A SWARM INSPIRED MULTIPATH DATA TRANSMISSION WITH CONGESTION CONTROL IN MANETS USING PROBABILISTIC APPROACH

Transcription

1 A SWARM INSPIRED MULTIPATH DATA TRANSMISSION WITH CONGESTION CONTROL IN MANETS USING PROBABILISTIC APPROACH Subhankar oarar 1, Vanana Bhattacheree 2 an Debasis Giri 1 1 Department of Computer Science an Engineering, Halia Institute of Technology,Halia, West Bengal, Inia subhankarranchi@yahoo.co.in, ebasis_giri@hotmail.com 2 Department of Computer Science an Engineering, Birla Institute of Technology,Mesra, Jharkhan, Inia vbhattacharya@bitmesra.ac.in ABSTRACT The maor causes of network congestion are the lack of network resources an the irrational allocation of network resources. The traitional congestion control methos such as rate control, winow mechanism, queue control an others can solve the congestion problem, but for the congestion ue to irrational allocation funamental solution is to make more effectively use of the network resources by austing the traffic routing epening on choice may be probabilistic, when congestion occurs. In mobile a hoc networks congestion creates elay in transmission an also loss of the packet that causes wastage of time an energy on recovery. The Wireless Networks have to play an important role to aopt an execute a large no of innovative application. New challenges have come consiering the maor limitations of the a hoc network like noe s limite processing power, balance the loa of network (to maintain the computation of the noe). To overcome the above problem some algorithm is invoke an there may be huge amount of packet loss an this leas to ecrease the lifetime of the network. Base on the concept of evolutionary cooperation in swarm Ant, we use the ants swarm intelligence to reinforce goo quality routes. The obective of our propose algorithm is to ientify the congestion areas between source an its neighboring noes to the estination an thus it will help to avoi the congestion of the network in the intermeiate links an also minimize the packet loss in the network. Using a new mathematical moel consiering the swarm-base ant intelligence concept, we foun an efficient congestion control mechanism (Ant s probabilistic transition rule). KEYWORDS A-Hoc Networks, Conitional Probability, Congestion, Distance Vector, Swarm Intelligence, Transmission Queue Length. 1. INTRODUCTION Mobile A hoc Networks are movable an also able to communicate within themselves using a wireless physical meium where there is no nee of pre existing network infrastructure. These networks can form stan-alone groups of wireless terminals, which can be again a part of any other cellular system or a fixe network. The main characteristic which raws the attention is that they are able to configure themselves without a centralize aministration. This paper is base on A Swarm inspire Probabilistic Path Selection with Congestion Control in MANETs by S.oarar, V.Bhattacheree, D.Giri which appeare in the proceeings of secon international conference on Computer Science, Engineering An Application, Springer-Verlag, Vol.2, pp , May 25-27, 2012 DOI : /iwmn

2 The Mobile a hoc network has several rawbacks that are not foun in fixe networks. Unlike in the traitional wireless networks, communication in such a ecentralize network is typically multi-hop, with the noes using each other as relay routers without any fixe infrastructure. This kin of network is very flexible an suitable for applications such as temporary information sharing in a conference, isaster rescue etc. However, multi-hop routing, ranom movement of mobile noes an other features in MANETs lea to The frequent change in network topology ue to the mobility of the noes causes a great eal of control information onto the network. The small capacity of batteries an the banwith limitation of wireless channels are other factors. The scalability is also a maor factor because the network performance egraes quickly as the no of noes increases Routing in Mobile A-hoc Wireless Networks Specially configure routing protocols are engage in orer to establish routes between noes which are more than a single hop. The ability to trace routes in spite of a ynamic topology is the unique feature of these protocols. These protocols can be categorize into two main types: Reactive (On-eman) an Proactive (Table-riven). Evaluating the routes continuously within the network is one by proactive protocols, so when a packet nees to be forware the route is alreay known an can be immeiately use. Reactive protocols appeal to a route etermination proceure on eman only Congestion in Mobile A-hoc Wireless Networks The network where the resources are share, to minimize the network overloa the ata rate must be auste use by the sener. The packets are roppe ue to some routers are not able to forwar the packets, so ue to the bottleneck conition leas to many packet rops. May be the roppe packets alreay travelle a long way in the network an thus consume significant amount of resources. So congestion leas to ecrease the network throughput. This type of situation occurre on the early Internet, the evelopment of the TCP congestion control mechanism. In mobile a hoc networks congestion occurs ue to the share wireless channel an ynamic topology, packet transmissions suffer from interference an faing. The network loa is burene through the transmission errors. There is an increasing eman of multimeia communications in MANETs so large amount of real-time traffic involves high banwith an it is liable to congestion. Congestion leas to packet losses an banwith egraation an also wastes time an energy on congestion recovery. Mobile A Hoc Networks have some unique properties that affect the esign of appropriate protocols an congestion control mechanism in particular. As it turne out, the vast ifferent environment in a mobile a-hoc network is highly problematic for TCP. The properties of MANETs are the noe mobility an a share, wireless multi-hop channel. Aapting new routes ue to noe mobility as well as unreliable meium result in unsteay packet elay an packet loses. Also these are not congestion loses. The wireless multi-hop channel allows only one ata transmission at a time within the transmission range of one noe. So close links as per the istance are not inepenent from each other. Internet routers typically are eicate hosts connecte by high banwith links. When congestion occurs on the Internet, it is usually concentrate on one single router, but in MANETs congestion affects a whole area because of the share meium. Not only noes are overloae, but regions of the networks are also overloae. 2. RELATED WORK The continuous research on congestion minimization for MANET still nees more new techniques. This section will illustrate the research relate to congestion control. 110

