Research On Informed Resource Searching Strategy In Unstructured MP2P Networks

With the continuous expansion of the Internet and the increasing popularity of mobile electronic devices,mobile peer-to-peer network technology,which combines wireless network communication with P2 P network technology,has become a hot issue in academic research and application.Decentralized unstructured P2 P network,which is more suitable for the mobile network environment with dynamic topology due to the self-organization and acentric characteristics,has been widely used in MP2 P network technology.Resource location is a core issue of MP2 P network technology.How to locate resources efficiently and accurately is a prerequisite for efficient operation of the entire system.So far,many resource search strategies have been proposed to solve this problem.Although these strategies can play a role in some specific network scenarios,most of them are obviously limited.Designing an efficient,stable,and secure resource search strategy is very important for MP2 P networks.In this paper,we propose an informed search strategy(IKRW)for unstructured MP2 P networks.In order to improve the success rate while reducing costs,and considering the dynamics and malicious behavior of MP2 P network nodes,IKRW defines three guiding factors: accessibility,stability and trust.Each node maintains a neighbor table that records its neighbor's guiding factor,and calculates them by using the interaction information.IKRW is an improved informed random walk search algorithm,requester employs k walkers to search for the required resource and each intermediate node forwards the query to one of its neighbors.The main difference with random walk is that a node makes use of guiding factor to probabilistically direct future walkers,instead of forwarding the walker at random.Since obtaining information of neighbors is a learning process,the search performance of IKRW search algorithm in the early network is poor.In order to solve this problem,we propose a supplementary flooding mechanism to ensure the success rate of the search when local information is insufficient at the beginning of the network while ensuring that nodes can collect more local information,so that the performance of the IKRW search algorithm can be quickly improved.Simulation experiment shows that our search strategy can adapt to the dynamic nature of MP2 P networks.It effectively reduces the number of redundant messages while ensuring the search success rate,and can also guarantee the service quality of the responder.