of nod SN are supplie y is a major pr mance. Takin s are more ef o less control 3 Efficient Ant Co rgy Ef ormanc Networ V. Kulkarni nstitute of Tech harashtra, Indi @yahoo.co.in cannot be use s important. S problems of se of WSN routi to reduce conn WSN architecture ed clustering ion (EBC_LE of the propos network energy ct the optimal CO and locati LATED WORK ls ls are proact ocol has an des in a netw ed with a limit roblem that inf ng energy in fficient as com overheads. B 3177 olony Optimizat fficient ce rk hnology a n ed in disaster Selection of ro ensor network ing are to imp nectivity failur localized en E-ACO) algor sed algorithm y consumptio path based o ion informatio tive, reactive impact on en work. Since m ted energy ba fluences the ov nto considera mpared to proa Both proactive tion… t areas outing to be prove re. nergy rithm is to on by on the on of e and nergy mobile attery, verall ation, active e and Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3177-3183 3178 www.etasr.com Chavan & Kulkarni: Event Based Clustering Localized Energy Efficient Ant Colony Optimization… reactive protocols are unaware of energy metrics and hence cause lowering of the battery energy of the nodes over the most heavily used routes in the network. B. Localization Based Routing Protocols Location information for sensor networks by most routing protocols needs to calculate the distance between two particular nodes so that energy consumption can be calculated. According to the dependency of range measurements, the existing localization schemes can be categorized into two major categories: range-based approaches and range-free approaches. Range-based and range-free schemes are again divided into anchor based and anchor free schemes. The anchor-free schemes do not assume that the node positions are known at first. On the other hand, the anchor-based schemes need some nodes that are aware of their positions (anchor nodes) to provide geographic information to unidentified nodes to localize [6]. C. Location Based Bio-Inspired Routing Algorithms Various location-based routing algorithms have been proposed, nevertheless, they have a relative shortcoming: either not guaranteeing to find a way to the destination or locating a path which is much longer than the shortest path. Position based ant colony (POSANT) routing algorithm is a collection of ant colony based routing algorithms that use the data about the position of nodes to improve the efficiency of an ant algorithm. Contrary to other position based algorithms, this algorithm does not fail when the network contains nodes with different transmission range. POSANT is a multipath routing algorithm using GPS to find position information which adds to cost of nodes and is not suited for indoor network [7]. Location based ant colony optimization (LOBANT) algorithm uses distance to consider routing metrics due to received signal strength indicator (RSSI), but energy aware metrics are not taken into account [8]. Autonomous localization based eligible energetic Path_with_Ant colony optimization (ALEEP with ACO) algorithm was developed in [9] by a combination of the advantages of the best exiting protocols. Authors used the location of the nodes, adaptive transmission power (ATP) and energy aware metrics to increase the efficiency of routing. After studying the related work, we came to conclusion and proposed an event based clustering localized energy efficient ant colony optimization (EBC_LEE-ACO) routing algorithm by combining the advantages of ACO, RSSI and clustering. III. PROPOSED SCHEME A. Problem Definition The main challenge of WSN is to discover efficient routing, as the sensor nodes are not static and change their position randomly. Limited battery life is another issue. A disaster situation is one more challenge for WSN by which the communications in the network may fail and lead to excessive packet drop and can hang the network. For solving these problems, EBC_LEE-ACO algorithm is proposed, which is ACO, based on geographical location with clustering approach. B. Objectives  To find and reconstruct the optimal path for routing in disaster situations smoothly and quickly.  To reconstruct communication links in case of link failure.  To reduce network energy consumption by selecting the least distance from source to destination node with localized and clustering approaches.  To improve network QoS.  To verify whether the proposed routing algorithm is more efficient than other routing algorithms like AODV, ACO and ACO using RSSI and present ALEEP with ACO routing algorithm. C. Methodology To achieve the objectives we considered the following implementation steps:  Design EBC_LEE-ACO routing algorithm by combining the advantages of ACO, RSSI and clustering.  