TY  - JOUR
T1  - Fuzzy Based Resource Management in Broker Architecture Using Trust and Credentials
AU - , P. Varalakshmi AU - , S. Thamarai Selvi AU - , P. Kanchana AU - , J. Ramesh AU - , V. Pavithra 
JO  - International Journal of Soft Computing
VL  - 2
IS  - 4
SP  - 494
EP  - 501
PY  - 2007
DA  - 2001/08/19
SN  - 1816-9503
DO  - ijscomp.2007.494.501
UR  - https://makhillpublications.co/view-article.php?doi=ijscomp.2007.494.501
KW  - Policy-based
KW  -trust-index
KW  -presence-index
KW  -broker
KW  -SOA
AB  - Multi-broker grid environment gives very high importance to the concept of trust. This architecture consists of entities (Consumers and Service Providers (SP)) spread across the brokers with each SP connected (logically) to more than one broker and the consumer is free to contact a broker of its choice, when it requires a service. In this study, we propose to introduce a trust management architecture that not only supports a choice of SP based on reputation (trust-index) but also on credentials (policy).Thus trustworthy SP is assigned to a consumer that not only satisfies the consumer’s requirements but also its policy constraints. Thereby the consumer is assigned with a trustworthy SP and the transaction is free from runtime failure as the policies have been matched. The concept of registration of an SP to a broker has been dealt with. The concept of de-registration of an SP, has been introduced in two scenarios, one is an SP deregisters by choice and the other is the broker forcing deregistration of an SP. Publishing of brokers has been dealt with based on SOA. The SP’s trust-index is updated based on both the consumer’s feedback and the broker’s feedback. SP’s trust-index is updated periodically instead of after every transaction. The time for periodic updation of the trust-index of SP is set proportional to SP’s current trust-index. This helps to reduce the brokers’ workload and the number of messages exchanged The consumer’s trust-index is updated based on feedback from the SP. The model proposed allows the consumers to assign priorities to parameters while requesting for a service. Fuzzy logic has been incorporated to calculate the trust-index of SP in our model. This ensures the use of fuzzy inferences in the system which can handle any imprecise linguistic terms, in this case trust-index calculation effectively. This model shows a marked improvement in cost-loss reduction to the consumers and reduction in the number of messages exchanged in the environment.
ER  -