A Survey on Trust Management for Mobile Ad Hoc Networks

DOI : 10.17577/IJERTCONV4IS24012

Download Full-Text PDF Cite this Publication

Text Only Version

A Survey on Trust Management for Mobile Ad Hoc Networks

J. Sathiya Jothi Department of Information

Technology,

Anjalai Ammal Mahalingam Engineering College,

Thiruvarur-614403, Tamil Nadu, India.

  1. Senthilkumar*

    Department of Computer Science and Engineering,

    University College of Engineering Pattukkottai,

    Thanjavur-614701, Tamil Nadu, India.

    1. Rajesh

      Department of Mathematics, University College of Engineering Pattukkottai,

      Thanjavur-614701, Tamil Nadu, India.

      Abstract Mobile Ad hoc Network (MANET) is featured by the absence of centralized communication system to access the resources and sharing the information in between the wireless mobile nodes. In MANET, every node is dependent on the neighbor nodes for relaying the data towards the destination. So every mobile node is responsible for forwarding the data to neighbor node because of non-routing device in the system. In some situations, due to resource limitations, some nodes are neglecting the data to forward to another node. In this circumstance, the vital role is to detect the misbehavior nodes and induce them to involve in the communication system. We give a survey of trust management plans created for MANETs and talk about normally accepted classifications, possible attacks, performance metrics, and trust metrics in MANETs. At last, we confer future investigate areas on trust management in MANETs based on the idea of social networks.

      KeywordsTrust management; Mobile ad hoc networks; Trust; Trust Attacks; Secure Routing; Trust Metrics.

      1. INTRODUCTION

        In an increasingly distributed collaborative network world, could lead to efficient data transmission and information sharing in mobile ad hoc networks. It is composed of a set of independent devices that work as network nodes agreeing to forward packets for each other and have no central coordinator, absence of support infrastructure, dynamic topology and resource constraints and so on. These characteristics (memory size, processing power, battery life and unique wireless characters and others), force a node to be cautious when communicating with other nodes and its behavior .Nodes must act as a router, server, client, compelling them to cooperate for the perfect operation of the network.

        A node should be capable of self-configuration, self- managing and self-learning by means of collecting local information and exchanging information with its trustworthy neighbors. In MANETS, decision making trust the terms through input, process and output of the information. This trust must often be derived under time critical conditions, Environmental condition, and in a distributed way.

        The notion of trust, the degree of subjective belief about the behavior of a particular entity, trust relationship among participating nodes are critical in cooperative and

        collaborative environments to enhance system goals in terms of scalability,reconfigurability,reliability,availability,maintainabi lity,confidentiality,integrity and safety.(i.e. survivability). For each node, a trust relationship to all neighbors .The trust is based on previous individual experience of the node and the recommendations of its neighbors. The formation and prolonged existence of MANET services are mainly based on an individual nodes cooperating in packet forwarding. Therefore identifying and qualifying behavior of nodes in the form of trust essential for ensuring proper co-operation of MANETS.

        First introduced the term Trust Management and identified it as a separate component of service in networks and clarified that trust management provides a unified approach for specifying and interpreting security policies, credential and relationships

        Trust management in MANETS is needed when participating nodes, forwarding data to others nodes, there is no introduction about other nodes, desire to establish a network with an acceptable level of trust relationship among them.

      2. RELATED WORK

        Mobile Ad Hoc Network (MANETs) is a collection of mobile nodes connected with wireless links. MANET has no fixed topology as the nodes are moving constantly from one place to another place. In ad hoc networks, nodes can perform several actions, like sending packets, forwarding packets, responding to routing messages, sending recommendations, among others. The set of performed actions define the nodes behavior. Therefore, the learning plan monitors the neighbors actions trying to evaluate their behavior. All the nodes must co-operate with each other in order to route the packets. Cooperating nodes must trust each other.

        Trust is characterized as the degree to which one gathering is willing to participate in a given activity with a given accomplice, considering the dangers and motivators included. Reputation is characterized as a discernment a gathering makes through past activities about its expectations and standards. In the greater part of the writing, reputation management is viewed as a component of trust management. Further, the terms trust management and trust foundation are

        likewise reciprocally utilized. Aivaloglou et al., (2006) trust establishment is a procedure to manage the representation, assessment, upkeep, and dispersion of trust among nodes. Trust management manages issues, for example, the definition of assessment guidelines and strategies, representation of trust confirmation, and assessment and management of trust connections among nodes. Agreeing George et al.,(2006) The detail of permissible sorts of proof, the era, conveyance, disclosure, and assessment of trust confirmation are all in all called Trust Establishment.

      3. TRUST MANAGEMENT FOR MANETS

        This section discusses classifications, attacks and trust based solutions for trust management. Before reviewing the literature, we would like to clarify (some terminologies that have been used interchangeably but sometimes confusingly in the context of trust management.) trust related terms.

