Study Of Mobile Data Offloading Technology

The exponentially increasing data traffic of mobile users in wireless communication networks has brought great challenges to the limited wireless network resources of the network operator.In order to cope with this challenge,in this dissertation we study data offloading schemes for the network operator by mobile data offloading technology to alleviate the contradiction between traffic growth and limited resources.First,we propose a mobile data offloading scheme via which the network operator caches popular files on mobile users' devices and reduces the burden of network backhaul links through device-to-device(D2D)communications between users.Considering the energy limitation of user equipment(UE),we propose an energy consumption model based on continuous-time Markov chain to evaluate the power consumption of UE.In order to allow users to autonomously adjust the energy used for D2D data offloading,we propose a threshold-based energy saving strategy.At the same time,taking into account the user's selfishness,we adopt the VCG-based reverse auction technology to compensate users for the cost of participating in D2D data offloading.We give a j oint design model of user resource purchase and caching strategy that maximizes the system utility,and propose an algorithm to obtain the optimal solution of the model efficiently.The simulation results not only verify that our scheme can cope with the selfishness of users,but also save UE's energy while maximizing the system utility.On the basis of the first work,we further study the impact of adding privacy-preserving policies for users on system performance.User preference information for files is a key factor that affects the performance of caching strategies.As users pay more and more attention to their own privacy protection,it is difficult for the network operator to directly obtain user preference information for files.For the protection of user file preference privacy,we propose a privacy-preserving scheme based on K-anonymity technology and local differential privacy(LDP)technology.According to the design of the privacy-preserving scheme,the network operator can obtain user file preference information with privacy protection.According to the statistical model,the network operator estimates the probability distribution of user file requests through the collected user information,and optimizes the joint design of caching strategies and resource allocation strategies based on the estimated distribution.The simulation results show that the privacy-preserving scheme can not only protect users' private information,but also improve the overall performance of the network.Finally,we propose a data offloading scheme in which network operators lease resources of third-party wireless access points(APs)to provide services for users.The network operator spends a certain amount of budget to purchase resources of different third-party APs to provide services for mobile users in its own scenario.Previous works mainly focus on the determination of network resource allocation by maximizing the benefit/performance of the network,thus ignoring the fairness of users.In order to protect the fairness of users,we adopt a participatory budgeting(PB)method to make the decision on the network operator's budget allocation.At the same time,based on this budget allocation,we design a third-party AP resource allocation strategy that maximizes the amount of data offloading.Simulation results show that this scheme can not only protect the fairness of all users,but also achieves better performance than other schemes that protect fairness.