A study on impact of untrustworthy and older recommendations over T-Grid computational model

Trust can be either direct or indirect based on the direct impression established by the recipient of the services or feedback given by recommenders over the service provider respectively. However, untrustworthy recommendations may lead to misuse of valuable resources, especially in grid like distributed environment where continuous dynamic sharing of the resources is involved. Further, a genuine recommendation may become older over a period of time in a trust database and using such values in trust evaluation may lead to misinterpreted estimates of the trust. Hence, eliminating such untrustworthy and older recommendations from the interactions can construct a robust and reliable grid computing environment. Proposed T-Grid model eliminates untrustworthy recommendations from the grid while considering parameters like credibility and trust decay of a recommending entity. This paper analyzes the performance evaluation of proposed T-Grid model in terms of trustworthiness of the resource providers, effect of decay and utilization ratio of the resources. Results of the analysis are produced.