        Trust Classifications:

        Adams et al.,(2005) propose three sorts of reputation frameworks: positive reputation, negative reputation, and a mix of the two. Positive reputation frameworks just consider perceptions or input of the positive practices of a node. Negative reputation frameworks just record objections or perceptions of the negative practices of a node. Companions are thought to be trusted thus input on practices is utilized to adversely mirror a node's reputation.

        Aivaloglou et al.,(2006) groups two sorts of trust establishment frameworks for MANETs: certificate-based framework versus behavior-based framework. In the previous, components are characterized for pre-sending learning of trust connections inside of the network, utilizing certificates which are conveyed, kept up and oversaw, either freely or agreeably by the nodes. Trust choices can be made taking into account a legitimate testament that demonstrates reliability of the objective node by a certificate authority or by different nodes that the backer trusts. In behavior-based framework, every node consistently screens practices of its neighboring nodes with a specific end goal to assess trust. The behavior-based framework is a responsive methodology, working under the supposition that the personalities of nodes in the system are guaranteed by preloaded confirmation mechanisms. For instance, if a node utilizes system assets as a part of an unapproved way, it will be viewed as a selfish or malicious node, and will at long last be disconnected from different nodes.

        Aivaloglou et al.,(2006) likewise order trust establishment plans as far as te sort of architectures utilized: hierarchical framework versus distributed framework. In the previous, a hierarchy exists among the nodes taking into account their abilities or levels of trust. In this framework, unified declaration powers or trusted third parties are normally accommodated on-line or disconnected from the net proof. Such a centralized infrastructure does not exist in a distributed system; subsequently, every node has a few, potentially measure up to, obligation regarding obtaining, keeping up, and appropriating trust evidence.

        George et al.,(2006) recommends two diverse ways to deal with register trust (i)Proactive trust calculation utilizes

        more transmission capacity for keeping up the trust connections precise. Thus, the trust choice can be come to immediately. (ii) Reactive systems compute trust values just when explicitly required. The decision depends generally on the particular circumstances of the application and the network. For instance, if nearby trust values change a great deal more frequently than a trust choice should be made, then a proactive calculation is not favored: The transmission capacity used to stay up with the latest will be squandered, subsequent to the greater part of the processed data will be out of date before it is utilized.

        Li et al.,(2007) and Li et al.,(2008) classify trust management as reputation-based framework and trust establishment framework. A reputation-based framework utilizes direct observations and second-hand information disseminated among nodes in a network to evaluate a node. A trust establishment framework assesses neighboring nodes based on direct observations while trust relations between two nodes without earlier direct interactions are built through a mix of suppositions from intermediate nodes.

        Yonfang (2007) suggests two diverse approaches to assess trust: policy-based trust management and reputation-based trust management. Policy-based trust management is based on well-built and objective security schemes such as logical rules and verifiable properties programmed in signed recommendation for access control of users to resources. In addition, the access decision is typically on the basis of mechanisms having a well defined trust management language that has well-built verification and evidence support. Such a policy-based trust management approach usually makes a binary decision according to which the requester is trusted or not, and accordingly the access demand is allowed or not. Due to the binary nature of trust evaluation, policy-based trust management has less suppleness. Moreover, the accessibility of (or access to) trusted certificate authorities (CA) cannot always be assured, mostly for disseminated systems such as MANETs. On the other hand, reputation-based trust management utilizes mathematical and computational mechanisms to evaluate trust. Typically, in such a system, trust is planned by collecting, aggregating, and disseminating reputation among the entities.

        According to Li and Singhal, (2007) trust management can be classified as evidence-based trust management and monitoring-based trust management. Evidence-based trust management considers something that proves trust relationships between nodes these could incorporate public key, address, identity, or any confirmation that any node can create for itself or different nodes through a test and reaction process. Monitoring based trust management rates the trust level of each taking part node taking into account direct information (e.g., watching the considerate or malicious practices of neighboring nodes, for example, packet dropping, and packet flooding prompting intemperate asset utilization in the network, or denial of service attacks) as well as indirect information (e.g., notoriety evaluations, for example, suggestions sent from different nodes).Despite the fact that reputation management is a piece of trust management, numerous scientists further order reputation management plans.