Simulate network considering variable number of nodes and variable node mobility.  This simulation will provide us data to perform network analysis of performance parameters like throughput, packet delivery ratio, packet drop and consumed energy.  Compare above network parameters between the proposed algorithm and AODV, ACO, ACO using RSSI algorithm and present ALEEP with ACO algorithm.  Network Simulator 2 (NS2) is used for simulation. IV. CURRENT METHODS A. Ad-hoc on Demand Distance Vector (AODV) Protocol As the name itself suggests, AODV protocol is an on demand routing protocol, which means that it determines a route to a destination only when a node wants to send a packet to that destination [10]. Essential objectives of the algorithm are:  To broadcast discovery packets only when necessary using RREQ message.  To perform local connectivity management, neighborhood identification and general topology maintenance using HELLO messages.  To spread data about changes in availability to neighboring nodes which will probably require the information using RERR message. AODV operation is divided into two phases, route discovery and route maintenance. B. Ant Colony Optimization (ACO) Algorithm ACO is a bio-inspired meta-heuristic algorithm introduced in [11, 12]. The main idea is to use ants as an inspirational source because they follow self-organizing principles which allow highly coordinated behavior. Ants have collective learning intelligence. Each ant communicates, learns and cooperates non-verbally with the others through pheromones. Different kinds of ant’s algorithm can be inspired by different ant behaviors, e.g. foraging, labor division, brood sorting, and Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3177-3183 3179 www.etasr.com Chavan & Kulkarni: Event Based Clustering Localized Energy Efficient Ant Colony Optimization… cooperative transport [13, 14]. The basic rule of ACO is the ability of ants to discover the shortest path between food sources and the anthill. In the beginning, the route the ants find may not be the shortest path, but with the passage of the time, more and more ants move cooperatively and the trail of their path becomes shorter and shorter until they get the shortest path. There are three phases of the ant based algorithm namely route discovery, route maintenance and route failure handling. C. Ant Colony Optimization using Received Signal Strength Indicator (ACO using RSS) Algorithm Location aware ACO routing is a high performance routing protocol for WSN design [15]. The main reason to seek location awareness in ACO routing is the dynamic network topology causing frequent link breakup that causes the source node to spend most of its time in route setup and route maintenance. In location awareness in ACO, each node will have a general idea about the network topology and its neighbors so that it can choose the nearest neighbor toward the destination [15, 16]. This new routing algorithm is based on ACO and uses location as a parameter to enhance its efficiency. From the RSSI value every node can determine the distance between nodes. In ACO using RSSI, the route is searched only when there is a collection of information packets to be sent from a source node (S), to a destination node (D), thus it is a reactive routing algorithm. Sending the information packets will begin after a route from the source to the destination node is built up. Before that, only forward ants are being exchanged with backward ants. To limit the time it needs to discover a route while keeping the quantity of generated ants as small as it could be expected under the circumstances, data about the position of nodes is utilized as a heuristic value [7, 8]. When there is a packet to be sent, the source starts a route discovery phase. At first, a route request (RREQ) broadcasts to each one of the nodes from S to D. When the D gets the first RREQ message, its answer is a route reply message to S. On receiving the RREP message, a node will extract RSSI value from it and would calculate the location of the neighboring nodes and in turn the location of the destination. The routing table is updated with distance information between the nodes utilizing RSSI value. Every sensor node in the WSN has a memory block in which the leftover energy, the location data of the node, its neighbors and the base station are stored. Route establishment using distance is described in [8, 20]. V. PROPOSED METHODS A. EBC_LEE-ACO Algorithm Previous algorithms have some limitations such as: