Research On Incentive Mechanisms In Large-scale Mobile Crowdsensing Via Social Network

The methods of data acquisition based on human-centered sensing and computation have become a new application mode and development trend of Internet,and can be used to solve many large-scale sensing and computing problems.Crowdsening is well known as an important data acquisition model in the age of big data.The mobile crowdsening applications rely on the participation and intelligence of a large number of smartphone users.To perform the crowdsensing tasks,the participants have to consume a variety of resources,such as traffic,electricity and memory resources.So incentive mechanism design is very necessary and important for most mobile crowdsensing systemsThe paper considers the context of the lack of users in the mobile crowdsensing system.The task diffusion in social network is processed to recruit more users to participate in mobile crowdsensing.According to the different demand,we build the system models,which follow reverse auction framework,for such scenarios,and formulate the problem.We propose the user selection and payment/reward computation algorithms based on game theory methord.The most important contribution of this thesis is that we present two incentive mechanisms for social network based mobile crowdsensing to achieve the properties of time sensitive and sybil proof.The time sensitive incentive mechanism uses second price rule to compute the payment to winners,and can obtain the optimal solution of maximizing the social efficience within polynomial time.To prevent the potential sybil attack by social users,a sybil-proof incentive mechanism is proposed.In sybil-proof incentive mechanism,the task subsets allocated can maximize the utility of users.Finally,through rigorous theoretical analysis and a large number of experimental simulations,it is proved that the two proposed incentive mechanisms achieve the target properties.