        Govindan et al.(2012) Trust computations can be widely requested into the going with categories:1.Distributed trust computations: Every node enlists its own specific estimation of trust on its neighbor's 2.Centralized trust computations: a Central master administers/helps the node in trust calculations. Disseminated trust counts can be appointed: Neighbor sensing speak to a trustor depends all alone experience around a trustee. Recommendations based trust speak to a trustor depends on different nodes experience of managing a trustee and Hybrid technique. There are three method in distributed trust computations.(i)Neighbor sensing through the way of watching the neighbors conduct over the time, routing based direct trust computations, past activities and present conduct are joined in Bayesian evaluation to decide trust. (ii)Recommendation based trust (Indirect trust) taking into account nearby voting, trust assessment in light of controlled flooding suggestions. (iii)Hybrid method based on input suggestion and own assessments in P2P system, taking into account proposal accumulation furthermore neighbor detecting, estimation taking into account packet sending conduct, and works in light of both direct associations furthermore proves gathered. In centralized trust computation can be grouping based trust calculations ,Nodes inquiry the specialists for the beginning trust and after that compute the last trust esteem taking into account averaging, Cluster head totals the trust reports got from individual nodes and decides the last trust, Centralized Trust Block which gathers votes and ascertains the trust.

        Govindan et al.,(2012) Trust dynamics can be characterized by trust propagation, prediction and aggregation

        1. Trust propagation using mechanism learning along with web of trust, tiny world net, public neighbors, dispersed hash table, individual meetings are used for trust information Exchange, rendezvous based trust propagation It can serve as a first level data to set up a node to have cooperation with any unusual node. It can help nodes to create a new group and together battle the getting into mischief exercises. (ii)Trust aggregation using trust properties like subjectivity, iterated belief trust values, Weighted Ordered Weighted Averaging (WOWA) operator trust values, sequence and parallel operator trust values. It enhances exactness on the trust estimation.(iii) Trust prediction Uses Kalman filter theory to predict the future trust values, Kalman filter model based on aggregation and prediction, past actions are used to predict the future trust value using mathematical inductions. In that some interior parameters of the objective node are used in trust prediction. It helps the node to be careful to avoid any possible hazard whereas communicating with abnormal nodes and also increase more accuracy.

          Zhexiong Wei et al.,(2014) recommends trust management plot, the trust model has two parts: trust from direct observation and trust from indirect observation. With direct observation from a viewer node, the trust quality is inferred utilizing Bayesian inference, which is a form of vague calculation when the complete probability form can be mentioned. On the supplementary give, through indirect observation, which is called as secondhand information with the function of is obtained from neighbor nodes of the observer node, the trust value is calculated using the DempsterShafer theory (DST), which is a different type of

          uncertain reasoning when the proposition of attention can be formed by an indirect method.

          Antesar and Keshav,(2015) to sort out the disobedient nodes as searching for a packet release route. Here trust was calculated by removing dishonest recommendations between certain time, number of connections, compatibility of information and closeness between the nodes. The recommending node is picked in light of three components to check its genuineness: Recommendations are gathered over a time frame to guarantee the consistency of proposals gave by a prescribing node with respect to the assessed node. Bunching straegy is embraced to progressively sift through proposals between certain time allotments in light of: (i) Identification of confidence value using number of interactions (ii) Compatibility of data with the assessed node through deviation test (iii) Closeness between the nodes. Distinctive nodes are picked in the trust evaluation system to test the execution of the filtering against different movable topologies and neighbor nodes.

          The utilization recommendation based trust method can be profitable to nodes in finding making trouble nodes before collaboration, in this way maintaining a strategic distance from a potential awful ordeal. Recommendation can be classified into four categories. (i)Dishonest recommendation used CONFIDENT model. It applies the deviation test on the got suggestions and rejects the ones veering off over the limit esteem.(ii)Positive recommendation used CORE model, which just acknowledges positive proposal by other nodes.(iii).Negative recommendation is the main data traded between nodes.(iv). False recommendation used a trust-based motivating force model for self-policing versatile spontaneous systems to diminish the effect of false proposal on the precision of trust worth.

      4. TRUST MANAGEMENT ATTACKS

        In this section, we discuss different type of attacks and explain various features from the viewpoint of trust management.

        Active Attacks

        An active attack occurs when an illegal person modifies a message, change data stream within the system, or introduce invalid data into a system; destroy data already stored in the he system or file. It is characterized by the attacker attempting to smash into the system. Comparison of different types of active attacks in MANETs is provided in Table 1.

        Passive Attacks

        A passive attack is a network attack in which a framework is observed and infrequently examined for open ports and vulnerabilities. The design is exclusively to pick up data about the objective and no information is changed on the objective. It is characterized by the attacker listening in on communication. Here the hacker does not try to break into the method or else modify data. Comparison of different types of passive attacks in MANETs is provided in Table 2.

