Improving Algorithm BM Capability By Optimizing Character Code

With the rapid development of Internet, people benefit from it more and more. But Internet comes into a more complex situation in security, and worm is a virus spreading through network. The destruct methods of worms are through copying themselves, transmitting in the network and using up all the system resources. Finally, worms make the whole network useless and users can not work normally. The whole world economy suffers losses because of every worm broken out, so worm's jeopardize is very serious, and the worm detecting technology becomes one of the most important research areas for our researchers in network security.Among the multitudinous algorithms of detecting worms, algorithm BM with some patterns is a better matching algorithm in the average performance. It makes use of jumping and hashing method, and uses few space to detect worm existing or not. The research scope of this paper is the technology of optimizing worm character code and improving algorithm by optimizing worm character code. This paper analyzes the capability of matching algorithm BM with several patterns by some experiment, we analyze the reasons of results gotten by experiment through some theoris. Then we introduce the eligible characteristic of worm character code. Moreover, this paper uses the statistic methods to calculate net data stream frequency from various of bytes. After studying these contents, we finds some weakness when we detect worms by algorithm BM with several patterns, and we find out improved methods based on two aspects.Finally, we validate that two above-mentioned methods are aright and logical while finishing research above.