With the growth of multimedia devices and applications, we have more and more available media data distributed in computers in a house or even a building, so we need an efcient way to access and integrate those data. In this thesis, we propose to use a P2P network architecture for multimedia streaming. In the P2P network, every- one can contribute his own resources such as bandwidth and storage capacity. Thus, we can integrate the multimedia data among a community. However, the resources in a P2P network are time-varying. To guarantee the streaming quality, we design a mechanism, Rate Adaption Bandwidth Allocation, to manage resources among the nodes in the system. Rate Adaption Bandwidth Allocation performs three main func- tions: (1) selecting reliable nodes for incoming multimedia streaming; (2) scheduling the chunk requests to meet timing constraints of video playback; (3) generating the bandwidth allocation for nodes to reserve bandwidths. We implement Rate Adaption Bandwidth Allocation in a real-world P2P system. We evaluate the performance of Rate Adaption Bandwidth Allocation and, and discuss lessons learned from our expe- rience with respect to the practicality and further optimization of our system.