        Trust Management

        Trust Classification Trust Attacks Trusted Based Solutions

        Trust Establishment Trust computation Trust update

        Active attacks

        Passive attacks

          1. Reputation based

          2. SORI

          3. CONFIDANT

          4. RFTrust

            Certificate based Behavior based Hierarchical based

            Distributed based

            Reputation based

            Policy based

            Monitoring based

            Monitoring based

            Evidence based

            Distributed trust

            Neighbor sensing (Direct trust)

            Recommendation based(Indirect trust)

            1. Dishonest

            2. Positive

            3. Negative

            4. Hybrid

              Centralized Trust

              Centralized Trust

              Hybrid Trust

              Trust Agent Model

              1. Traffic Monitoring

              2. Traffic Analysis

              3. Eaves dropping

        1. Black hole Attack [29]

        2. Wormhole Attack [29]

        3. Replay attack [29]

        4. Bad mouthing attack (BMA). [18]

        5. Location-dependent attack [18]

        6. Camouflage attack (CA) [16]

        7. Ballot stuffing attack (BSA) [18]

        8. Sybil attack [16] [30] 9. [32]

        1. Blackmail [29]

        2. Denial of service attack[16]

        3. Gray-hole attack [29]

        4. Man-in-the-middle attack

        5. Routing loop attacks [29]

        6. false recommendation [29]

        7. Packet modification/insertion

        8. Newcomer attacks[16][32][33]

        9. Selective misbehavior attacks[18]

        1. CORE

        2. E-Hermes

        3. REP

        4. Recommendation

        5. Clustering based

        6. Trust-based incentive

        7. TSR

        8. TARP

        9. OLSR

        10. AODV

        11. CTrust

          Proactive Method

          Reactive Method

          1. Trust Propagation

          2. Trust Prediction

          3. Trust Aggregation

        Table 1: Types of Active Attacks

        ACITVE ATTACKS

        DESCRIPTION

        Newcomer attacks(2009)(2012)

        The attacker fools, by joining and leaving the system alternatively to gain a new hope and hide its prior bad records.

        Conflicting behavior attacks(2009)(2012)

        The reference for a good node can be spoiled by a wicked node by carrying out information in a different way to different peers.

        On-off attacks(2009) (2012)(2015)

        In order to get unobserved while making trouble to services, a malicious node may alternatively appear as outstanding and awful node.

        Blackmail (2011)

        A particular node is threatened by malicious node that it will propagate bad intuition about it .

        Gray-hole attack(2011)

        It is a unique case of black hole attack in which a frightful node may fussily drop packets.

        Black hole Attack (2011)

        A malicious node respond positive information for all kind of route requests

        Wormhole Attack (2011)

        It makes up connection with low-latency, disturbs routing information.

        Replay attack (2011)

        A replay attack (also known as playback attack) in which a legitimate data transmission is unspitefully repeated or delayed.

        Routing loop attacks (2011)

        The routing packets are amended by the malicious node. As a result it cannot reach the intended recipient.

        False information or false recommendation(2011)

        A malicious node plan to isolate good nodes by offering false Inference/ Information, at the same time keeping dreadful nodes connected.

        Incomplete information(2011)

        Possibly, a malicious node may not lend a hand in providing appropriate or comprehensive information.

        Denial of service attack(2012)

        The normal utilization or process of communication services may be obstructed by a malevolent node , for example-extreme resource utilization

        Camouflage attack (CA) (2012)

        Here, the fraudulent users tries to make trust by constantly informing what they observed .After gaining trust they behave fraudulently for particular instance.

        Sybil attack (2012) (2013)

        It can damage the entire topology through multiple references such as location based routing.

        Bad mouthing attack (BMA) (2015)

        A malicious node may unlawfully broadcast bad ratings for good nodes in order to smudge their character.

        Location-dependent attack (2015)

        From the individual property of trust i.e.( deeds at one location cannot affect assessing fidelity of nodes at another location) this attack has been derived

        Ballot stuffing attack (BSA) (2015)

        In this attack, a malicious node tries to provide fair rating for scantily performing node in order to deceive trust mechanism.

        Selective misbehaving attacks(2015)

        This attack proliferate fake ranking for some faithful nodes. Whereas, behave normal to other nodes.

        Man-in-the-middle attack

        An Intruder tries to access information passed between sender and receiver without their knowledge.

        Sometimes intruder tries to pretend itself as either sender or receiver.

        Packet modification/insertion

        Depraved packets such as packets with erroneous routing information may be inserted by malicious node on other hand it may also alter packets.

        Table 2: Types of Passive Attacks

        PASSIVE ATTACKS

        DESCRIPTION

        Traffic Monitoring

        It could provide information to launch further attacks.

        Traffic analysis

        Analyzing patterns of data transmission.

        Eavesdropping

        Attacker observe transmissions of message content, and may also include fake message in the network

      5. TRUST BASED SOLUTIONS

        In this section, we review various trust based solutions for secure routing.

        Secure Routing:

        Marti et al.,(2000) proposed a reputation based trust management plot that contains watch dog strategy utilizing which the practices of the nodes are observed notwithstanding that pathrater plan accumulates notoriety and give an answer about independent bamboozling node so as to discover find terrible assaults. It includes direct observation in light of expanded DSR.

