Copy­right Notice & Dis­claimer

© Atul Patil, 2015. All rights reserved. This arti­cle, titled “A Study of Rout­ing Pro­to­cols in Mobile Ad-hoc Net­work for Secu­ri­ty Con­sid­er­a­tion”, was authored and pub­lished by Atul Patil. It was orig­i­nal­ly fea­tured in the Inter­na­tion­al Jour­nal of Advanced Research in Com­put­er Sci­ence and Soft­ware Engi­neer­ing (IJARCSSE), ISSN: 2277–128X, Vol­ume 5, Issue 3 (March 2015). The orig­i­nal pub­li­ca­tion can be accessed at https://www.researchgate.net/publication/306358387_A_Study_of_Routing_Protocols_in_Mobile_Ad-hoc_Network_for_Security_Consideration and www.ijarcsse.com.

Dis­claimer: This arti­cle is repub­lished here by the orig­i­nal author, Atul Patil, in accor­dance with the copy­right poli­cies of the Inter­na­tion­al Jour­nal of Advanced Research in Com­put­er Sci­ence and Soft­ware Engi­neer­ing (IJARCSSE). The con­tent remains unchanged to pre­serve its orig­i­nal­i­ty. For any inquiries or copy­right-relat­ed mat­ters, please con­tact the author direct­ly.

Abstract-An ad-hoc net­work is a mul­ti-hop wire­less net­work where all nodes coop­er­a­tive­ly main­tain net­work con­nec­tiv­i­ty­with­out a cen­tral­ized infra­struc­ture. If these node­schange their posi­tions dynam­i­cal­ly, it is called a mobile ad-hoc net­work (MANET). Since the net­work topol­o­gy­changes fre­quent­ly, effi­cient adap­tive rout­ing pro­to­colssuch as AODV, DSR, and TORA are used. As the net­work is wire­less, secu­ri­ty becomes the major issue in Mobile Ad hoc Net­works. Some of the attacks such as mod­i­fi­ca­tion, fab­ri­ca­tion, imper­son­ation and denial of ser­vice attacks are due to mis­be­hav­ior of mali­cious nodes, which dis­rupts the trans­mis­sion. To avoid such attacks some of cryp­to­graph­ic algo­rithms and key man­age­ment schemes and some exist­ing secu­ri­ty pro­to­cols are used. In this paper we rep­re­sent a sur­vey of per­for­mance based secure rout­ing tech­niques in MANET. The secu­ri­ty tech­niques are cat­e­go­rized based upon dif­fer­ent approach­es. The secu­ri­ty type is bor­rowed from intru­sion detec­tion as either mis­use detec­tion or anom­aly detec­tion. This paper pro­vides the major improve­ment in the secure tech­niques in MANET research using these approach­es the fea­tures and cat­e­gories in the sur­veyed work.

Key­words: MANET, Rout­ing Pro­to­col, Secu­ri­ty.

I.          INTRODUCTION

All com­mu­ni­ca­tions in MANETs take place over the wire­less medi­um. The wire­less chan­nels are open, shared and with rel­a­tive­ly less pow­er. First, due to the “open” nature of wire­less medi­um, the wire­less com­mu­ni­ca­tion in MANETs is sus­cep­ti­ble to eaves­drop­ping that may lead to crit­i­cal infor­ma­tion leak­age. The require­ment of promis­cu­ous mode raised by many MANET pro­to­cols, i.e. con­tin­u­ous mon­i­tor­ing of the shared medi­um, fur­ther facil­i­tates the prac­ti­cal­i­ty of eaves­drop­ping. Addi­tion­al­ly, wire­less trans­mis­sions can be inter­cept­ed. Once cap­tur­ing ongo­ing trans­mis­sion, adver­saries with suf­fi­cient knowl­edge of MANET pro­to­cols can mean­ing­ful­ly per­form var­i­ous mali­cious behav­iours.

Some typ­i­cal exam­ples are: alter key infor­ma­tion in pack­ets, dis­card and/or forge mes­sages, inject mali­cious mes­sages, gen­er­ate floods of spu­ri­ous mes­sages, and replay con­trol and data traf­fic. Such mis­be­hav­iours have severe impact on MANETs. For exam­ple, MANET rout­ing process requires all nodes duti­ful­ly par­tic­i­pate in for­ward­ing pack­ets and pro­vide valid rout­ing infor­ma­tion. Adver­saries who per­form either of above mali­cious behav­iours can ruin the rout­ing func­tion­al­i­ty [1] and [3].

