International Journal of Engineering Trends and Technology (IJETT) – Volume 4 Issue 10 - Oct 2013 An Empirical Trust Computation Method With Integrated Load Balancing Factor In P2P Networks Pallae Raja Sree Marthanda Kulapathy1, ,Vegi Srinivas2 1 M.Tech Scholar,2 Associate Professor 1,2 Dept of CSE, Dadi Institute of Engineering & Technology. Abstract:- For the recent developments in the scalable networks, security and privacy are the most important factors like p2p networks. Most of our approaches works on statistical measures like trust computations, (i.e. Direct and indirect trust computations), these approaches are not optimal, because anonymous user may not have (or use) same set of characteristics as previous connection. In this paper a proposal to an efficient and empirical trust computation method for identifying and allowing the nodes based on trust measures was done. Here, we are proposing an improved trust computation method by enhancing the traditional trust computation method with balancing factor and the results are shown at the last for the efficiency of proposed system. I. INTRODUCTION Peer-to-peer (P2P) electronic commerce (E-Commerce) communities can be seen as truly distributed computing applications in which peers (members) communicate directly with one another to exchange information, distribute tasks, or execute transactions. P2P E-Commerce communities can be implemented either on top of a P2P network [1, 2, 3] or using a conventional client-server platform. Gnutella is an example of P2P E-Commerce communities that are built on top of a P2P computing platform. Person-to-person online auction sites such as eBay and many business-to-business (B2B) services such as supply-chain-management networks are examples of P2P communities built on top of a client server computing architecture. Reputation Systems: In E-Commerce settings P2P communities are often established dynamically with peers that are unrelated and unknown to each other. Peers of such communities have to manage the risk involved with the transactions without prior experience and knowledge about each other’s reputation. One way to address this uncertainty problem is to develop strategies for establishing trust and develop systems that can assist peers in accessing the level of trust they should place on an E-Commerce transaction. For example, in a buyerseller market, buyers are vulnerable to risks because of potential incomplete or distorted information provided by sellers[4][5]. Trust is critical in such electronic markets as it can provide buyers with high expectations of satisfying exchange relationships. A recent study [9] reported results from both an online experiment and an online auction ISSN: 2231-5381 market, which confirmed that trust can mitigate information asymmetry (the difference between the amountsof information the two transacting parties possess) by reducing transaction-specific risks, therefore generating price premiums for reputable sellers. Recognizing the importance of trust in such communities, an immediate question to ask is how to build trust. There is an extensive amount of research focused on building trust for electronic markets through trusted third parties or intermediaries [14, 8]. However, it is not applicable to P2P e-Commerce communities where peers are equal in their roles and there are no entities that can serve as trusted third parties or intermediaries. Reputation systems [11] provide a way for building trust through social control without trusted third parties. Most research on reputationbased trust utilizes information such as community-based feedbacks about past experiences of peers to help making recommendation and judgment on quality and reliability of the transactions. Community based feedbacks are often simple aggregations of positive and negative feedbacks that peers have received for the transactions they have performed and cannot accurately capture the trustworthiness of peers. In addition, peers can misbehave in a number of ways, such as providing false feedbacks on other peers. The challenge of building a trust mechanism is how to effectively cope with such malicious behavior of peers. Another challenge is that trust context varies from communities to communities and from transactions to transactions. It is important to build a reputation based system that is able to adapt to different communities and different situations. Furthermore, there is also a need for experimental evaluation methods of a given trust model in terms of the effectiveness and benefits. Most traditional trust models only give an analytical model without any experimental validation due to the subjective nature of trust. There is a need of general metrics for evaluating the effectiveness and benefits of trust mechanisms. With these research problems in mind, we develop Peer Trust, a peer-topeer trust model for quantifying and assessing the trustworthiness of peers in P2P E-Commerce communities. Our goal is to build a general trust metric that provides an effective measure for capturing the trustworthiness of peers, addresses the fake or misleading