        Buchegger et al.,(2002) Initiated another outline called CONFIDANT (Cooperation of Nodes-Fairness in Dynamic Ad-hoc Networks) a notoriety based trust establishment plan which tosses fiendish node through DSR by giving deceptive proposal. It considers any kind of suggestions of trust esteem than the obsession of limit worth negative

        proposals. At the point when the season of transmission among the nodes the wrong message is infiltrated .Hence this technique is badly designed.

        Michiardi and Molva,(2002) proposed CORE (Collaborative REputation) show, that consider idealistic suggestion instead of negative suggestion which prompts wasteful transmission.

        He et al.,(2004) planned SORI (Secure and Objective Reputation-based Incentive) a notoriety based trust management plan utilizing an incentive based method. This technique advance transmission of packets and deject self absorbed practices which relies on upon taking into account counted target measures and notoriety multiplication through a restricted hash chain based acceptance. However this might have a constraint in the existence of malicious node.

        Nekkanti and Lee,(2004) created AODV (Ad hoc On demand Distance Vector) using trust variable and security level at each node. Their procedure deals unmistakably with each course request considering the node's trust variable and

        security level. In a common arrangement, coordinating information for every requesting would be encoded provoking tremendous outlay; they recommend using unmistakable levels of encryption in light of the trust variable of a node, along these lines diminishing overhead. This procedure adjusts the security level in light of the apparent opposing vibe level and therefore can spare resources; regardless, the technique does not treat evaluation of trust itself.

        Li et al.,(2004) also enlarged AODV and received a trust model to make arrangements for malignant practices of nodes at the system layer. They address trust as supposition beginning from subjective method of reasoning. The inclination reflects the characteristics of trust in MANETs, particularly dynamicity. The key component is to consider system execution perspectives by dealing with each request in light of its level of trust. Dependent upon the level of trust of nodes incorporated into the request, there is no necessity for a node to request and check verifications always, thusly provoking paramount diminishment of estimation and correspondence overhead. This work advances trust organization by considering a non particular trust organization structure for MANETs.

        Pisinou et al.,(2004) considered a safe AODV-based routing convention for multi-hop specially appointed systems for finding a secured end-to-end course free of any exchanged off nodes. Their trust-based guiding tradition discovers trust values develop just in light of direct observations, expecting that trust must have the property of transitivity.

        Asad Amir Pirzada et al.,(2006) The primary thought of trust-based routing protocols is to find trusted course as opposed to secure courses. It is broke down by the trustworthiness of the nodes in the connection. Routing circles can be disposed of through utilization of arrangement numbers in AODV. In this method permit the portability operation which depict about the current circumstance of system.

        Wang et al.,(2008) prescribe a trust-based incentive model for self-policing versatile system frameworks to decrease the impact of false proposal on the exactness of trust quality. In any case, the execution of the model is not attempted against specific attacks, for instance, ballot- stuffing.

        Moe et al.,(2008) proposed a trust-based routing protocol as an increase of DSR considering a persuading power part that actualizes cooperation among nodes and diminishes the points of interest that self important nodes can acknowledge (e.g., saving resources by particularly dropping groups). This work is novel in that they used a covered Markov model (HMM) to quantitatively measure the dependability of nodes. In this work, extremist nodes are agreeable and particularly drop packets.

        Abusalah et al.,(2008) proposed a Trust-Aware Routing Protocol (TARP) and added to a trust metric in light of six trust parts including program structure, architecture of equipments, battery power, record as a purchaser, presentation and legitimate hierarchy of leadership. On the other hand, no musing was given to trust decay after some time and space to reflect powerlessness in view of

        components and divided information in MANET circumstances.

        Luo et al., (2009) proposed RFSTrust, which compute the reliability of node through a trust taking into account fuzzy suggestion. They propose relationship among nodes in light of the closeness idea. There is a relentless assessment between nodes when there is more similitude in the middle of assessing and suggesting node. It is not suitable to manage distinctive sorts of assault including immature node assault.

        Hermes (2009) is a proposition in light of suggestion trust demonstrates that uses an additional parameter known as a sufficiency edge (in association with the confidence level). The thought of value is used as a part of the computation of recommendation to ensure that adequate view of the behavior of taking an interest center point have been gained. Regardless, the determination of value is a tradeoff between getting more exact dependability regard and the combining time required to procure it.

        Recommendation Exchange Protocol was proposed by Velloso et al.,(2010) to allow nodes to send and get recommendations from neighboring center points. It displays the thought of relationship advancement considering to what degree nodes have known each other. Recommendations sent by whole deal accomplices are weighed higher than that from transient accomplices. The advancement of relationship is evaluated on the reason of a lone variable by considering only the period of time of relationship.