By sup­port­ive infra­struc­ture, we mean enti­ties (or author­i­ties) that per­form admin­is­tra­tive and man­age­ment func­tion­al­i­ties in MANETs. In a pure at MANET, there is no par­tic­u­lar node that is des­ig­nat­ed as a cen­tral author­i­ty to exe­cute admin­is­tra­tive and man­age­ment func­tion­al­i­ties. Instead, all net­work oper­a­tions, includ­ing secu­ri­ty relat­ed con­trol, are on the self-con­fig­u­ra­tion base and in a decen­tral­ized way. Whether the secu­ri­ty con­trol (e.g., authen­ti­ca­tion and autho­riza­tion) or not can be achieved heav­i­ly relies on the coop­er­a­tion of net­work nodes. How­ev­er, in the ful­ly dis­trib­uted and open envi­ron­ment of ad hoc net­work­ing, nodes trust­wor­thy are fair­ly dif­fi­cult to iden­ti­fy. This pro­vides pos­si­ble oppor­tu­ni­ties for mis­be­hav­ing nodes to harm the secu­ri­ty con­trol oper­a­tion. Mean­while, the absence of admin­is­tra­tive or domain bound­aries make the enforce­ment of any secu­ri­ty mea­sures an even more com­plex prob­lem. In this paper we dis­cussed a sur­vey of per­for­mance based secure rout­ing pro­to­col tech­niques in MANET.

II.  BACKGROUND TECHNIQUES

MANET Security

The the­o­ry and expe­ri­ences have indi­cat­ed that, due to its unique char­ac­ter­is­tics, MANETs are suf­fer­ing from a wide range of secu­ri­ty threats and attacks, not only the same attacks their infra­struc­ture coun­ter­part are fac­ing, but also new ones par­tic­u­lar­ly tar­get­ing MANETs.

Unsecured Wireless Channel

All com­mu­ni­ca­tions in MANETs take place over the wire­less medi­um. The wire­less chan­nels are open, shared and with rel­a­tive­ly less pow­er. First, due to the “open” nature of wire­less medi­um, the wire­less com­mu­ni­ca­tion in MANETs is sus­cep­ti­ble to eaves­drop­ping that may lead to crit­i­cal infor­ma­tion leak­age. The require­ment of promis­cu­ous mode

raised by many MANET pro­to­cols, i.e. con­tin­u­ous mon­i­tor­ing of the shared medi­um, fur­ther facil­i­tates the prac­ti­cal­i­ty of eaves­drop­ping. Addi­tion­al­ly, wire­less trans­mis­sions can be inter­cept­ed. Once cap­tur­ing ongo­ing trans­mis­sion, adver­saries with suf­fi­cient knowl­edge of MANET pro­to­cols can mean­ing­ful­ly per­form var­i­ous mali­cious behav­iours. Some typ­i­cal exam­ples are: alter key infor­ma­tion in pack­ets, dis­card and/or forge mes­sages, inject mali­cious mes­sages, gen­er­ate floods of spu­ri­ous mes­sages, and replay con­trol and data traf­fic. Such mis­be­hav­iours have severe impact on MANETs. For exam­ple, MANET rout­ing process requires all nodes duti­ful­ly par­tic­i­pate in for­ward­ing pack­ets and pro­vide valid rout­ing infor­ma­tion. Adver­saries who per­form either of above mali­cious behav­iours can ruin the rout­ing func­tion­al­i­ty [1] and [2].

Dynamic Mobility

In MANETs, freely roam­ing nodes join and leave the net­work inde­pen­dent­ly, pos­si­bly fre­quent­ly, and with­out notice. This dynam­ic mobil­i­ty rais­es sev­er­al chal­lenges. First, the net­work topol­o­gy is con­stant­ly chang­ing. More impor­tant­ly, the mobil­i­ty makes it dif­fi­cult in most cas­es to have a clear pic­ture of the mem­ber­ship. Trust rela­tion­ship among mobile nodes can­not be assumed to be held in all time, which may lead secu­ri­ty solu­tions with sta­t­ic con­fig­u­ra­tion not to pro­duce expect­ed results. Sec­ond­ly, the net­work mobil­i­ty also makes it dif­fi­cult to clas­si­fy nodes as inter­nal nodes or exter­nal nodes, which can be eas­i­ly achieved in tra­di­tion­al infra­struc­ture net­works. The clas­si­fi­ca­tion of inter­nal and exter­nal nodes (that is, nodes that belong to the net­work or not) is impor­tant for estab­lish­ing a line of defence, such as authen­ti­ca­tion and autho­riza­tion. Assist­ed by the absence of trust rela­tion­ship and clas­si­fi­ca­tion facil­i­ties, adver­saries can eas­i­ly infil­trate MANETs and launch var­i­ous attacks from inside.

