Additionally, the network topologies of these two scenes are different. In high speed mobile application scene, all users share common in and out communication links, while each user can be connected directly to base stations in dense users application scene.The content of this study includes: (1) Modeling P2P file sharing system in a network circumstance with bottleneck links. (2) Investigating utility-based resource allocation for mixed traffic in wireless access networks.The contributions of this paper are listed as follows: (1)In a Network Circumstance with Bottleneck Links, after analyzing the characteristic of P2P file sharing system (take BitTorrent system for an example), an improved fluid model is proposed, this model can be used to analyze the evolution of peers and predict the average file download delay of BitTorrent system which previous works seldom focus on. (2) Resource allocation for mixed traffic in wireless access networks is investigated and a Utility-based Resource Dynamic Allocation (URDA) algorithm is proposed. This algorithm considers the QoS characteristics, real-time resource requirements and wireless channel qualities that belong to different types of applications, and then generates utility functions of them. The resource is allocated dynamically according to these utility functions. Simulation results indicate that URDA works much better than traditional algorithms and achieves the target of resource allocation optimization. (3) After analyzing the QoS characteristic of P2P video and file sharing applications, these two types of P2P applications are both added into the utility-based resource dynamic allocation framework.
