Research Of Search Algorithmsbased On The Characteristics Of Complex Network Topologies

With the rise of information technology which represented by Internet, the study of complex networks has gradually aroused people’s attention. In many research areas of the complex network, searching problem is one of the most practical subjects. Involving finding the specified data or files or determining the shortest path between the nodes in the network, it provides theoretical guidance for such problems as designing the Internet search engine and the communicative action among person in social network. As the complex network has many different topological characteristics in general, this paper, search algorithms were studied deeply, analyzed and improved combined with different topology characteristics of complex networks.First of all, the maximum degree–the minimum distance searching algorithm was put forward. A parameters-adjustable scale-free network space model was established for the actual complex network containing metric space. On this basis, combined with the measure of the distance and the degree, this algorithm can not only make the direction of information transmission not deviate from the notes’ searching destination, but transfer the searching information to greater degree nodes which contain long-distance connection.Secondly, the minimum clustering coefficient- maximum degree searching algorithm was proposed.A boundary value that can make the searching order of nodes in the range as the assumption "according to the degree of sequence" was found through analyzing the cause of the maximum degree searching strategy. According to the boundary value, the notes’ degree and the aggregation coefficient were combined, which make the algorithm not only take the advantages of the maximum degree searching algorithm of the power index in a certain range, but also take into consideration the high efficiency of minimum clustering coefficient searching algorithm when beyond the range.Finally, using the C++ programming language in this paper, the simulation was done for several real complex network datasets which has different topological characteristic. According to two effective indexes, the average searching time and the average searching steps, several complex network strategies were compared, analyzed and evaluated to verify the correctness and the effectiveness of the maximum-minimum distance searching algorithm and maximum-minimum clustering coefficient searching algorithm.