feedbacks, and has the capability to adapt to different communities and situations. http://www.ijettjournal.org Page 4502 International Journal of Engineering Trends and Technology (IJETT) – Volume 4 Issue 10 - Oct 2013 II. TRADITIONAL TRUST MEASURES A) Interaction trust Interaction trust (IT) models the trust that ensues from the direct interactions between two agents. Here we simply exploit the direct trust component of Regret [10] since this meets all our requirements for dealing with direct experiences. In more detail, consider a commercial transaction where agent a buys a particular product from agent b. The outcome of the transaction may consist of the product price, product quality, and the delivery date. From this outcome, agent a may give ratings about agent b’s service in terms of price, quality, and delivery for that particular interaction. Ratings are thus tuples in the following form: r = (a, b, I, c, v); where a and b are the agents that participated in the interaction i, and v is the rating a gave b for the term c (e.g. price, quality, delivery). The range of v is[¡1; +1], where 1 means absolutely negative, +1 means absolutely positive, and 0 means neutral or uncertain. In order to calculate IT from past experiences, an agent needs to record its past ratings in a (local) rating database. This database stores at maximum the H latest ratings that the agent gave to the each partner that it has interacted with. Here H is called the local rating history size. When calculating the IT value for agent b with respect to term c, agent a has to query its database for all the ratings that have the form (a, b, c ), where the ‘ ’ symbol can be replaced by any value. We call the set of those ratings RL(a, b, c). Then the IT (denoted by TI ) is calculated as the weighted mean of the rating values of all the ratings in the set: TI (a, b, c) =Xri2RL(a, b, c)(ri) ¢ vi; (1)where vi is the value of the rating ri and (ri) is the weight corresponding to ri. The weight (ri) for each rating is selected such that it gives more weight to more recent ratings, with a constraint that Pri2RL(a ,b, c) (ri) = 1. This is to ensure that the trust value TI (a, b, c) is in the range[1,+1]. B) Role Based Trust Role-based trust (RT) models the trust resulting from the role-based relationships between two agents (e.g. owned by the same company, a service provider and its registered user, friendship relationship of their owners). Since there is no general method for computationally quantifying trust based on this type of relationship, we use rules to assign RT values. This means end users can add new rules to customize this component to suit their particular applications. Rules are tuples of the following form :rul= (role a, role b, c, vD , eD), which describes a rule that if role a and role b are the roles of agent a and b respectively, then the expected performance of b in an interaction with ais vD(vD 2 [1, 1]) with respect to the term c; eD 2 [0, 1] is the default level of influence of this rule on the resulting RT value. ISSN: 2231-5381 For example, possible rules may be:rul1 = (buyer; seller; quality;0:2,0:3);rul2 = (friend-buyer; friend-seller; quality; 0;,0:6), andrul3 = ( ; government-seller; quality; 0, 0:9): rul1 expresses an agent’s belief that an ordinary seller will usually sell a product of slightly lower quality than agreed, but the reliability of this belief is low (0:3); rul2 is the belief that in a close partnership the buying agent can expect the seller to do what is agreed in terms of product quality; and this is also true for a governmental seller almost all of the time (rul3).Each agent has its own set of rules which are stored in a (local) rule database. In order to determine the RT with an agent b, agent a looks up the relevant rules from its rule database. Then the value of RT is given by the following formula: TR(a, b, c) =Pruli2Rules(a;b;c) eDi ¢ vDiPruli2Rules(A;B;c) eDi; (5)where ruli= (role a; role b; c; vDi; eDi) is a rule in the set of rules Rules(A;B; c). This set is a subset of the rule database in which only the rules that are relevant to the roles of a, the roles of b, and the term care selected. Since the rules for RT are specified by the agent’s owner, the reliability of RT also needs to be set by the agent’s Owner. We use TR(a; b; c), again in the range [0, 1], to denote this value. A reliability measure that can be used for role-based trust is: TR(a; b; c) =Pruli2Rules(a;b;c) eDijRules(a, b, c)j; (6)where jRules(a, b, c)j is the cardinality of the set of rules. C) Certified reputation Certified reputation(CR) are ratings presented by the rated agent (agent b) about itself which have been obtained from its partners in past interactions [6]. These ratings are certifications(provided by the rating agents) of agent b’s past performance (somewhat like a reference when applying fora job). They allow an agent to prove its achievable performance as viewed by previous interaction partners2. Since agent b can choose which ratings it puts forward, a rational agent will only present its best ratings. Therefore, we should assume that CR information