The Research Of Multimedia Data Redundancy Mode Analysis System

With the more developed network and a big increase in people’s needs for it, data traffic in the network is growing rapidly, thus the redundant transmission in the network of the content of relative hot spots has become an important feature for the current delivery network. The current research for network data redundant elimination technology focuses on non-discriminatory network data, and it aims to increase the efficiency of data redundant elimination by improving the selection algorithm of redundant data chunk. The research of different types of network data redundant mode can provide some references for increasing the efficiency of data redundancy elimination, and can also offer an idea for analyzing network data from different dimensions.Based on the need of redundant mode analysis for different types of network data in the lab project "the research of DRE technology application in CDN network traffic compression", this paper designs and realizes each module for the multimedia data redundancy mode analysis system in CDN network traffic. The whole system is divided into the network data type analysis module, redundancy mode analysis module, the distributed computing environment module, and the results presentation module. The captured network data is categorized into different types through the analysis of the network data type analysis module. Then through the redundancy mode analysis module, the best redundancy elimination strategy for each type of data will be analyzed via single-mode redundant dictionary mode. Next, the redundancy elimination efficiency of the whole data will be analyzed by multi-mode redundancy dictionary, and the result is to be compared with the redundancy elimination efficiency of single-mode redundancy dictionary. At last, the analysis conclusion will be shown in the form of charts through the results presentation module.This paper firstly analyzes the current research focus of DRE technology, and secondly, it studies and summarizes the Rabin-Karp algorithm which is used to calculate the data fingerprint, the Winnowing and MAXP algorithm that are used to select the redundant chunk, and the generation and replacement mechanism of cache dictionary. Based on the existing DRE engine, this author completes the multi-redundant dictionary that enables different types of network data to adopt different redundancy elimination strategies, achieves the combination of DRE engine with the Hadoop distributed computing environment, and designs and realizes the multimedia data redundancy mode analysis system. This paper introduces the system from the aspect of technology used in the system, the system design, system processes and system modules, from the aspect of system design, including the use of technology,and the system processes and the design and implementation of system modules,I introduce this system completely,finally, it presents the analyzed results of the captured CDN network data by this system, and describes the conclusions and results.Based on the analysis need for the redundancy mode of different types network data in CDN network, this paper applies DRE technology, HTTP packet content type identification, multi-redundant dictionary mechanism and Hadoop distributed computing framework in redundancy mode analysis for different types of network data of HTTP protocol data, and completes a multimedia data redundancy mode analysis system. The system makes it possible that different strategies to eliminate redundancy are used for different types of network data, and it can be applied in the research field which analyzes network data redundancy mode according to the type of network data, thus bearing certain value in engineering.