Absence of Central Supportive Infrastructure

By sup­port­ive infra­struc­ture, we mean enti­ties (or author­i­ties) that per­form admin­is­tra­tive and man­age­ment func­tion­al­i­ties in MANETs. In a pure at MANET, there is no par­tic­u­lar node that is des­ig­nat­ed as a cen­tral author­i­ty to exe­cute admin­is­tra­tive and man­age­ment func­tion­al­i­ties. Instead, all net­work oper­a­tions, includ­ing secu­ri­ty relat­ed con­trol, are on the self-con­fig­u­ra­tion base and in a decen­tral­ized way. Whether the secu­ri­ty con­trol (e.g., authen­ti­ca­tion and autho­riza­tion) or not can be achieved heav­i­ly relies on the coop­er­a­tion of net­work nodes. How­ev­er, in the ful­ly dis­trib­uted and open envi­ron­ment of ad hoc net­work­ing, nodes trust­wor­thy are fair­ly dif­fi­cult to iden­ti­fy. This pro­vides pos­si­ble oppor­tu­ni­ties for mis­be­hav­ing nodes to harm the secu­ri­ty con­trol oper­a­tion. Mean­while, the absence of admin­is­tra­tive or domain bound­aries make the enforce­ment of any secu­ri­ty mea­sures an even more com­plex prob­lem. For exam­ple, because mobile nodes move through dif­fer­ent net­work areas and become asso­ci­at­ed with dif­fer­ent do mains, it may be dif­fi­cult to estab­lish the trust asso­ci­a­tions of nodes. In turn, MANETs may lack the ground for the estab­lish­ment of some type of a secret (or keys), so that cryp­to­graph­ic mech­a­nisms can be employed [4] and [6].

Limited Resources

In order to be light and portable, mobile wire­less devices in MANETs can­not be equipped with many resources, such as mem­o­ry, bat­tery and CPU capac­i­ty. The lack of suf­fi­cient resources could result in sev­er­al secu­ri­ty risks. First of all, lim­it­ed com­pu­ta­tion­al capa­bil­i­ties of mobile nodes can­not sup­port com­pli­cat­ed cryp­to­graph­ic oper­a­tions, espe­cial­ly if they have to be per­formed for each pack­et and over each link of the tra­versed path. Sec­ond­ly, mobile nodes have less phys­i­cal pro­tec­tion, and there-fore are eas­i­ly stolen, cap­tured and com­pro­mised. In many cas­es, adver­saries exploit the com­pro­mised nodes to launch the attack. In addi­tion, node’s trans­mis­sion pow­er is typ­i­cal­ly lim­it­ed. An adver­sary with suf­fi­cient trans­mis­sion pow­er and knowl­edge of the phys­i­cal and medi­um access con­trol (MAC) lay­er mech­a­nisms can obstruct its neigh­bours from gain­ing access to the wire­less medi­um. More impor­tant­ly, mobile devices could become ide­al tar­gets of DoS attacks due to their lim­it­ed com­pu­ta­tion­al capa­bil­i­ty, mem­o­ry and bat­tery. An adver­sary could gen­er­ate bogus pack­ets, forc­ing the vic­tim to con­sume a sub­stan­tial por­tion of its resources. Even worse, a mali­cious node with valid cre­den­tials could fre­quent­ly gen­er­ate con­trol traf­fic, such as route queries, at a high rate not only to con­sume band­width, but also to impose cum­ber­some cryp­to­graph­ic oper­a­tions on a net­work node

[5] and [7].

Routing Attacks in MANET

Rout­ing pro­to­cols (Net­work lay­er pro­to­cols) extend con­nec­tiv­i­ty from neigh­bour­ing 1‑hops nodes to all oth­er nodes in MANET. The con­nec­tiv­i­ty between mobile hosts over a poten­tial­ly mul­ti-hop wire­less link strong­ly relies on coop­er­a­tive reac­tions among all net­work nodes. A vari­ety of attacks tar­get­ing the net­work lay­er have been iden­ti­fied and heav­i­ly stud­ied in research papers. By attack­ing the rout­ing pro­to­cols, attack­ers can absorb net­work traf­fic; inject them­selves into the path between the source and des­ti­na­tion, and thus con­trol the net­work traf­fic flow, as shown in Fig­ure 1 (a) and (b), where a mali­cious node M can inject itself into the rout­ing path between sender S and receiv­er D.

The traf­fic pack­ets could be for­ward­ed to a non- opti­mal path, which could intro­duce sig­nif­i­cant delay. In addi­tion, the pack­ets could be for­ward­ed to a non-exis­tent path and get lost. The attack­ers can cre­ate rout­ing loops, intro­duce severe net­work con­ges­tion, and chan­nel con­tention into cer­tain areas. Mul­ti­ple col­lud­ing attack­ers may even pre­vent a source node from find­ing any route to the des­ti­na­tion, caus­ing the net­work to par­ti­tion, which trig­gers exces­sive net­work con­trol traf­fic, and fur­ther inten­si­fies net­work con­ges­tion and per­for­mance degra­da­tion.

Fig 1 Illus­tra­tion of Rout­ing Attack

Attacks at the routing discovery phase

There are mali­cious rout­ing attacks that tar­get the rout­ing dis­cov­ery or main­te­nance phase by not fol­low­ing the spec­i­fi­ca­tions of the rout­ing pro­to­cols. Rout­ing mes­sage flood­ing attacks, such as hel­lo flood­ing, RREQ flood­ing, acknowl­edge­ment flood­ing, rout­ing table over­flow, rout­ing cache poi­son­ing, and rout­ing loop are sim­ple exam­ples of rout­ing attacks tar­get­ing the route dis­cov­ery phase. Proac­tive rout­ing algo­rithms, such as DSDV and OLSR, attempt to dis­cov­er rout­ing infor­ma­tion before it is need, while reac­tive algo­rithms, such as DSR and AODV, cre­ate routes only when they are need­ed. Thus, proac­tive algo­rithms are more vul­ner­a­ble to rout­ing table over­flow attacks. Some of these attacks are list­ed below.