probably overestimates an agent’s expected behaviour. Thus, although it cannot guarantee agent b’s performance in future interactions, the CR information does reveal a partial perspective on agentb’s past behavior. The main benefit of this type of information is its high availability. With the cooperation of its partners, agent b can have CR information from just a small lnumber of interactions. Therefore, CR is available to agents in most circumstances; even in situations where the other components may fail to provide a trust measure. Now a day’s security and privacy issues hav ebecome critically important with the fast expansion of multi agent systems(MAS).Most network applications such as grid computing , and p2p n/w ‘s can be viewed as MAS which are open, anonymous, and dynamic in nature. Such characteristics of MAS introduce vulnerabilities and threats to providing secured communication when malicious agents appear. One feasible way to minimize the threats is to evaluate the trust and reputation of the interacting agents http://www.ijettjournal.org Page 4503 International Journal of Engineering Trends and Technology (IJETT) – Volume 4 Issue 10 - Oct 2013 Most trust/computation models works on statistical measures(i.e. Direct and indirect trust computations)which are not optimal to address this issue. So, to cope with the strategically altering behavior of malicious agents; we present in this paper an efficient trust computation method, which is based on classification for analyzing the nodes or user when connects to the network. D) Trust Computation with Load Balancing Factor In this paper we are proposing an efficient trust computation method with load balancing factor, when nodes or agents tries to connect with service providers, connection needs to optimize the optimal service providers, for trust worthy various mechanisms implemented as trust measures for measuring the trust of the service providers. Even though trust measure is the basic measure we need to consider the load balancing factor, because even though service provider is good, if load is high, agent can not choose as its service provider because it may not leads to optimal results. Now a day’s identifying the unauthorized user in network is still an important research issue during the peer to peer connections. Networks are protected using many firewalls and encryption software’s. But many of them are not sufficient and effective. Most trust computation systems for mobile ad hoc networks are focusing on either routing protocols or its efficiency, but it fails to address the security issues. Some of the nodes may be selfish, for example, by not forwarding the packets to the destination, thereby saving the battery power. Some others may act malicious by launching security attacks like denial of service or hack the information. The ultimate goal of the security solutions for wireless networks is to provide security services, such as authentication, confidentiality, integrity, anonymity, and availability, to mobile users. This paper incorporates agents and data mining techniques to prevent anomaly intrusion in mobile adhoc networks. Home agents present in each system collect the data from its own system and using data mining techniques to observe the local anomalies. The Mobile agents monitoring the neighboring nodes and collect the information from neighboring home agents to determine the correlation among the observed anomalous patterns before it will send the data. This system was able to stop all of the successful attacks in an adhoc networks and reduce the false alarm positives. We define transaction success rate as a metric to measure the productivity and security level of a community. A transaction is considered successful if both of the participating peers cooperate. Otherwise one or both of the peers is faced with the risk of malicious behaviors from the other peer. The successful transaction rate is defined as a ratio of the number of successful transactions over the total number of transactions in the community up to a certain time. ISSN: 2231-5381 A community with a higher transaction success rate has a higher productivity and a stronger level of security. We expect that a community with an effective trust mechanism should have a higher transaction success rate as peers are able to make informed trust decisions and avoid unreliable and dishonest peers. Trust Computation Model (TCM) is based on a set of required factors to be known prior to making a trust decision [10]. These factors include information about other agents’ knowledge base, and interactions during the current task. A key feature of this model is that it considers how much knowledge one should have about the trustee agents in order to make a trust decision. TCM model defines direct trust based on three concepts: familiarity, similarity and past experience, whereas indirect trust is defined based on recommendations. Overall trust metric can be calculated with expected trust(i.e Expected trust reflects expected performance of the target agent and it is deduced from both recent and historical trust.) and deviation reliability.