Study On Learning Driven Crowdsensing Privacy Protection And Incentive Mechanism

With the development of the Internet of things(IoT)technology,various smart devices with sensing and computing capabilities have increased dramatically,which has promoted the great development of crowdsensing.Smart devices in crowdsensing collect data,analyze data and provide computing power to complete the sensing tasks delivered by the platform.However,in this process,participants may disclose their own sensitive information.Due to the intention of protecting privacy,the participants' enthusiasm for participating in sensing tasks is reduced,thereby seriously reducing the effectiveness of the platform.In addition,due to the diversity of participants and terminals in the crowdsensing network,therefore,it is particularly important to design a reasonable sensing task allocation strategy and incentive mechanism to improve the participants' task completion.In view of the outstanding problems of the current crowdsensing network in terms of privacy protection,low participation of participants,and unreasonable allocation of sensing tasks,the main innovative contributions of this article include the following two aspects:(1)In response to problems such as the low utility of the platform and the low participation enthusiasm of participants due to suspected participants and insufficient compensation of the platform participants,a payment privacy protection level(PPL)game is proposed,in which each participant uses a specific PPL submit your own sensing data,and then the platform selects the corresponding payment(Payment).Additionally,we derive the Nash equilibrium(NE)point of the game.Considering that the payment-PPL model is unknown in practice,we employ a reinforcement learning technique,i.e.,Q-learning to obtain the payment-PPL strategy in a dynamic payment-PPL game.We further use deep Q network(DQN),which combines a deep learning technique with Q-learning to accelerate learning speed.Through extensive simulations,we verify that our proposed algorithm using DQN achieves superior performance in terms of utilities of both platform and participants and data aggregation accuracy compared with the using Q-learning.(2)Directing at the problem of extremely low sensing quality due to the leakage of participants'privacy,an incentive mechanism for participants is proposed,which protects participants' privacy from leakage by assigning different sensing tasks to different participants.To ensure the availability of sensing data,so as to maximize the utility of the platform and participants.More specifically,we formulate the interactions between platforms and participants as a multi-leader multi-follower stackelberg game and derive the stackelberg equilibrium(SE)of the game.Due to the difficulty to obtain the optimal strategy,a reinforcement learning algorithm,i.e.,Q-learning is adopted to obtain the optimal sensing contributions of participants.In order to accelerate learning speed and reduce overestimation,a deep learning algorithm combined with deep Q network in a Dueling architecture,i.e.,double deep Q network with Dueling architecture(DDDQN)is proposed to obtain the optimal payment strategies of platforms.To evaluate the performance of our proposed mechanism,extensive simulations are conducted to show the superiority of our proposed mechanism compared with state-of-the-art approaches.