For proper monitoring, dense and large WSN will be used for different types of applications. There is a high probability of redundant data being recorded by neighboring nodes during an event. As many nodes might sense the same event, they will establish a route separately. Routing algorithms based on ant colony are considered to have a high percentage in terms of packet delivery, but the drawback is overhead of the control messages required for discovering the route. Cluster-heads use only forward route discovery control messages. Their limitation is the dynamic topology of the network, which limits the bandwidth availability and energy constraints. To overcome these problems, clustering technique is used, as the clustering approach has an advantage of spatial reuse of resources to increase system capacity, data aggregation, reduce energy consumption etc. [17-19]. In the present research work an event based clustering localized energy efficient ant colony optimization (EBC_LEE-ACO) algorithm is proposed. The main focus of the proposed EBC_LEE-ACO algorithm is to minimize the energy consumption of the network by cluster formation and to select the optimal path based on the biological inspired routing ACO and location information of nodes. B. Three Phases (Stapes) of Proposed Algorithm 1) Hop Tree Formation Phase In WSN, data transmission takes place in multi-hop fashion where each node forwards its data to a neighbor node nearer to the sink. The node doesn’t have a full understanding of the network, but has only knowledge of its neighboring nodes. Each node only knows the hop level in which it is in a hop tree [20]. In this phase, the distance metric used is the hop count (i.e. the number of nodes from A to B). The distances between the sink and different nodes are calculated. The algorithm is initiated by the sink node broadcasting a hop configuration message (HCM) to its neighboring nodes with a hop value. This hop value gets incremented every time the message is transmitted and is stored in their routing table. This process is continued until all the nodes are configured with a hop value within a tree. The HCM has two parts ID and HopToTree (HTT). ID is the node identifier whereas the HTT is the distance in hops. In this approach the sink node broadcasts HCMs having values of HTT as 1 and hop count 0. The receiving nodes forward messages to their neighboring nodes. Initially all nodes set the value of HTT as infinity. On receiving the HCM, each node compares the value of HTT in the HCM with the value of HTT that node already has in it. If the above conditions are met, then the node updates its internal stored values by the value of the field ID as well as the value of the HTT variable of the HCM. The node broadcasts the HCM with the new values. If the condition is not met which means that the node received the HCM from a shorter distance then the node discards it. The step described above occurs repeatedly until the whole network is configured. Initially there is no recognized route and the value of HTT variable has the smallest distance to the sink. When the first event is triggered the variable still stores the smallest distance, but a new route is established. After the event, the variable stores the lower among the 2 values: the distance to the sink or the distance to the closest already recognized route. Hop tree formation phase is shown in Figure 2. 2) Cluster Formation and Cluster Head (CH) Selection Phase When an event is detected by one or more nodes, the cluster formation and the cluster head selection algorithm gets started. A set of nodes that have detected an event forms a cluster. Once clusters are formed, the next process is to select the cluster head within each cluster. The main process for forming the cluster is the selection of leader node which is called cluster head (CH) by using the cluster configuration message (CCM). CCM has four attributes (Type, ID, HTT, State), where ID is the sto Fig netw the no sel bet cou i.e on tie wi me col the wh sam CH the nu red agg Clu AC 3) rou coo wo des bei sea to loc rou no inv the me des no are me Engineerin www.etasr e identifier of ores state value g. 2. Flooding work where node If this is the e cluster head de or to the no lection is based tween the cur unt that the no ., the one with e can use the . After the end ll be selected embers which llects all infor em to the sink hole set of info me event is ac H. In addition, erefore large mber of con duced, the n gregation and uster formatio CO algorithm i Route Discov The final st ute discovery ordinator) to D ould be done stination. Befo ing exchanged arched when t a