Rout­ing table over­flow attack: A mali­cious node adver­tis­es routes that go to non-exis­tent nodes to the autho­rized nodes present in the net­work. It usu­al­ly hap­pens in proac­tive rout­ing algo­rithms, which update rout­ing infor­ma­tion peri­od­i­cal­ly. The attack­er tries to cre­ate enough routes to pre­vent new routes from being cre­at­ed. The proac­tive rout­ing algo­rithms are more vul­ner­a­ble to table over­flow attacks because proac­tive rout­ing algo­rithms attempt to dis­cov­er rout­ing infor­ma­tion before it is actu­al­ly need­ed. An attack­er can sim­ply send exces­sive route adver­tise­ments to over­flow the victim’s rout­ing table.

Rout­ing cache poi­son­ing attack: In route cache poi­son­ing attacks, attack­ers take advan­tage of the promis­cu­ous mode of rout­ing table updat­ing, where a node over­hear­ing any pack­et may add the rout­ing infor­ma­tion con­tained in that pack­et head­er to its own route cache, even if that node is not on the path. Sup­pose a mali­cious node M wants to poi­son routes to node X. M could broad­cast spoofed pack­ets with source route to X via M itself; thus, neigh­bour­ing nodes that over­hear the pack­et may add the route to their route caches.

Attacks at the routing maintenance phase

There are attacks that tar­get the route main­te­nance phase by broad­cast­ing false con­trol mes­sages, such as link- bro­ken error mes­sages, which cause the invo­ca­tion of the cost­ly route main­te­nance or repair­ing oper­a­tion. For exam­ple, AODV and DSR imple­ment path main­te­nance pro­ce­dures to recov­er bro­ken paths when nodes move. If the des­ti­na­tion node or an inter­me­di­ate node along an active path moves, the upstream node of the bro­ken link broad­casts a route error mes­sage to all active upstream neigh­bours. The node also inval­i­dates the route for this des­ti­na­tion in its rout­ing table. Attack­ers could take advan­tage of this mech­a­nism to launch attacks by send­ing false route error mes­sages.

Attacks at data forwarding phase

Some attacks also tar­get data pack­et for­ward­ing func­tion­al­i­ty in the net­work lay­er. In this sce­nario the mali­cious nodes par­tic­i­pate coop­er­a­tive­ly in the rout­ing pro­to­col rout­ing dis­cov­ery and main­te­nance phas­es, but in the data for­ward­ing phase they do not for­ward data pack­ets con­sis­tent­ly accord­ing to the rout­ing table. Mali­cious nodes sim­ply drop data pack­ets qui­et­ly, mod­i­fy data con­tent, replay, or flood data pack­ets; they can also delay for­ward­ing time- sen­si­tive data pack­ets selec­tive­ly or inject junk pack­ets.

Wormhole attack:

An attack­er records pack­ets at one loca­tion in the net­work and tun­nels them to anoth­er loca­tion. Rout­ing can be dis­rupt­ed when rout­ing con­trol mes­sages are tun­nelled. This tun­nel between two col­lud­ing attack­ers is referred as a worm­hole. Worm­hole attacks are severe threats to MANET rout­ing pro­to­cols. For exam­ple, when a worm­hole attack is used against an on-demand rout­ing pro­to­col such as DSR or AODV, the attack could pre­vent the dis­cov­ery of any routes oth­er than through the worm­hole.

Attacks on particular routing protocols

There are attacks that tar­get some par­tic­u­lar rout­ing pro­to­cols. In DSR, the attack­er may mod­i­fy the source route list­ed in the RREQ or RREP pack­ets. It can delete a node from the list, switch the order, or append a new node into the list. In AODV, the attack­er may adver­tise a route with a small­er dis­tance met­ric than the actu­al dis­tance, or adver­tise a rout­ing update with a large sequence num­ber and inval­i­date all rout­ing updates from oth­er nodes [1–3] and [6- 8].

III.  SURVEY OF SECURE ROUTING TECHNIQUES

Trust Based Secure Routing in AODV Routing Protocol

A per­fect trust mod­el in the net­work lay­er, and estab­lished secure route between source and des­ti­na­tion with­out any intrud­ers or mali­cious nodes. In this paper, exist­ing AODV rout­ing pro­to­col has been mod­i­fied in order to adapt the trust based com­mu­ni­ca­tion fea­ture. Pro­posed trust based rout­ing pro­to­col is equal­ly con­cen­trates both in node trust and route trust. In this pro­posed mod­el, route trust plays an equal role with node trust. Using trust val­ue, secure route can be estab­lished in the MANET. Here, net­work secu­ri­ty enhance­ment is com­plete­ly per­formed in the lime light of trust val­ue. In the dynam­ic envi­ron­ment, node can change its char­ac­ter­is­tics at any time. After suc­cess­ful par­tic­i­pa­tion in the route estab­lish­ment process, the neigh­bor may behave like as a mali­cious node. To avoid this, route trust process (one of the process in the mod­i­fied pro­to­col) con­tin­u­ous­ly mon­i­tor the active routes and cal­cu­late the cur­rent route trust val­ue or the sta­tus of the route. But most of the pre­vi­ous works have been con­cen­trat­ed only in the node trust for estab­lish­ing com­mu­ni­ca­tion. This paper explains three main oper­a­tions; Node trust cal­cu­la­tion, Route trust cal­cu­la­tion and Trust based route estab­lish­ment and route mon­i­tor­ing process.

