Research On Complex Network Search Technology Combined With Max-degree And Random Walk Strategy

Complex networks exist in many fields of our life. Searching problem has always been one of an important research programs in complex network system theory. It is with a great deal of practical applications now. There is not useful global information in the target node in network searching actually. Then each node uses neighbor node’s identity or degree of local information to search, such as World Wide Web, so as to find the shortest path or a shorter path. It is important to study search program in complex network which focus on local search strategies. And local search strategies’ performance affects the searching efficiency and cost.Basic topological characteristics, topological models and search strategies in complex network were introduced and realized in this paper. We designed a new search strategy based on the defect causes of Max-degree seeking, verified its effectiveness on the existing date set and compare the results with some close strategies. The specific works as follows.ⅰCharacters of complex network and performance of part searching strategy were realized by starting from the basic network model and part searching strategy. The existing local search strategy was improved and a combing method of Max-degree and Random Walk strategy was proposed. By means of short processing time and quick advancing step from random walk, the final optimized result is received. The effectiveness of the envisaged large-scale real network data set is proved by simulation.ⅱAs for the disadvantage showed by average searching steps, the similarities and differences of several hybrid search strategy was analyzed. We found that they always choose another strategy used together with Max-degree strategy, when the current node didn’t meet the condition. But they ignore whether the next node is eligible. So we designed a improved strategy also based on the complex network search strategy combined with Max-degree and Random Walk, which added the consideration of next node and reduced the frequency of using Random Walk search strategy to optimize the average search steps. And the simulation has confirmed its effectiveness.