The Approach Of Message Broadcast And Large-scale Data Transfering For Distributed Content Audit System

For distributed network content audit system, there are two important aspects in its communication, the first is the issuing of the audit rules, and the second is the data transferring between audit nodes. The above two problems involves about message broadcast and large-scale date transferring. In response to solve the two problems, this paper presents a thorough study of the message broadcast and large-scale date transferring, and gives our research achievement.Regarding the broadcasting algorithm, this paper first analyzes the existing broadcasting algorithm and their advantages and disadvantages are summarized and compared, then, we introduce our broadcasting algorithm: the APDG (Advanced Ferfect Difference Graph) broadcasting algorithm on the basis of the established broadcasting algorithm is proposed in this paper. This broadcasting algorithm will determine a rebroadcast node from its backward nodes, then, the rebroadcast node will broadcast the message to the rest of its backward nodes. This improvement can reduce the performance of the source node, and increase broadcasting efficiency. In addition, the APDG broadcasting algorithm can construct the topology of the distributed content audit system. The experimental results show that the APDG has better performance than existing broadcasting algorithm.For large-scale data transferring, splitting a large-scale data into multiple data block and transferring by multiple paths are primary ways. This paper analyzes the main methods of large-scale data transferring and summarizes their defect. Relying on the task allocation algorithm in grid, we give our large-scale data transferring algorithm: the large-scale data transferring algorithm based on the performance of relay node. This algorithm will find relay nodes by APDG broadcasting algorithm, then, it will split the large-scale data into multiple parts according to the number of relay nodes and the performance of the relay nodes. Experimental results show that our large-scale data transferring has preferable flexibility and transmission performance.