This mod­el requires some ade­quate changes in the exist­ing source ini­ti­at­ed rout­ing pro­to­col, AODV. Mod­i­fied AODV rout­ing pro­to­col estab­lish­es route among nodes based on the trust val­ue.

Node Trust Cal­cu­la­tion Process:

Var­i­ous meth­ods have been pro­posed for cal­cu­lat­ing node’s trust­wor­thi­ness. Dif­fer­ent trust met­rics have been eval­u­at­ed to iden­ti­fy the trust lev­el of node. Each node has opin­ion about oth­er node’s (neigh­bor) trust­wor­thi­ness. Node X has an opin­ion about trust­wor­thi­ness of one of its neigh­bor node Y based on Y’s pre­vi­ous and cur­rent behav­iors. a new data struc­ture Neigh­bor is intro­duced in each node of the MANET. All the nodes in such envi­ron­ment already main­tain Rout­ing Table. Addi­tion­al­ly added Neigh­bor Table should be main­tain­ing in all the nodes for keep tracks the dynam­i­cal­ly chang­ing neigh­bor list and its cor­re­spond­ing node trust val­ue. Trust com­pu­ta­tion involves the process of assign­ing weights (utility/importance fac­tor) to the events that they can mon­i­tor and quan­ti­fied. Weight assign­ment process depends on the type of appli­ca­tion demand­ing the trust lev­el. Nodes are dynam­i­cal­ly assign­ing these weights based on their own cri­te­ria and cir­cum­stances. These weights have a con­tin­u­ous range from 0 to +1 rep­re­sent­ing the sig­nif­i­cance of a par­tic­u­lar event from unim­por­tant to most impor­tant. The trust val­ues for all the events of a node can be com­bined using indi­vid­ual weights to deter­mine the aggre­gate trust lev­el for anoth­er node.

Route Trust Cal­cu­la­tion Process:

Route trust is com­put­ed by every node for each route in its rout­ing table. Mod­i­fied extend­ed Rout­ing sup­ports Route Trust cal­cu­la­tion process. Exist­ing Rout­ing Table extend­ed with one more field; Route Trust. In this approach, source node selects the route which is hav­ing the high­est Route Trust val­ue. Route Trust field of every Rout­ing Table entry is updat­ed at some reg­u­lar inter­val. In this method, only one addi­tion­al field is enough to mon­i­tor the route trust wor­thi­ness. The pro­posed approach is the exten­sion of exist­ing AODV rout­ing pro­to­col for cre­at­ing secure route for com­mu­ni­ca­tion. Pro­posed mod­i­fi­ca­tions are in accept­able lim­it. With this min­i­mum over­head, we can eas­i­ly elim­i­nate the mali­cious node as well as they can estab­lish a best trust­ed route between source and des­ti­na­tion. Also it cre­ates a secure com­mu­ni­ca­tion in this envi­ron­ment with­out any inter­nal attack­ers. Using sim­u­la­tion results, the per­for­mance of this pro­to­col is not suf­fi­cient jus­ti­fied. In the future, it will be incor­po­rate with oth­er MANET rout­ing pro­to­cols [1].

DAAODV: A Secure Ad-hoc Routing Protocol based on Direct Anonymous Attestation

Wen­chao Huang, Yan Xiong, Depin Chen et. al. pro­posed a nov­el secure rout­ing pro­to­col DAAODV which is based on Ad-hoc On-demand Dis­tance Vec­tor rout­ing (AODV). DAAODV takes full advan­tage of trust­ed com­put­ing tech­nol­o­gy, par­tic­u­lar­ly the Direct Anony­mous Attes­ta­tion (DAA) and Prop­er­ty-based Attes­ta­tion (PBA) pro­to­cols. DAAODV is an anony­mous pro­to­col with­out require­ment of Trust­ed Third Par­ty (TTP). More­over, we pro­pose an effi­cient sign­ing and ver­i­fi­ca­tion scheme to over­come the poten­tial DoS attacks trig­gered by the low effi­cien­cy of DAA and PBA. In the sim­u­la­tion, the results show that DAAODV is still effi­cient in dis­cov­er­ing secure routes com­pared with AODV pro­to­col. In this paper, based on AODV and pro­posed a nov­el secure ad hoc rout­ing pro­to­col DAAODV which is anony­mous and avoids TTP, and pre­vents from mali­cious nodes and self­ish nodes. The basic idea is to use Direct Anony­mous Attes­ta­tion (DAA) to accom­plish full anonymi­ty in the rout­ing pro­to­col and use issuer instead of TTP, and to use prop­er­ty based attes­ta­tion (PBA) to guar­an­tee that only nodes whose plat­form is trust­ed can join the group. The main chal­lenge of imple­ment­ing this pro­to­col is the cost of DAA and PBA pro­to­col is a lit­tle high, so we choose an effi­cient DAA pro­to­col and pro­pose a new light- weight­ed sign­ing and ver­i­fy­ing pro­to­col to ease the prob­lem. Exper­i­ments proves that it is still very effi­cient com­pared with AODV pro­to­col.

