Research On P2P Network Resource Monitoring Using Improved Kademlia Algorithm

The shared data resources in P2P networks are directly transmitted between network nodes without the control of central server.Although P2P network has more advanced performance than those centralized system in fault tolerance,attack resistance and resource utilization due to its decentralization feature,it also exists the problem that the resources transmitted in P2P networks are difficult to monitor and track.Therefore,how to efficiently monitor the concerned information in P2P networks,how to increase the rate of resource messages received by nodes,and how to reduce the system overhead have become key issues in P2P network content monitoring.The thesis researches on P2P technology and Kademlia protocol,aiming at problems that bottlenecks in traditional resource discovery model that are summarized as being unable to adapt to the randomness of P2P networks,the low efficiency of resource discovery,poor comprehensiveness and the lack of efficient monitoring when a large number of resources exist in networks.Based on Kademlia protocol,an improved Kademlia algorithm was proposed in the thesis.The improved algorithm interacts with P2P node in the network by dynamically changes the node ID,stores to the specific location of the routing table of the other nodes,which increase the rate of receiving resource searching messages or resource downloading messages between nodes,so as to find more resource messages transmitted in P2P network.Experimental results show that the improved Kademlia algorithm finds the location of node routing table receiving the largest number of resource messages between nodes,and verifies the feasibility and efficiency of the improved algorithm.Aiming at problems that the high failure rate of resource collection and high repetition rate of resource collection in resource monitoring,the thesis designed a new resource collection strategy to improve the resource collection efficiency and reduce the system overhead.To solve the problems of low discrimination efficiency and low accuracy in the process of discriminating illegal resources,the deterministic finite automaton algorithm is used to improve the accuracy of resource discrimination and keep the processing time at the millisecond level to achieve real-time and accurate resource information discrimination.The thesis developed a set of P2P network content monitoring software based on the above research.According to resource discovery,resource collection,resource discrimination,and resource query statistics,the software employed modular development method to complete the overall design and detailed design of the entire system.The developed software is tested in the real P2P network,which proved that the functions and performances satisfied the actual requirements of the monintoring system.