destination. cation informa ute. When the de, the source volves a broad eir neighboring essage each no stination node de initiates a R e at 1 hop dista essage reaches ng, Technology r.com the node that e of the node. g of HCM messag es are deployed ra e first event, a d is the one w ode which is c d on paramete rrent energy v ode has been se h smaller ID w energy level d of the proce d as the CH h also are a p rmation gathe k. The main be ormation gathe ccumulated or , collected info amount of re ntrol message number of o d reliable tran on and CH se is shown in Fi very and Selec tep involves t and optimal p D (sink). Actu once the rou fore that, only d for the rout the packets are . The main pa ation of the no source node w e node initiates dcast message g nodes which ode checks wh e in its routin RREQ messag ance. The proc s destination no y & Applied Sci Chava started the C ge from the sink andomly all sensing nod which is closes closest to an es ers such as hop versus the ini elected as CH wins over the o to select the C ess, one of the H. The other part of the sa ered by memb enefit of this a ered by variou r aggregated a formation is hi edundant data es for buildin overlapping r nsmission of election phase igure 3. ction Phase the route esta path selection ual sending an ute is establish y forward and te discovery p e required to b arameter for th ode which is u wants to send d s a route disco e to be sent fro h are at 1 hop hether it conta ng table. If th ge to its neigh cess will conti ode D. ience Research an & Kulkarni: CM. HTT and k to the last node des are eligibl st to either th stablished rout p count, energy tial energy an . In the case of others. Alternat CH in the cas e nodes in the nodes will b ame event. Th ber nodes and algorithm is th us nodes sensin at a single nod ighly correlate a gets cleaned ng routing tr route for eff data is maxim e of the EBC_ ablishment th n for S (in ou nd receiving o hed from sou d backward an process. A ro be sent from s his algorithm used to discov data to a destin overy process om source no p distance. Wit ains the entry f here is no entr hboring nodes inue until the R h V Event Based C d state e of the le and e sink te. CH y ratio nd the f a tie, tively, se of a group be the he CH sends hat the ng the de, the ed and d, the ree is fficient mized. _LEE- hrough ur case of data urce to nts are oute is source is the ver the nation which des to th that for the ry the which RREQ F nod S b Rou nod extr nod this (FA divi info ben this info pro bro form esta the algo low of t with 1. 2. 3. 4. 5. 6. 7. Vol. 8, No. 4, 20 Clustering Loca Fig. 3. Nodes If the entry o des a reply me by following b ute reply pack de receives th racted which de and the rou s algorithm div ANTs) and ba ide ant agents ormation coll nefiting from t s principle, F ormation, but o cess is done adcasting a FA m the route fo ablishment by table whose orithm focuse w as possible a the link betwee h each other a The S (coo launching FA FANTs disco nearer to S b repeated until The FANT cr node. Upon receivi remaining ene When the des it initiates a During this t received FAN transition rule Once D is rea takes the stack The BANT fo path from the 018, 3177-3183 alized Energy Ef s detected the sam of the D is fo essage (RREP) back the path ket (RREP) co he RREP m will help in uting table will vides ant agent ackward ants s in two secti lected by th the information FANTs do only collect th by BANTs. T ANT agent to r the event dis y sending FAN first two neig s on keeping and the pherom en the nodes. as follows: ordinator) ini ANTs to destin over the route y analyzing th l an ant reache reates a stack, ng a FANT e ergy with the t stination node timer called F time the D n NTs, D will c e. ached, FANT k and follows ollows the sta e sink to the so 3 Efficient Ant Co me event form a cl ound in any o ) is generated h that was fol ontains RSSI essage, RSSI determining t l be updated. ts into two sec (BANTs). Th ions is to take he other sec n collected by not create o he information The process s wards the sink ssemination. T NTs to two p ghbor nodes ar the number o mone trail show These agents a tiates route nation at regula e to D based he routing tab es D. , pushing in tr each of the n threshold ener or sink receiv FET (forward node accepts calculate the o is converted i it. ack entries and ource. 