3 A istribute multi-path DSR protocol (MPDSR) was evelope to enhance the quality of service [1]. This protocol forwars the outgoing packets using multiple no of paths an with point to point reliability. Split Multi-path Routing protocol [2] uses a per-packet allocation technique to istribute the ata packets into multiple paths which prevents noes from being congeste in heavily loae traffic conitions. A novel technique Congestion Aaptive Routing in Mobile A Hoc Networks (CRP) [3] was propose. In CRP each noe on a route in the network sens a warn message to its previous noe when it is to be congeste. Then the previous noe uses a new route to bypass the congestion an get the first non congeste noe on the route. Traffic will be istribute probabilistically over these routes an there will be less chance of congestion occurrence. A ynamic loa aware base loa-balance routing for a hoc networks (DLBL) algorithm was evelope. The DLBL [4] uses intermeiate noe routing loa metric that helps the protocol to iscover a route with less network congestion. When a link fails because of the noe movement the algorithm uses the path maintenance to re oin the broken links to get a route from the source to the estination. Another novel algorithm propose is a Congestion-aware routing metric (CARM) [5] which consiers Mac-overhea, ata-rate an buffer queue elay to select low congestion with high throughput routes. Another mechanism that CARM applie was the use of link ata- rate categories to prevent the mismatche link ata-rate. CARM performe locally. An ant base DSR [6] will inform the source noe about QoS available to estination noe such as acceptable elay, itter an energy in the case of multimeia an real time application. They propose DSR using ACO an calle AntDSR(ADSR). Another novel algorithm propose a QoS-aware routing protocol [7] which consier two scheme amission control an feeback for the Qos requirements. Their maor works on the approximation of banwith estimation to calculate the rest amount of banwith available at each noe. A biologically inspire routing algorithm [8] for multi-hop networks consiers the concept of stigmergy an evaluates the optimal an sub-optimal routes without the whole system-wie connectivity information. They also incorporate MAC layer information into the routing ecision can prevent the forwar ata traffic into the congeste areas uner situation like various noal mobility, network ensity an ata loas. The success of ata transmission in Multipath routing [9] consiers the concept of traffic istribution strategy that efines how concurrently available paths are utilize an ata to the same estination as split an istribute over multiple paths. This paper shows when the no of path is single the entire ata packet is transmitte through it an maximum probability of success as the no of paths increases the probability of success of ata transmission will ecrease. 3. PROPOSED WORK In this section we will iscuss Swarm intelligence, Distance metric an the transmission Queue length Swarm Behaviour This algorithm is inspire by the foraging behaviour of biological ant colony [10], when they fin paths to foo sources. Each ant eposits a chemical calle pheromone when they move from source to the estination an the foragers follow such pheromone trails. By that more ants are attracte by these pheromone trails an in turn reinforce them even more. This natural phenomenon (stigmergy) plays a role in eveloping an manipulating local information Congestion Metric The maor parameters to control congestion areas probabilistically are the two metrics in a noe. 111