        Yu et al.,(2011)propose a clustering method to filter through trustworthy recommendations from deceptive ones. They take after the larger part direct by selecting the group with the greatest number of proposals as tried and true one. They attempted their model against a couple strikes like upbraiding and ticket stuffing. Then again, predominant part regulate could truly be destructive as a couple of nodes can plot to perform a strike, and not give a honest to goodness judgment about various nodes. Clustering strategy to dynamically filter through ambushes related to conniving recommendations between certain time in light of number of correspondences, comparability of information and closeness between the nodes.

        Huanyu Zhao(2013) proposed layout a response for cyclic improvement plan nodes in trust administration manets for upgrading precision and profitability. It takes the issue of trust affiliation and collection issues with cTrust Distributed Trust Aggregation Algorithm. It assembles the trust rate in trust route for each one of the nodes in the framework than the neighbors trust associations. Trust rating limits considering distinctive parts in the recorded trades, for instance, noteworthiness, nature of organization, time, and area. In additionally, it compute how to rate an administration and how to make the exact and stable direct trust assessments. The cTrust all out arrangement impacts a stochastic coursed BellmanFord calculation to achieve fast and lightweight trust rating mixture. The idea of Ctrust, depends on two categories.1) The improvement plans and trust associations in cMANET as a trust outline and show the most trustable way (MTP)- discovering process as the Markov decision technique (MDP). 2. Trust trade limit,

        regard cycle work, and spread trust aggregation estimation to deal with the MTP-finding issue. This figuring uses a stochastic Markov-chain-based methodology, which fundamentally diminishes the message overhead. It requires simply adjacent correspondence between neighbor nodes and gets a brief delineation of the whole framework from each node's point of view.

        Zhexiong Wei et al.,(2014) proposed new thought of trust organization plan for uncertain deduction to upgrade more accuracy trust regard in the viewer node. It incorporates two sorts of trust modules.(i) Direct observation viewer node can choose trust estimations of its neighbors by using Bayesian obstruction, which is a general framework to infer the estimation of the dark probability by using examination. (ii) In the indirect observation, which is in like manner called utilized information is gotten from viewer nod and convenient node, the trust worth is induced using the DempsterShafer therory(DST)

        Antesar and Keshav,(2015) Proposed a suggestion based trust management to deal with the getting out of hand nodes and secure the coordinating tradition amidst source and destination nodes when the season of package transport by using Bayesian quantifiable philosophy, filtering algorithms. The issue of data sparsity can be shed by disengaging recommendations using dynamic grouping methods. It relies on upon number of correspondences, pleasing information with the appraisal center nodes, and closeness between the center nodes.

      6. CONCLUSION

Trust and its management are energizing fields of examination. The rich literature developing around trust gives us a well-built indication that this is an imperative range of research. Trust as an idea has a wide collection of adjustments and applications, which causes dissimilarity in trust management phraseology. The objective of this paper is to give MANETs architects various points of view on the idea of trust, a comprehension of the properties that ought to be considered in developing a trust metric, and bits of knowledge on how trust can be processed. We begun this paper by introducing different meanings of trust also, measurements utilized for assessing trust.

We then exhibited a exhaustive overview of different trust computing approaches, their correlations regarding different attack models and computational requirements.

We discussed different literature on the trust establishment, for example, trust computation, updating and predictions. In this paper, we studied and dissected existing trust management plans in MANETs to give MANET trust system convention fashioners with different points of view on the idea of trust, a comprehension of trust properties that should be seen in creating trust measurements for assessing trust, and bits of knowledge on how a trust metric can be tweaked to meet the necessities and objectives of the focused on framework. A composite trust metric that catches parts of correspondences and interpersonal organizations, and comparing trust estimation, trust conveyance, and trust administration plans are intriguing exploration headings. For dynamic networks, such as

military MANETs, these plans ought to have alluring credits, for example, capacity to adjust to ecological progress, adaptability, unwavering quality, and reconfigurability.

We expect that the near potential will bring consolidation around a set of basic principles for building trust and its a range of related issues, and that these will be realized in practical and commercial applications.

REFERENCES

  1. Abusalah.L,Khokhar.A,&Guizani.M.,(2008).A Survey of Secure Mobile Ad Hoc Routing Protocols. IEEE Commun. Surveys and Tutorials, vol.19, no. 4, pp.78-93.

  2. Adams.W.J Hadjichristofi G.C.,&Davis. N.J.,(2005).Calculating a Nodes Reputation in a Mobile Ad Hoc Network. Proc. 24th IEEE Intl Performance Computing and communications Conference, Phoenix, AX, pp. 303-307.

  3. Aivaloglou.E,Gritxalis.S.,&C.Skianis.,(2006).Trust Establishment in Ad Hoc and Sensor Networks Proc. 1st Intl Workshop on Critical Information Infrastructure Security, Lecture Notes in Computer Science,vol.4347, pp. 179-192, Samos, Greece, 31, Springer.