3180 olony Optimizat luster and select C of the neighb by D and is se llowed to reac value. When I value woul the location o For route sele ctions: forward he main reaso e the advantag ction i.e. BA y FANTs. Base or update ro n and route cre starts with the k or base stati The S initiates ossible paths re closer to S f generated an ws the edge w are communic establishmen ar time interva on which no ble. This proce rip times for e nodes compare rgy. ves the first FA d expiration ti all FANTs. optimal path u into BANT. B d traces the re tion… CH boring ent to ch D. each ld be of the ection d ants on to ge of ANTs ed on outing eation e CH ion to route from . The nts as weight cating nt by als. ode is ess is every es its ANT, mer). With using BANT verse Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3177-3183 3181 www.etasr.com Chavan & Kulkarni: Event Based Clustering Localized Energy Efficient Ant Colony Optimization… 8. Pheromone edge updates depend upon the residual energy of the node and location. 9. Once the optimal path is found, D backs up all possible paths in case the path is failed. VI. PRACTICAL ANALYSIS This section presents a practical analysis of network performance metrics like throughput, packet delivery ratio, packet drop, consumed energy and node mobility. A. Network Scenarios and Simulation Parameters Scenarios and simulation parameters are shown in Table I. TABLE I. SCENARIOS AND SIMULATION PARAMETERS Network Parameters Values Routing Protocol / Algorithm AODV, ACO, ACO using RSSI, EBC_LEE-ACO Traffic Patterns CBR (Constant Bit Rate) Network Size 1000 × 1000 (X x Y) MAC Protocol 802.11 Initial Energy 200J (for each node) Simulation Time 30s Simulation Platform NS-allinone-2.32 Node Variables Number of Nodes 10/30/60/100 Node Speed 3m/s Variable Mobility Number of Nodes 50 Maximum Speed 1/2/3/4/5 m/s B. Results and Analysis 1) Results on Varying Number of Nodes It is observed from Figures 4 to 7 that for increasing number of nodes and constant mobility, throughput and packet delivery ratio decrease and consumed energy and the number of dropped packets increase because:  the probability of success in accessing the channel decreases,  as hop count increases, congestion and delay increases and collision and transmission error increase. Fig. 4. Throughput versus number of nodes Fig. 5. PDR versus number of nodes Fig. 6. Packets drop versus number of nodes Fig. 7. Energy versus number of nodes 2) Results of Varying Node Mobility Figures 8 to 11 show that for increasing mobility with constant nodes, throughput, packet delivery ratio and consumed energy decrease and packet drop increases because the probability of path breakage increases and the construction of a new path takes time. Results show that the performance parameters of the network are improved by the use of the proposed EBC_LEE-ACO algorithm in comparison with AODV, ACO and ACO using RSSI algorithms due to the following characteristics of the EBC_LEE-ACO algorithm: Nodes 10 20 30 40 50 60 70 80 90 100 50 55 60 65 70 75 80 85 90 95 Packet Delivery Ratio AODV ACO ACO RSSI EBC LEE ACO P ac ke ts D ro p( N o. ) Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3177-3183 3182 www.etasr.com Chavan & Kulkarni: Event Based Clustering Localized Energy Efficient Ant Colony Optimization…  No back propagation.  Multipath routing.  No packet flooding.  Shortest distance routing.  Ideal nodes.  Reduce overhead.  Data aggregation.  No redundant data transmission. Fig. 8. Throughput versus node mobility Fig. 9. Energy versus number of nodes VII. CONCLUSIONS In this work, the EBC_LEE-ACO routing algorithm is implemented and the network performance, by varying number of nodes and node mobility, is analyzed. This algorithm is extensively compared to other algorithms like AODV, ACO and ACO-RSSI by considering network metric parameters like throughput, packet delivery ratio, packet drop and consumed energy. Simulation results show that the EBC_LEE-ACO algorithm outperformed the other algorithms in disaster situations. ACO achieves better performance compared to AODV, as ACO allows rerouting to another link in the case of existing link failure (no back propagation). ACO using RSSI routing algorithm improves the routing by minimizing the flooding of routing packets because, it has the location information of nearby nodes. Fig. 10. PDR versus node mobility Fig. 11. Packet drop versus node mobility Fig. 12. Energy versus node mobility The EBC_LEE-ACO algorithm has achieved better performance due to clustering technique and location information of nodes. Clustering data is aggregated and sent to the sink through CH which reduces overheads. Also, location Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3177-3183 3183 www.etasr.com Chavan & Kulkarni: Event Based Clustering Localized Energy Efficient Ant Colony Optimization… information of nodes is useful to send data to shortest distance node in less