4 3.2.1 Distance Metric The first metric is the istance between the single hop by using the Time of arrival [11] estimating the time of signal travelling between noes (TOA) provies the istance between two noes. We can use ultrasoun sensor for measuring range between noes. For this purpose very accurate rubiium-base oscillators must be use. In orer to calculate the TOA parameter the noes must have a common clock, or exchange timing information by certain protocols such as two-way ranging protocol. The istance metric i where is the neighbor noe of i. i = C*( remote local )/2, C 3*10 8 m/s (1) The time between starting the transmission of a ata packet an receiving the corresponing immeiate acknowlege will refer as remote elay( remote ) an receiving one ata packet an sening out the immeiate acknowlegement will be the uration as local elay( local ). C is measure as the spee of light Transmission Queue Length Metric The secon metric is a heuristic value transmission queue length in the i th noe towars the neighbouring noes. The transmission queue length [8] can affect the packet latency an packet rop because of the size of the packet length. Figure 1 : Mobile noes with istance an out going transmission queue T q = 1 N i b q (2) 112

5 where q is the outgoing queue length in terms of bits waiting to be sent to the link [ i ] between i an an N is all the noes in the network an N is the neighbouring noes of b the noe i. We have consiere some ranom values as transmission queue in terms of bits. Using (2), we fin the gooness of heuristic values means small queue ata length will have higher heuristic value. From the figure 1, all the noes have symmetric link between them. We consier some hypothetical value as istance an transmission queue length, using (1), we foun the congestion metric T in the table 1 [8]. Table 1: Gooness of congestion metric T i Ant base probabilistic path selection When a source noe s, nees to route a estination without knowing the global topology, using the initial pheromone information available the ant will chose its next hop with probability P. We have propose an algorithm to fin the istance of neighbour noe from the i th noe an also we calculate the pheromone eposition, pheromone evaporation an resiual pheromone on the link l 1 τ = mvl, N [ i] i b i, (3) The istance from the noe i to its neighbouring noes can be calculate by using the equation 1, we assume some ranom istance metric for our algorithm. mvl is the minimum istance selecte from the neighbour noes of i. The τ i is the reciprocal of the istance on the link. We foun that the less istance have more reciprocal value of rate of pheromone eposition an vice versa. τ i = τi + τ i ρ (4) 113

6 e τ = τ (1 ρ) i i (5) τ i = 0.1 is the initial pheromone eposition an the ρ [12] specifies the rate at which pheromone evaporates means ant forget previous ecision for value ρ = 1 the pheromone evaporates rapily an search is ranom, when ρ = 0 results in slower evaporation rates, ρ [ 0,1]. We take ρ = 0.6. The final pheromone eposition is τ i. The final e pheromone evaporates from the link is τ. The resiual pheromone on the link is. i τ r = τ τ e i i i (6) Using (3) : τ = 3*1/5 =.6, τ = 3*1/4 =.75, 1,2 1,3 τ 1,4 = 3*1/3 =1, τ 1,6 = 3*1/6 =.5 Using (4) : ( ) τ 1,2 t =.1+(.6*.6)=.46, ( ) τ 1,3 t =.1+(.75*.6)=.55, τ ( ) 1,4 t =.1+(1*.6)=.7, τ ( ) 1,6 t =.1+(.5*.6)=.4 e Using (5) : ( ) e τ 1,2 t =.46*(1-.6)=.18, ( ) e τ 1,3 t =.55*(1-.6)=.22, τ ( ) 1,4 t =.7*(1-.6)=.28, e τ ( ) 1,6 t =.4*(1-.6)=.16 Using (6) : l ( t ) = =.28, l ( t ) = =.33, l ( t ) =.7-.28=.42, 1,2 1,3 1,4 l ( ) 1,6 t =.4-.16=.24. The link l from noe i shows when the istance is less compare to other neighbor noes, more pheromone eposition occurs. The noes resiual pheromone eposition metric in the table 2. i Table 2: Resiual pheromone metric τ i 114

