A reputation computation model for trusted networks considering uncertainties based on cloud theory

In trustable networks, the process that nodes submit service satisfaction degree to form reputation of service provider has uncertainties of fuzziness and randomness. A cloud theory based model is proposed for the computation of node reputation. Weights of service satisfaction degrees in the model include an attenuation coefficient part and a certainty degree part. The certainty degree part is obtained through algorithm of positive reputation cloud of evaluated node. It is a random number even under a given service satisfaction degree value. The randomness can describe the uncertainties of submitting service satisfaction degrees. Weights designed under such certainty degrees also have randomness under given service satisfaction degree value. The randomness indicates the trustable extent of a submitted service satisfaction degree. Results compared with other reputation models show that the proposed model has good stability.