The model of interaction among consumers and remote computational resources by means of brokers

A new mathematical model is proposed for fl ow control and pricing analysis as well as for optimization in distributed multiple-access computational resource systems. The model represents dynamic balance fl ow relations. System participants are independent agents pursuing their own goals stated in terms of cost and quality of service. Distributed decentralized gradient algorithm is offered for optimization task solution. Numerical experiments results and model improvement directions are discussed. Extensive references list is included.