7 When the pheromone information are available the ant will chose its next hop with probability P. We aopt this concept by using Baye s probability theorem α β P τ i T i =, α β τ T for i = 1 to N, (7) i i [ i] Nb where α 0 is the relative importance of the pheromone trail. β 0 is the relative importance of the congestion. α an β are the two tuneable parameter will control relative weight of pheromone trail an heuristic value T i. In the following, we calculate the probability corresponing to the noe i using. P 1,2 = (.28 3 *.85 1 ) / (.28 3 *.85 1 ) + (.33 3 *.78 1 ) + (.42 3 *.71 1 ) + (.24 3 *.64 1 ) =.17 P 1,3 = (.33 3 *.78 1 ) / (.28 3 *.85 1 ) + (.33 3 *.78 1 ) + (.42 3 *.71 1 ) + (.24 3 *.64 1 ) =.25 P 1,4 = (.42 3 *.71 1 ) / (.28 3 *.85 1 ) + (.33 3 *.78 1 ) + (.42 3 *.71 1 ) + (.24 3 *.64 1 ) =.48 P 1,6 = (.24 3 *.64 1 ) / (.28 3 *.85 1 ) + (.33 3 *.78 1 ) + (.42 3 *.71 1 ) + (.24 3 *.64 1 ) =.08 Analogously, we can calculate the probability corresponing to other noes which are shown in table 3. Table 3: Probability metric on l P, ( t ) i α > β Pi, ( t ) α < β P, i α > β Pi, α < β 1, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

8 3.4. Algorithm P 2 st : Noe i Processing for istribution of traffic Algorithm P 2 st( ) Input: i : Distance between the noes. q i : Transmission queue length. α : Relative importance of pheromone eposition. β : Relative importance of congestion. ρ : Rate at which pheromone evaporates. Output: P ( τ ) : Gooness of Congestion i P ( T ) : Resiual pheromone BEGIN 1. Upate parameter for Probabilistic path selection 2. τ Pheromone resiual on the link. i 3. Ti, Probability of congestion it iscars. 4. if (α>β) then 5. while( τ > τ ( ) ) i t i Nb 6. Distribute.Traffic( Max.Prob ( τ ) ) 7. wen 8. if(max( τ )= ( ) ) then i τ t i Nb 9. Distribute.Traffic( Max.Prob ( T ) ) 10. en if 11. else 12. if (α<β) then 13. while( T > Ti, N ) then b 14. Distribute.Traffic( Max.Prob ( T ) ) 15. wen 16. if (max( T ) = Ti, N ) then b 17. Distribute.Traffic( Max.Prob ( τ ) ) 18. en if 19. en if 20. en if 21. END 3.5. The Route Discovery i When initialization process starts the single-hop HELLO messages creates the neighbourhoo for each noe. The HELLO packet contains source IP aress an hop count (Initialize to 0). Whenever a traffic source s nees a route to a estination, it broacasts route request packets (RQ) across the network. The RQ packet contains estination IP aress, source IP aress, an broacast ID. For the route iscovery the solution woul be through flooing the RQ packets, we aopt the probabilistic broacast scheme in combine with the outgoing queue length in terms of bits an the istance metric. When a noe first receives a packet, Q D i, i 116

9 with probability p it broacasts the packet to its neighbours an with probability 1 p it iscars the packet. The probability value p is calculate as Q i, = q i, N i b q i, (8) D i, = i, N i b i, (9) P CONG = e σ ( Qi, ) (10) P Dist = e σ ( Di, ) (11) P = β ( P CONG ) + (1-β)( P Dist ), (12) where β lies between 0 an 1. Table 4: Probability metric of istance an congestion P CONG P Dist P CONG P Dist 1, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

10 Table 4: Probability metric on P P P P P P P β =.3 β =.5 β=.7 β =.3 β =.5 β=.7 1, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , Algorithm P 2 scm : Noe i Processing for selection of path Algorithm P 2 scm( ) Input: i : Distance between the noes. q i : Transmission queue length. Output: P CONG : Gooness of Congestion P Dist : Gooness of istance P : Probability of path selection BEGIN 1. Upate parameter for Probabilistic path selection 2. Q i Relative Pheromone eposition on the link., 3. D Relative congestion on the link i, 4. if ( pi ( t ) > pi N b ) then 5. Max.Prob ( pi ( t ) ) 6. en if 7. END where σ (σ > 0) is the coefficient. where [ 0,1] β is the relative importance of the istance an congestion in the link. When β is equal to.5 the probability of gooness of congestion an istance metric can be calculate. This broacast scheme helps to iscover new routes avoiing congestion areas an aopting less istance probabilistically. If the probability value P CONG is less means more congestion in the path, ue to maximum probability is set on P the other 118

