This paper analyzes and compares different incentive mechanisms for a masterto motivate the collaboration of smartphone users on both data acquisition anddistributed computing applications. To collect massive sensitive data fromusers, we propose a reward-based collaboration mechanism, where the masterannounces a total reward to be shared among collaborators, and thecollaboration is successful if there are enough users wanting to collaborate.We show that if the master knows the users' collaboration costs, then he canchoose to involve only users with the lowest costs. However, without knowingusers' private information, then he needs to offer a larger total reward toattract enough collaborators. Users will benefit from knowing their costsbefore the data acquisition. Perhaps surprisingly, the master may benefit asthe variance of users' cost distribution increases. To utilize smartphones' computation resources to solve complex computingproblems, we study how the master can design an optimal contract by specifyingdifferent task-reward combinations for different user types. Under completeinformation, we show that the master involves a user type as long as themaster's preference characteristic outweighs that type's unit cost. Allcollaborators achieve a zero payoff in this case. If the master does not knowusers' private cost information, however, he will conservatively target at asmaller group of users with small costs, and has to give most benefits to thecollaborators.
展开▼
