P2P traffic problem was analyzed to point out the reasons of P2P traffic problem firstly.Then the information of Internet Service Provider (ISP) and geographic position was used to design an intelligent node selection mechanism.In this mechanism, one node chose nodes which belonged to the same ISP and were in geographical proximity to the objects for exchanging data, so the mechanism could control the data flow in the local network and the same ISP's network to the greatest extent, thus reducing the network load, and improving data transmission performance.To prove the validity of intelligent node selection mechanism, Round-Trip Time (RTT) and number of hops were tested.The simulation results show that the mechanism can effectively reduce the "flow travel" and improve the P2P transmission efficiency.%对对等网络(P2P)的流量问题进行了分析,指出P2P流量问题产生的原因.融合网络运营商、地理位置等信息,提出了一种节点智能选择机制.该机制优先选择属于相同运营商和地理位置接近的节点作为数据交换对象,最大限度地将数据流量控制在本地网络和同一运营商的网络中,从而减少骨干网和网络出口的负载,同时提高了数据传输性能.在Bit Torrent(BT)系统中融合智能选择机制后,Tracker智能选择的节点在往返时延、路由跳数上均得到优化,实验数据表明该机制能有效减少"流量旅行",提高P2P传输效率.
展开▼
