Algorithms Based Trust And Economic Model For Manufacturing Grid Resource Scheduling

In manufacturing grid environment applying economic theories to resource scheduling is necessity from theory research to practical application.Economic model introduces economic concept to manufacturing grid resource scheduling and uses economic law to optimize resource provider's and resource user's allocation.In the promotion of maximum use of resources,economic model also encourages resource provider and user to participate in the construction and application of manufacturing grid initiative.As manufacturing grid is established on the internet,with its trading characteristics of the obvious characteristics of online transactions,as owner of money,buyer dominates transaction process through a repeated comparison of goods, then finally chooses a quality goods at low cost.In addition,when the manufacturing grid introduced into the practical application of economic models,the internet security issues of application also have to be worried about.From the user's point of view is that network security system evaluation is based on the trust relationship between users,which could forecast a direct relationship of trust grid users in the transaction by the two sides of the credibility of commitments to improve the success rate of the transaction.This kind of trusting relationship can directly forecast grid user both sides' confidence level of behavior which promised in the transaction,raise the transaction success ratio,and provide transaction process with the certain extent safeguard.A kind of manufacturing grid resource scheduling based on trust and economic model has been discussed in this paper as follow:choose an economic model giving priority to buyer's interests,given the evaluation mechanism of trust,which is to against the malicious node coordination fraudulent practice in the transaction though establishes the total consumption trust amount to suppress the malicious node to enhance the trust level through the ordinary service behavior;introduce the standard normal distribution in the evaluation of the recommendation confidence weight,and evaluate the probability of occurrence instead of the importance.The proportion weight of the recommendation confidence is not a constant anymore.This method accord with the feature of the trade in reality is that with the increase of the frequency of direct transaction among buyers and suppliers,the proportion weight of the recommendation confidence will decrease;uses the most superior comparison method to revise a series of trust value to obtain the trust index and solve standardization and unification of trust value in reverse auction;uses trust index as a form of trust value in reverse auction.Finally,an algorithm based on trust and reverse auction for MGrid resource scheduling is given.Based on analysis of eight grid simulators' features,in view of the real grid environment's characteristic,this tool kit GridSim is used to establish our simulation environment.The final result of this experiment proved that the algorithm is reasonable.