(i.e Deviation reliability is a measure of how much deviation we are willing to tolerate. Malicious agents sometimes strategically oscillate between raising and milking their reputation which seriously affects the performance of the network),coming to the load balancing factor, we are enhancing the previous load balancing factor with additional trust measures as by integrating the transaction trust and again recursively calculate the distance again if it is minimum then lamda. The following algorithm shows the load balancing factor with enhances trust measure as follows. III. LOAD BALANCING FACTOR Algorithm 1: Selection of service providing agent(p,S) Input: Evaluating agent p and the set of agents responding to a service request S Output: Service providing agent q. For each x € S do Trust(p,x) = ET(p,x)*DR(p,x) If(Trust(p,x)<¥) then Trust(p,x) = ET(P,x)*DR(p,x)+TransTrust(p,x). If(Trust(p,x)<¥) then UUu {x} Else GG u {x} End if Else http://www.ijettjournal.org GG u {x} Page 4504 International Journal of Engineering Trends and Technology (IJETT) – Volume 4 Issue 10 - Oct 2013 End if End for If G ≠ Ø then For each x € G do N(p,x)= I(p,x)+∑ FC(p,y)*I(y,x). End for Sort G in increasing order of lad N return agent q with the smallest load N else Total_trust0 For each x € U do Total_trustTotal_trust+ Trust(p,x) End for If Total_trust>0 then For each x € U do Compute Prob(p,x) End for return agent q with probability Prob(p,q) End if End if IV. RESULTS Our experimental results show the efficient performance than the traditional trust computation with previous load balancing factor. Performance evaluation of previous traditional and our proposed approach as follows: 5 4 3 2 1 0 Time Load Trust measures TA with TA with PA with out LB LB Factor LB Factor Factor TA-Traditional approach PA-Proposed approach LB-Load Balance V. CONCLUSION REFERENCES [1] N.R. Jennings, “An Agent-Based Approach for Building ComplexSoftware Systems,” Comm. ACM, vol. 44, no. 4, pp. 35-41, 2001. [2] R. Steinmetz and K. Wehrle, Peer-to-Peer Systems and Applications.Springer-Verlag, 2005. [3]Gnutella, http://www.gnutella.com, 2000. [4] Kazaa, http://www.kazaa.com, 2011. [5] edonkey2000, http://www.emule-project.net, 2000. [6] I. Foster, C. Kesselman, and S. Tuecke, “The Anatomy of the Grid:Enabling Scalable Virtual Organizations,” Int’l J. High PerformanceComputing Applications, vol. 15, no. 3, pp. 200-222, 2001. [7] T. Berners-Lee, J. Hendler, and O. Lassila, “The Semantic Web,”Scientific Am., pp. 35-43, May 2001. [8] D. Saha and A. Mukherjee, “Pervasive Computing: A Paradigmfor the 21st Century,” Computer, vol. 36, no. 3, pp. 25-31, Mar. 2003. [9] S.D. Ramchurn, D. Huynh, and N.R. Jennings, “Trust in Multi-Agent Systems,” The Knowledge Eng. Rev., vol. 19, no. 1, pp. 1-25,2004. [10] P. Dasgupta, “Trust as a Commodity,” Trust: Making and BreakingCooperative Relations, vol. 4, pp. 49-72, 2000. [11] P. Resnick, K. Kuwabara, R. Zeckhauser, and E. Friedman,“Reputation Systems,” Comm. ACM, vol. 43, no. 12, pp. 45-48, 2000. [12] A.A. Selcuk, E. Uzun, and M.R. Pariente, “A Reputation-BasedTrust Management System for P2P Networks,” Proc. IEEE Int’lSymp.Cluster Computing and the Grid (CCGRID ’04), pp. 251-258, 2004. [13] M. Gupta, P. Judge, and M. Ammar, “A Reputation System forPeer-to-Peer Networks,” Proc. 13th Int’l Workshop Network andOperating Systems Support for Digital Audio and Video (NOSSDAV’03), pp. 144-152, 2003. [14] K. Aberer and Z. Despotovic, “Managing Trust in a Peer-2-PeerInformation System,” Proc. 10th Int’l Conf. Information and KnowledgeManagement (CIKM ’01), pp. 310-317, 2001. [15] L. Mui, M. Mohtashemi, and A. Halberstadt, “A ComputationalModel of Trust and Reputation for EBusinesses,” Proc. 35th Ann.Hawaii Int’l Conf. System Sciences (HICSS ’02), pp. 2431-2439, 2002. We are concluding our proposed work by enhancing the load balancing factor with integrated trust computation for selection of the optimal service provider with balancing load factor and Secured Trust can ensure secured communication among agents by effectively detecting strategic behaviors of malicious agents. ISSN: 2231-5381 http://www.ijettjournal.org Page 4505 International Journal of Engineering Trends and Technology (IJETT) – Volume 4 Issue 10 - Oct 2013 BIOGRAPHIES Pallae Raja Sree Marthanda KulapathY is Pursuing M.Tech (CSE) from DIET College (JNTU Kakinada) and he obtained B.E.(Electrical & Electronics Engg) with Distinction from S.T.J.I.T. College Dharwad Dist,Karnataka.(Karnatak University).India. Vegi Srinivas received his M.Sc. and M.Tech.in Computer Science and Technologyfrom the Dept. of Computer Science and Systems Engineering, College of Engineering, Andhra University. He is currently doctoral candidate in the Department ofComputer Science & Engineering, JNTU, Kakinada, India. He is working as Associate Professor in the Dept. of Computer Science and Engineering, Dadi Institute of Engineering and Technology, Anakapalli, Visakhapatnam, India. His main areas of interests are Security, Privacy, Trust and Cloud Computing. ISSN: 2231-5381 http://www.ijettjournal.org Page 4506