  4. Antesar M. Shabut, Keshav P. Dahal, Senior Member, IEEE, Sanat Kumar Bista, & Irfan U. Awan,(2015).Recommendation Based Trust Model with an Effective Defence Scheme for MANETs.IEEE TRANSACTIONS ON MOBILE COMPUTING, VOL. 14, NO. 10.

  5. Ben- Jye Chang, Member, IEEE, & Szu-Liang Kuo,(2009).Markov Chain Trust Model for Trust-Value Analysis and Key Management in Distributed Multicast MANETs.IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 58, NO. 4.

  6. Buchegger.S & Le Boudec.J,(2002).Performance Analysis of the CONFIDANT Protocol: Cooperation Of Nodes- Fairness In Dynamic Ad-hoc NeTwork Proc. 3rd IEEE/ACM Symposium on Mobile Ad Hoc Networking and Computing, Lausanne, CH, 9-11,pp.226-236.

  7. George.G ,Theodorakopoulos & Baras.J.S,(2006).On Trust Models and Trust Evaluation Metrics for Ad Hoc Networks. IEEE J. Sel. Areas Commun., vol. 24, no. 2.

  8. Grandison.T & Sloman.M ,(2000).A Survey of Trust in Internet Applications. IEEE Comm. Surveys and Tutorials, vol. 3, no. 4, pp. 2-16.

  9. He.Q ,Wu.D,& P. Khosla,(2004)SORI: A Secure and Objective Reputation-based Incentive Scheme for Ad-Hoc Networks, Proc. IEEE Wireless Communications and Networking Conf., vol. 2, pp. 825- 830.

  10. Huanyu Zhao, Xin Yang, and Xiaolin Li, Member, IEEE,(2013).cTrust: Trust Management in Cyclic Mobile Ad Hoc networks. IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 62, NO. 6.

  11. Jin-Hee Cho, Member, IEEE, Ananthram Swami, Fellow, IEEE,& Ing-Ray Chen, Member, (2011).A Survey on Trust Management for Mobile Ad Hoc Networks, IEEE COMMUNICATIONS SURVEYS & TUTORIALS, VOL. 13, NO. 4.

  12. Kamvar.S, Scholsser.M & H. Garcia-Molina,(2003).The Eigen Trust Algorithm for reputation Management in P2P Networks, Proc. 12th Intl Conf. World Wide Web (WWW).

  13. Kannan Govindan,Member IEEE & Prasant Mohapatra, Fellow IEEE,(2012).Trust Computations and Trust Dynamics in Mobile Adhoc Networks:A Survey IEEE COMMUNICATIONS SURVEYS & TUTORIALS, VOL. 14, NO. 2.

  14. Karim Rostamzadeh,Student Member, IEEE, Hasen Nicanfar, Student Member, IEEE, NarjesTorabi, Student Member, IEEE, Sathish Gopalakrishnan, Member, IEEE, & Victor C. M. Leung,Fellow,IEEE,(2015).A Context-Aware Trust-Based Information Dissemination Framework for Vehicular Networks, IEEE INTERNET OF THINGS JOURNAL,VOL. 2, NO. 2.

  15. Lee. S, Sherwood.R,& Bhattacharjee.B,(2003).Cooperative Peer Groups in NICE, Proc. IEEE INFOCOM, pp. 1272-1282.

  16. Li.R,Li.J, Liu.P, Chen.H.H,(2007).An Objective Trust Management Framework for Mobile Ad Hoc Networks, Proc. IEEE 65th Vehicular Technology Conf., 22-25, pp. 56-60.

  17. Li.J,Li.R ,& Kato.J,(2008).Future Trust Management Framework for Mobile Ad Hoc Networks: Security in Mobile Ad Hoc Networks, IEEE Commun. Mag., vol. 46, no. 4, pp. 108-114.

  18. Li.R, Li.J, Liu.P,& Kato.J,(2009).A novel hybrid trust manaement framework for MANETs in Proc. 29th IEEE Int. Conf. Distrib.Comput. Syst. Workshops , pp. 251256.

  19. Li.X, Lyu.M.R &Liu.J,(2004).A Trust Model Based Routing Protocol for Secure Ad Hoc Networks,Proc. IEEE Aerospace Conf., Bug Sky, Montana, 6-13, vol. 2, pp. 1286-1295.

  20. Li.H & Singhal.M.,(2007).Trust Management in Distributed Systems Computers, vol. 40, no.2, pp. 45-53.

  21. Michiardi.P &Molva.R.,(2002).Core: A Collaborative Reputation mechanism to enforce node cooperation in Mobile Ad Hoc Networks, in Proc.Commun. Multimedia Security Conf.,pp. 107121.

