Home Engineering Modeling and Optimization for Mobile Social Networks
Strategy of Media Cloud in Stage I
The cloud resource can be sold to the broker to obtain revenue by media cloud. Thus, the media cloud hopes to choose a proper price of cloud resource to obtain the maximum utility. For media cloud, the optimization problem can be formulated as
where p* is the optimal strategy of media cloud on the price per cloud resource unit, E = [Ei, E2,E/]r is the vector of cloud resource purchased by each broker.
Similar to the broker iteration, we also present a media cloud iteration to adjust the cloud resource price to obtain the maximum utility. The media cloud updates its price by
where mr is used to control the speed of adjustment on the price of cloud resource price.
The marginal payoff can be calculated by
Here, when all users obtain the maximum utilities with the optimal strategies, the evolutionary game reaches the equilibrium. If someone tries to adjust his selection to connect the broker, the number of connection of this broker will become larger and the utilities of users in the same community to connect with this broker will decrease. If the equilibrium state is not Pareto efficiency, the utility of a user can be larger by adjusting strategy, where other users in the same community may imitate this selection to obtain higher utilities with the result that all users in the same community have the identical utility. In this case, the state in evolution game is not stable, which is not the equilibrium. Therefore, as the equilibrium can be obtained which is opposite to the above assumption, the equilibrium of evolution game in our work is Pareto efficiency. In addition, when the Stackelberg game reaches to the equilibrium, each broker or media cloud only has one optimal strategy. Therefore, each party can not adjust strategy to obtain higher utility when other two parties choose the optimal strategy. It also proves the Pareto efficiency of the proposed scheme.
|< Prev||CONTENTS||Next >|