DAAODV presents almost a ful­ly pro­tec­tion of rout­ing process and it can be more eas­i­ly analysed than oth­er pro­to­cols for the hosts that could par­tic­i­pate in the rout­ing pro­to­col have to run in an antic­i­pat­ed way. The main extra cost of DAAODV via AODV is the estab­lish­ment of secure link which uses DAA and PBA pro­to­cols. The DAA adopt­ed in this paper is very effi­cient in DAASign and DAAVer­i­fy though not effi­cient in join pro­to­cols. How­ev­er, hosts have already got the cer­tifi­cate in join pro­to­col before deployed, which means only the cost of DAA Sign and DAA Ver­i­fy are con­sid­ered in our pro­to­col. Mean­while, hosts with the cer­tifi­cate could make DAASign which means the bot­tle­neck TTP is no longer need­ed in the pro­to­col. Addi­tion­al­ly, hel­lo mes­sages should be broad­cast­ed after a few sec­onds for con­trol­ling the CHV, which increas­es the time inter­val of estab­lish­ment of secure link. How­ev­er, only the process­es of estab­lish­ment cost an extra time, and oth­er mes­sages are han­dled effi­cient­ly for they are encrypt­ed by sym­met­ric keys between hosts. They pre­sent­ed a secure ad hoc rout­ing pro­to­col which can pre­vent most attacks includ­ing worm-hole attacks, ver­tex cut attacks, and traf­fic analy­sis attacks, and adopt a new effi­cient sign­ing and ver­i­fy­ing scheme pre­vent­ing DoS attacks. This pro­to­col doesn’t use TTP, and doesn’t add much over­head in ns‑2 sim­u­la­tion. In future work is to make a fine-grained con­struc­tion of the rout­ing soft­ware, as the design of DAAODV on soft­ware lev­el is a lit­tle coarse-grained. For exam­ple, we should make a con­crete scheme of oper­at­ing the PCRs, and should prove that the DAAODV can avoid attacks at the soft­ware lev­el [2].

AODVsec: A Multipath Routing Protocol in Ad- Hoc Networks for Improving Security

Cuirong Wang, Shux­in Cai et. al. pro­posed a secure rout­ing pro­to­col based on mul­ti­path rout­ing tech­nol­o­gy, name­ly AOD­Vsec, which divides a data unit into sev­er­al data pieces and trans­mits these pieces through dif­fer­ent paths. By set­ting secu­ri­ty lev­el on each node, AOD­Vsec lim­its the max­i­mum num­ber of data pieces an inter­me­di­ate node can for­ward. In this way, the mali­cious node can­not get enough data infor­ma­tion for break­ing the encryp­tion algo­rithm. Sim­u­la­tion results show that AOD­Vsec improves secu­ri­ty with neg­li­gi­ble rout­ing over­head by com­par­i­son of the tra­di­tion­al mul­ti­path AODV rout­ing pro­to­cols.

Design and Imple­men­ta­tion of AOD­Vsec:

In AOD­Vsec, each node is set a trust lev­el to lim­it the max­i­mum data piece num­ber that can be trans­mit­ted through. Mul­ti­ple paths are gen­er­at­ed from the source to the des­ti­na­tion, and the path infor­ma­tion is stored in source node’s rout­ing table. Before send­ing each data unit at the source node, each data unit is split into sev­er­al pieces. AOD­Vsec assigns a data piece to a safer path select­ed from the local rout­ing table.

Reverse path: Dif­fer­ent from tra­di­tion­al AODV, AOD­Vsec does not look up from the broad­cast list when gen­er­at­ing the reverse path. The rout­ing table’s update time should fol­low the three fol­low­ing prin­ci­ples. To estab­lish con­nec­tion to the des­ti­na­tion node, the source node broad­casts a request PREQ. On the receipt of source’s PREQ at the first time, the inter­me­di­ate node inserts a reverse path to the local broad­cast list. When it receives sources PREQ from oth­er path, it stops look­ing up local broad­cast list and adds anoth­er reverse path. Only the fol­low­ing three con­di­tions sat­is­fy, update the rout­ing table

• If there is no route to the source in the rout­ing table, AOD­Vsec adds this new route to the rout­ing table.
• If the num­ber of the rout­ing paths to the source has not hit the max­i­mum num­ber, which should be set accord­ing to prac­ti­cal require­ments and node num­ber, AOD­Vsec adds this new route to the rout­ing table.
• If there is a route update request which trans­mits through less hops, even the rout­ing path num­ber hits the max­i­mum lim­it, AOD­Vsec updates rout­ing table by adding this new rout­ing path.