11 parameter like P Dist must be high means shorter istance is available in the link. When the link have the same probability, this broacast scheme by changing the tuneable parameter β as.7 sets a new probability value to fin a new path. During the course of flooing the intermeiate noe receives a Route Request (RQ) packet. By capturing the source aress an the previous hop noe in the message cache it first set the reverse path to the source. If a vali route to the estination is available, that is, there is at least one link associate with the pheromone trail greater than the other neighbour noe or least istance within the neighbour noe, the intermeiate noe generates a route reply (RP). The RP is route back to the source s via the reverse paths. Otherwise, the RQ is rebroacast. When the estination noe receives RQ it will sen a RP to all the neighbours from which it gets a RQ. The intermeiate noe can maintain multiple loop-free paths by recoring all the new paths that have the latest sequence number but a lower hop-count in its routing table, an also sen a RP to all the neighbours from which it gets a RQ. During the process of the RP back to the source s will assigns an initial pheromone value τ to the corresponing neighbour noe, which inicates a vali route to the estination Description of the Algorithms Our propose p 2 scm algorithm presents the pseuo coe of how a noe i process the out going queue an by using swarm intelligence how it will select the path an base on the routing table, the p 2 st will istribute the ata traffic accoring to the probabilities for each neighbour noe in the routing table[12]. The noe i first calculate the istance i between all its neighbour noes. By using the smallest istance mvl the noe i evaluate the relative istance between all its neighbour noes. P 2 st is also evaluate the τ i which is reciprocal to the istance i where the relative pheromone eposition occurs on the link l. Now p 2 st evaluate the resiual pheromone on the link l from the line 2 of the above algorithm. Seconly p 2 st evaluate the probability of gooness of q congestion ( 1 ). q [ i ] Nb In our propose algorithm from line no 4-10 when α>β it first checks the pheromone ( τ ) eposition in a particular link have the highest value within its neighbour or not. If foun yes then the maximum probability P will be on that link l. Next it checks a pair of paths with highest eposition an same value, if foun yes within that, less congeste (T ) path will have the highest probability[13]. Similarly from line no when the α<β, if Ti, have the highest value within its neighbour then the highest probability will be on that link l. Next it checks a pair of paths with highest an same value of Ti, or not if foun yes within that, more pheromone ( τ i ) eposite path will have the highest probability on that link l. 4. ANALYSIS P 2 st at first from noe i an the two tuneable parameter α>β we observe the higher pheromone on the link l i will have higher probability of path selection an higher gooness of congestion means small queue ata length will have low probability, if the eposition is same on both the link, we observe then p 2 st by using α>β an conitional probability evaluates the less congeste path means queue ata length is small. In our example noe 3 to its neighbour noe 1 an 2 from table 1, table 2 an table 3. Also by making α<β p 2 st consier the less congeste path probabilistically means when the queue ata length is same on both the link, the higher probability comes with higher pheromone i 119

12 concentration on the link an if the congestion is same within a pair of paths then it evaluates the paths which have more pheromone eposition. By using two tuneable parameters p 2 st ecies the path probabilistically which have less congeste or more pheromone eposition when the noes have ynamic topology at real time situation. Similarly P 2 scm evaluates the gooness of congestion as well as istance. If pi ( t ) > pi N b means the average value of pi ( t ) is maximum within all the links of the neighbouring noes. Another important feature of P 2 scm is that if the probabilistic value of gooness of congestion is low, the probabilistic value of gooness of istance must be high. As we have seen from the table CONCLUSION In this paper, we propose a new mathematical moel for congestion control. We have evelope a single hop congestion aware probabilistic path selection algorithm which employs the swarm intelligence of biological ant. Using istance an congestion metric the algorithm invokes route iscovery process with less congestion an shorter istance in a single hop at run time. By probabilistic approach the algorithm can prevent forwar ata traffic into the congeste areas an also fins the shortest istance probabilistically within a pair of noes. For ata transmission, if more congestion occurs in all the links of a noe, the algorithms by using the tuneable parameter that can shift the moe of transmission an can easily avoi the congeste areas. By using the above state mathematical moels an some hypothetical ata we trie to control the congestion at local noe level. REFERENCES [1] R.Leung, Jilei Liu, E.Poon, C.A.Chan & Baochun L (2001) MPDSR : a QoS-aware multipath ynamic source routing protocol for wireless a-hoc networks, In proceeings Local Computer Networks, 26 th Annual IEEE Conference, pp [2] S.Lee & Mario Garla, (2001) Split Multi-path Routing with Maximally Disoint Paths in A Hoc Networks, In proceeings International IEEE Conference on Communication, pp [3] Duc. A.Tran & Harish Raghavenra, (2006) Congestion Aaptive Routing in Mobile A Hoc Networks, IEEE transaction on parallel an Distribute systems, 17: pp [4] Xiangquan Zheng, Wei Guo, R.Liu & Yongchun Trian, (2004) A New Dynamic Loa-aware Base Loa-balance Routing for A Hoc Networks, IEEE, pp [5] Xiaoqin Chen, H.M.Jones & A.D.S.Jatyalath, (2007) Congestion- Aware Routing Protocol for Mobile A Hoc Networks, In proceeings of IEEE conference on vehicular Technology, Doi /VETECF , pp [6] R.Asokan, A.M.Nataraan & C.Venkatesh, (2008) Ant Base Dynamic Source Routing Protocol to Support Multiple Quality of Service(QoS) Metrics in Mobile A Hoc Networks, International Journal of Computer Science an Security, vol.2 no.3, pp [7] L. Chen & Weni B. Heinzelman, (2005) Qos-Aware Routing Base on Banwith Estimation for Mobile A Hoc Networks, IEEE on Selecte Areas in Communication, vol.23, pp [8] L. Zhenyu, M.Z.Kwiatkowska & Costas Constantinou, (2005) A Biologically Inspire Congestion Control Routing algorithm For MANETs, In Proceeings of the Thir IEEE International Conference on Pervasive Computing an Communications, pp

