Service-oriented Dynamic Trust Model And Trust Management

With the in-depth researches of large scale distributed systems such as Gridcomputing Ubiquitous computing P2P computing Ad hoc networks etc thetechnology of dynamic trust management has become a significant requirement froma network economics’ point of view and trust evaluating and predicting mechanismhas become a determining factor for any given service’s success because staticmechanism such as public key infrastructure based on certificate authority issufficient no longer.The dynamic nature of trust creates the biggest challenge in measuring trust valueand predicting trust relationship amongst entities. In recent years many ofstate-of-the-art trust models have been proposed and some of them are veryinnovative and elaborate but most of the studies still have some limitations. Manycurrent trust models use simple or one-sided trust decision factors to quantify andpredict trustworthiness of service providers or requesters which may lead toinaccurate or unfair outcome of trust decision. When trust relationship betweenentities cannot be fairly defined it is unstable and difficult to manage and predict. Inmany of previous studies the subjective assigning method to weight s of trustdecision factors cannot reflect trust decision scientific and reasonable and may leadto misjudgment of trust decision result.Focusing on these problems in this paper a novel dynamic trust quantizationmodel with multiple decision factors is proposed in which multiple decision factorsare incorporated to reflect trust relationship’s complexity and uncertainty in variousangles. Also the weight of classification is set up for these decision factors, whichovercomes the shortage of traditional method in which the weight is set up bysubjective manners and makes the model has a better rationality and a higherpracticability.An important challenge regarding entities’ trust valuation is how to cope withstrategically altering behaviors and dishonest feedbacks of malicious peers efficiently. However，the trust models employed by the existing systems do not provide adequatesupport to coping with quick changes in entities’ behavior and aggregating feedbackinformation. In this paper，the author researches and improves dynamic trust models.After incorporating time dimension， which captures experience andrecommendation’s time-sensitivity，the author also introduce two trust parameters incomputing trustworthiness of entities，namely，short time trust and long time trust.Together，the parameters such as study factor and feedback weight are adjusted intime to reflect the dynamics of the trust environment using feedback controlmechanism，thus，the trust evaluation has better adaptability to the dynamics of trust．Trust is not only time-sensitivity，but also context-sensitivity. In this paper，theauthor systematically put forward a framework for evaluation，prediction and controlof user behaviors，including reliable evaluation of user behavior trust，trust decisionmaking to meet different prediction combinations of performance and security forservice providers. The mechanism can reduce the monitoring and preventionadditional costs for mutual trust between the service providers and requestors．In this paper，moreover，a recommended trust management model performs trustevaluation by compressing records of entities’ behaviors which contain moreinformation including time dimension，so it can exhibit good performance in trustevaluation and detecting malicious especially oscillatory behaviors. Moreover，thecapability of filtering false feedbacks is also improved. Theoretical analysis andsimulation show that，compared to the existing trust quantization metrics the model inthis paper has advantages in modeling dynamic trust relationship and aggregatingfeedback information．It is highly effective in countering malicious entities regardingstrategic altering behavior and dishonest feedbacks of malicious entities which hasremarkable enhancement s in the system’s security.