For­ward path: In AODV pro­to­col, before send­ing back the response pack­et RREP, the node looks up the reverse path exist­ed in the rout­ing table, through which RREP is sent back, and final­ly the for­ward path is gen­er­at­ed. While in AOD­Vsec, if we query rout­ing table to choose the reverse path based on AODV’s mech­a­nism, we will get the same

result on every attempt of query. In this way, they gen­er­ate a sin­gle for­ward path. When a node sends RREP, AOD­Vsec uses the new route query func­tion for choos­ing the par­tic­u­lar reverse path with least sent RREP mes­sage. Thus, all the reverse paths would be used as round robin, and mul­ti­ple for­ward paths could be estab­lished.

The infra­struc­ture less and dynam­ic nature of MANET demands new set of net­work­ing analy­sis in order to pro­vide diverse appli­ca­tion in many dif­fer­ent sce­nar­ios. So, it is pos­si­ble that some appli­ca­tion demands less over­head as well as fast pro­cess­ing with effi­cient trans­mis­sion. This paper, presents the pro­to­col being pro­posed which uti­lizes the dual coop­er­a­tive mobile agents and sta­tion­ary agents for rout­ing in dynam­ic net­works as MANET. Every mobile agent com­putes the trans­mis­sion capac­i­ty of all the nodes so that Rout­ing Agent Sys­tem (RAS) can take the effi­cient reli­able deci­sion which rout­ing path is more effi­cient and reli­able. Each node has its own sta­tion­ary agents but num­ber of mobile agents in the net­work depends on the net­work archi­tec­ture or the pro­to­col used.

The trans­mis­sion capac­i­ty fac­tor into the net­work­ing as MANET of the pro­to­col will need to improve in future [6].

Sur­vey­ing dif­fer­ent tech­niques we define the Advan­tages and Dis­ad­van­tages of tech­niques in the table:

Tech­niques	Advantages/                                Mer­its	Dis­ad­van­tages /Future Improve­ment
MANET, AODV, Trust­ed Net­works; Trust Mod­el	The pro­posed approach is the exten­sion of exist­ing AODV rout­ing pro­to­col for cre­at­ing secure route for com­mu­ni­ca­tion. Pro­posed mod­i­fi­ca­tions are in accept­able lim­it. With this min­i­mum over­head, we can eas­i­ly elim­i­nate the mali­cious node as well as they can estab­lish a best trust­ed route between source and des­ti­na­tion.	Using sim­u­la­tion results, the per­for­mance of this pro­to­col is not suf­fi­cient jus­ti­fied. In the future, it will be incor­po­rate with oth­er MANET rout­ing pro­to­cols [1].
DAAODV, Secure Rout­ing Pro­to­col	They pre­sent­ed a secure ad hoc rout­ing pro­to­col which can pre­vent most attacks includ­ing worm-hole attacks, ver­tex cut attacks, and traf­fic analy­sis attacks, and adopt a new effi­cient sign­ing and ver­i­fy­ing cheme pre­vent­ing DoS attacks.	This pro­to­col doesn’t use TTP, and doesn’t add much over­head in ns‑2 sim­u­la­tion. In future work is to make a fine-grained con­struc­tion of the rout­ing soft­ware, as the design of DAAODV on soft­ware lev­el is a lit­tle coarse- grained [2].
Mul­ti­path Rout­ing, Ad-hoc Net­works, AOD­Vsec	The results show that AOD­Vsec out­per­forms tra­di­tion­al mul­ti­path rout­ing on ensur­ing secu­ri­ty. As a com­mon case, attack­er can­not inter­cept all the paths, AOD­Vsec avoids mali­cious­ly access­ing a entire data pack­et, so it improves system’s secu­ri­ty with neg­li­gi­ble rout­ing over­head.	The AOD­Vsec still has some imper­fect points. As a future work, it will need to focus on design­ing the syn­chro­niza­tion con­trol Mech­a­nism to solve this prob­lem [3].
lessSsecurity1,; MANE, IEEE 802.11b4	The effi­cient secu­ri­ty algo­rithm ES-AODV enhances the secu­ri­ty in ad hoc wire­less net­works. Accord­ing to the analy­sis of the results obtained from exten­sive sim­u­la­tion, it con­cludes that the secure rout­ing solu­tion scales well to both mobil­i­ty and net­work size.	The rout­ing pro­to­col per­forms Does not bet­ter than the exist­ing secure AODV rout­ing pro­to­col with increased mobil­i­ty in the net­work. It should be improve in future exten­sion [4].