  22. Moe.M.E.G.,Helvik.B.E., &Knapskog.S.J.,(2008).TSR: Trust-based Secure MANET Routing using HMMs, Proc. 4th ACM Symposium on QoS and Security for Wireless and Mobile Networks, Vancouver, British Columbia, Canada, 27-28 , pp. 83-90.

  23. Nekkanti.R.K.,&Lee.C.,(2004).Trust-based Adaptive On Demand Ad Hoc Routing Protocol, Proc. 42th Annual ACM Southeast Regional Conf., Huntsville, Alabama, pp. 88-93.

  24. Pirzada &McDonald.C,(2004).Establishing trust in pure adhoc networks, in Proc. 27th Australasian Conf. Comput. Sci.,vol. 26, pp. 4754.

  25. Pirzada.A.,McDonald.C.,&Datta.A.,(2006).Performance Comparison of Trust-based Reactive Routing Protocols, IEEE Trans. Mobile Comput., vol. 5, no. 6, pp. 695-710.

  26. Pedro Velloso.B., Rafael Laufer.P.,Daniel de Cunha.O.,Otto Carlos Duarte.M.B.,&Guy Pujolle(2010). Trust Management in Mobile Ad Hoc Networks Using a Scalable Maturity-Based Model, IEEE TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT, VOL. 7, NO. 3.

  27. Pisinou.N.,Ghosh.T., &Makki.K.,(2004). Collaborative Trust-based Routing in Multi-hop Ad Hoc Networks,Proc. 3rd Intl IFIP-TC06 Networking Conf., Lecture Notes in Computer Science, Athens, Greece, 9-14 ,vol. 3042, pp. 1446-1451.

  28. Pissinou.N.,Ghosh.T.,Makki.K., (2004).Collaborative trust-based secure routing in multi hop adhocnetworks,inProc.Netw.Netw.Technol.,Services,Protocols;Perfor m.Comput.Commun.Netw.,Mobile Wireless Commun., pp. 14461451

  29. Sohail Abbas, MadjidMerabti, David Llewellyn- Jones,&KashifKifayat,(2013).Lightweight Sybil Attack Detection in MANETs,IEEE SYSTEMS JOURNAL, VOL. 7, NO. 2.

  30. Velloso.P.B.,Laufer.R.P., Cunha.D.,Duarte.O.C.M., & Pujolle.G., (2010). Trust management in mobile ad hoc networks using a scalable maturity-based model,IEEE Trans.Netw. Service Manage., vol. 7, no. 3, pp. 172185.

  31. Luo.J., Liu.X., & Fan.M.,(2009).A trust model based on fuzzy recommendation for mobile ad-hoc networks,Comput. Netw., vol. 53, no. 14, pp. 23962407.

  32. Marti.S,Giuli.T.,Lai.K.,,& Baker.M.,(2000).Mitigating Routing Misbehavior in Mobile Ad Hoc Networks,Proc. 6th Annual ACM/IEEE Mobile Computing and Networking, Boston, MA, pp.255-265.

  33. Wang.K.,Wu.M.,&Shen.S.,(2008). A trust evaluation method for node cooperation in mobile ad hoc networks,in Proc. 5th Int. Conf. Inform. Technol., New Generations, pp. 10001005.

  34. Younghun Chae, Lisa Cingiser DiPippo, Member, IEEE Computer Society,&Yan Lindsay Sun,Member,IEEE,(2015).Trust Management for Defending On-Off Attacks.,IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, VOL. 26, NO. 4, 32.

  35. Yu.H.,Liu.S.,Kot.A.C.,Miao.C.,,&Leung.C.,(2011).Dynamic witness selection for trustworthy distributed cooperative sensing in cognitive radio networks, in Proc. IEEE 13th Int. Conf. Commun. Technol.,pp. 16.

  36. Yu.H.,Shen.Z.,Miao.C.,Leung.C.,&Niyato.D.,(2010).A survey on trust and reputation management systems in wireless communications,Proc. IEEE, vol. 98, no. 10, pp. 17551772.

  37. Yunfang.F., Adaptive Trust Management in MANETs, (2007). Intl Conf. on Computational Intelligence and Security, Harbin, China, 15- 19, pp. 804-808.

  38. Zhexiong Wei, Helen Tang, Member, IEEE, Richard Yu.F., Senior Member, IEEE, Maoyu Wang, & Peter Mason.,(2014). Security Enhancements for Mobile Ad Hoc Networks with Trust Management Using Uncertain Reasoning, IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 63, NO. 9.

  39. Zouridaki.C, Mark.B.L,Hejmo.M, & R. K. Thomas.,(2009). EHermes: A robust cooperative trust establishment scheme for mobile ad hoc networks, Ad Hoc Netw., vol. 7, no. 6, pp. 11561168.

Leave a Reply