13 [9] A. K. Monal & P. K. Baneree, (2008) The Success of ata Transmission in Multipath Routing for MANET, Mobile an Pervasive Computing, CoMPC-2008, pp [10] R. Beckers, J.L.Deneubourg & S.Goss, (1992) Trails an u turns in the selection of the shortest path by the ant lasius niger, Journal of Theoretical Biology, pp [11] A. Gunther & Christian hoene, (2005) Measuring Roun trip times to etermine the istance between WLAN noes, In proceeings of Networking 2005, Springer-Verlag, pp [12] A.C.S.Jawahar, Ant Colony Optimization for Mobile A-Hoc Networks, aaychak@een.rutgers.eu. [13] S. Soarar, V. Bhattacheree & D. Gir (2012) A Swarm inspire Probabilistic Path Selection with Congestion Control in MANETs, In Proceeings of secon international conference on Computer Science, Engineering an Application, Springer-Verlag, Vol.2, pp Authors Subhankar Joarar is currently working as Assistant Professor at Halia Institute of Technology, Halia. He complete his MCA an M. Tech from Birla Institute of Technology, Mesra. Currently he is pursuing PhD From BIT, Mesra. His current research areas are Mobile a Hoc Networks, Swarm Intelligence an Avance Algorithm esign. DR. Vanana Bhattacheree is a member of the IEEE an the IEEE Computer Society. She is currently working as Associate Professor at Birla Institute of Technology, Ranchi. She complete her B. E. in Computer Science an Engineering from BIT Mesra in 1989, an M. Tech an PhD from Jawaharlal Nehru University, New Delhi in 1991 an 1995 respectively. Her current research areas are Software Process Moels, Software Cost Estimation, Data Mining an Mobile a Hoc Networks. Dr. Debasis Giri is presently Professor in the Department of Computer Science an Engineering, Halia Institute of Technology, Halia , Inia. He receive his Ph.D egree from Inian Institute of Technology, Kharagpur , Inia 1n He i his masters (M. Tech an M. Sc) both from IIT, Kharagpur in 2001 an 1998 respectively. He has tenth All Inia Rank with percentile score in the Grauate Aptitue Test in Engineering (GATE) Examination in He receive certificate from All Inia Science Teachers' Association for success in Science Aptitue & Talent Search Test in He has publishe more than 20 technical papers in the referre ournals/conferences. He is presently Eitorial Boar Members of many repute Journals, namely Journal of Convergence, International Journal of Computers an Applications, Journal of Security an Communication Networks (Wiley InterScience) etc. Further, he is a Reviewer of many repute International Journals. He is also Program Committee member of many International Conferences. He is a Life member of Cryptology Research Society of Inia. His current research interests inclue cryptography, Network security, Information security, E- commerce security, Security in Wireless Sensor Networks, Security in VANETs an Wireless Networks. 121