MANET, Rout­ing, Secu­ri­ty	In the imple­men­ta­tion of such rout­ing pro­to­cols, the need is to elim­i­nate the short­com­ing     of     these                    pro­to­cols     by eval­u­at­ing per­for­mance of them on a sim­u­la­tion                      plat­form.      To     min­i­mize      the asso­ci­at­ed                    over­head    like                     delay,   rout­ing over­head demands an inten­sive opti­miza­tion in both the pro­to­cols.	In future it will require more specif­i­cal­ly SAODV                                      to decrease            the                          pro­cess­ing require­ments to tack­le hash chains and dig­i­tal sig­na­tures to imple­ment the secu­ri­ty [5].
MANET, Secure AODV	This paper, presents the pro­to­col being pro­posed                     which       uti­lizes                     the            dual coop­er­a­tive mobile agents and sta­tion­ary agents for rout­ing in dynam­ic net­works as MANET. Every mobile agent com­putes the trans­mis­sion capac­i­ty of all the nodes so that Rout­ing Agent Sys­tem (RAS) can take the effi­cient reli­able deci­sion which rout­ing path is more effi­cient and reli­able.	The trans­mis­sion capac­i­ty fac­tor into the net­work­ing as MANET of the pro­to­col will need to improve in future [6].

CONCLUSION

A MANET (Mobile Ad-hoc Net­work) is an autonomous col­lec­tion of mobile users that offers infra­struc­ture-free com­mu­ni­ca­tion over a shared wire­less medi­um. It is formed spon­ta­neous­ly with­out any pre­plan­ning. Mul­ti­cas­t­ing is a fun­da­men­tal com­mu­ni­ca­tion par­a­digm for group ori­ent­ed com­mu­ni­ca­tions such as video con­fer­enc­ing, dis­cus­sion forums, fre­quent stock updates, and video on demand, view pro­grams, and adver­tis­ing. The com­bi­na­tion of an ad hoc envi­ron­ment with mul­ti­cast ser­vices induces new chal­lenges towards the secu­ri­ty infra­struc­ture in rout­ing pro­to­cols. In order to secure mul­ti­cast com­mu­ni­ca­tion, secu­ri­ty ser­vices such as authen­ti­ca­tion, data integri­ty, and access con­trol and group con­fi­den­tial­i­ty are required. Among which group con­fi­den­tial­i­ty is the most impor­tant ser­vice for sev­er­al appli­ca­tions. These secu­ri­ty ser­vices can be facil­i­tat­ed if group mem­bers share a com­mon secret. Dur­ing the sur­vey on secure rout­ing pro­to­cols in MANET, we con­clude some points that can be fur­ther explored in the future using advanced secure tech­nique and it will improve the per­for­mance of secure MANET to achieve more effi­cient accu­ra­cy in net­work con­ges­tion, reduce the end to end delay time, over­head and through­put.

REFERENCES

• A.Menaka Push­pa, “Trust Based Secure Rout­ing in AODV Rout­ing Pro­to­col”, IEEE 2009.

• Wen­chao Huang, Yan Xiong, Depin Chen, “DAAODV: A Secure Ad-hoc Rout­ing Pro­to­col based on Direct Anony­mous Attes­ta­tion”, 2009 Inter­na­tion­al Con­fer­ence on Com­pu­ta­tion­al Sci­ence and Engi­neer­ing, IEEE 2009, pp. 809–916.

• Cuirong Wang, Shux­in Cai, and Rui Li, “AOD­Vsec: A Mul­ti­path Rout­ing Pro­to­col in Ad-Hoc Net­works for Improv­ing Secu­ri­ty”, 2009 Inter­na­tion­al Con­fer­ence on Mul­ti­me­dia Infor­ma­tion Net­work­ing and Secu­ri­ty, IEEE 2009,

pp. 401–404.

• Zeyad M. Alfawaer and Saleem Al_zoubi, “A pro­posed Secu­ri­ty sub­sys­tem for Ad Hoc Wire­less Net­works”, 2009 Inter­na­tion­al Forum on Com­put­er Sci­ence-Tech­nol­o­gy and Appli­ca­tions, IEEE Com­put­er Soci­ety 2009, pp. 253–255.

• Muham­mad Naeemv, Zah ir Ahmed, Rashid Mah­mood, and Muham­mad Ajmal Azad, “QOS Based Per­for­mance Eval­u­a­tion of Secure On-Demand Rout­ing Pro­to­cols for MANET’s”, 20 1O IEEE, ICWCSC 2010X.

• Preeti Bhati, Rin­ki Chauhan and R K Rathy, “An Effi­cient Agent-Based AODV Rout­ing Pro­to­col in MANET”, Inter­na­tion­al Jour­nal on Com­put­er Sci­ence and Engi­neer­ing (IJCSE), Vol. 3 No. 7 July 2011, pp. 2668–2673.

• Ming Yu, Mengchu Zhou, and Wei Su, “A Secure Rout­ing Pro­to­col Against Byzan­tine Attacks for MANETs in Adver­sar­i­al Envi­ron­ments”, IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 58, NO. 1, JANUARY 2009, pp. 449–460.

• D. Sug­anya Devi and Dr. G.Padmavathi, “IMPACT OF MOBILITY FOR QOS BASED SECURE MANET”,

Inter­na­tion­al jour­nal on appli­ca­tions of graph the­o­ry in wire­less ad hoc net­works and sen­sor net­works, pp. 46–57.