time. The proposed algorithm reduces energy consumption by approximately 7%. An improvement in throughput, packet delivery ratio and increase in packet drop has been observed in comparison with present network routing algorithms, i.e. autonomous localization based eligible energetic Path_with_Ant colony optimization (ALEEP with ACO) [9]. Use of IEEE 802.11 standard increased packet drop. Hence, our EBC_LEE-ACO algorithm is useful for improvement in QoS and reduction in energy consumption of the WSN. It is most suitable for information monitoring in disaster situations. The extension of the proposed algorithm will be considered for varying network areas as well as increasing number of nodes and mobility in future work. REFERENCES [1] I. F. Akyildiz, W. Su, Y. Sankarasubramaniam, E. Cayirci, “Wireless sensor networks: a survey”, Computer Networks, Vol. 38, No. 4, pp. 393–422, 2002 [2] J. Yick, B. Mukherjee, D. Ghosal, “Wireless sensor network survey”, Computer Networks, Vol. 52, No. 12, pp. 2292–2330, 2008 [3] S. K. Gupta, P. Sinha, “Overview of Wireless Sensor Network: A Survey”, International Journal of Advanced Research in Computer and Communication Engineering, Vol. 3, No. 1, pp. 5201-5207, 2014 [4] A. K. Gupta, H. Sadawarti, A. K. Verma, “Review of Various Routing Protocols for MANETs”, International Journal of Information and Electronics Engineering, Vol. 1, No. 3, pp. 251-259, 2011 [5] H. S. A. Hamatta, N. I. Zanoon, R. M. Al-Tarawneh, “Comparative Review for Routing Protocols in Mobile Ad-Hoc Networks”, International Journal of Ad hoc, Sensor & Ubiquitous Computing, Vol. 7, No. 2, pp. 13-31, 2016 [6] A. Mesmoudi, M. Feham, N. Labraoui, “Wireless Sensor Networks Localization Algorithms: A Comprehensive Survey”, International Journal of Computer Networks & Communications, Vol. 5, No. 6, pp. 45-64, 2013 [7] S. Kamali, J. Opatrny, “A Position Based Ant Colony Routing Algorithm for Mobile Ad-hoc Networks”, Third International Conference on Wireless and Mobile Communications (ICWMC'07), Guadeloupe, France, March 4-9, 2007 [8] R. Vallikannu, S. E. Jubin, “A Location Based ACO Routing Algorithm For Mobile Ad Hoc Networks Using RSSI”, IEEE International Conference on Communication and Signal Processing, Chennai, India, April, 3-5, 2013 [9] R. Vallikannu, A. George, S. K. Srivatsa, “Autonomous localization based energy saving mechanism in indoor MANETs using ACO”, Journal of Discrete Algorithms, Vol. 33, pp. 19–30, 2015 [10] P. K. Maurya, G. Sharma, V. Sahu, A. Roberts, M. Srivastava, “An Overview of AODV Routing Protocol”, International Journal of Modern Engineering Research, Vol. 2, No. 3, pp. 728-732, 2012 [11] M. Dorigo, C. Blum, “Ant colony optimization theory: A survey”, Theoretical Computer Science, Vol. 344, No. 2-3, pp. 243–278, 2005 [12] C. Blum, “Ant colony optimization: Introduction and recent trends”, Physics of Life Reviews, Vol. 2, No. 4, pp. 353–373, 2005 [13] S. Binitha, S. S. Sathya, “A Survey of Bio inspired Optimization Algorithms”, International Journal of Soft Computing and Engineering, Vol. 2, No. 2, pp. 137-151, 2012 [14] O. Deepa, A. Senthilkumar, “Swarm Intelligence from Natural to Artificial Systems: Ant Colony Optimization”, International Journal on Applications of Graph Theory in Wireless Ad hoc Networks and Sensor Networks, Vol. 8, No.1, pp. 9-17, 2016 [15] X. Wang, Q. Li, N. Xiong, Y. Pan, “Ant Colony Optimization-Based Location-Aware Routing for Wireless Sensor Networks”, in: Lecture Notes in Computer Science, Vol. 5258, pp. 109-120, Springer, 2008 [16] C. Dominguez-Medina, N. Cruz-Cortes, “Energy-Efficient and Location- Aware Ant Colony Based Routing Algorithms for Wireless Sensor Networks”, 13th Annual Conference on Genetic and Evolutionary Computation, Dublin, Ireland, pp. 117-124, July 12-16, 2011 [17] S. K. Popat, M. Emmanuel., “Review and Comparative Study of Clustering Techniques”, International Journal of Computer Science and Information Technologies, Vol. 5, No. 1, pp. 805-812, 2014 [18] S. Mahajan, P. K. Dhiman, “Clustering in Wireless Sensor Networks: A Review”, International Journal of Advanced Research in Computer Science, Vol. 7, No. 3, pp. 198-201, 2016 [19] S. K. Gupta, N. Jain, P. Sinha, “Clustering Protocols in Wireless Sensor Networks: A Survey”, International Journal of Applied Information Systems, Vol. 5, No. 2, pp. 41-50, 2013 [20] L. Aparecido Villas, A. Boukerche, H. Soares Ramos, H. A. B. Fernandes de Oliveira, R. Borges de Araujo, A. A. Ferreira Loureiro, “DRINA: A Lightweight and Reliable Routing Approach for in-Network Aggregation in Wireless Sensor Networks”, IEEE Transactions on Computers, Vol. 62, No. 4, pp. 676-689, 2013