Bayesian Network Worm Detection Method Based on Time Model

The network worms have been a severe threat to the Internet with their occurring frequently, and also has brought about huge economic losses to the society. According to analysis the network worm characteristic, a Bayesian network worm detection method based on time model is put forward. This technology treats the failure rate of destination unreachable as the worm detection target. It determines whether the host is infected with worms by calculating the probability of failure. In the meantime, the time interval of connection request is considered into it, and the piecewise-function of time is introduced, so that the test results can be more accurate. Experiments show that the improved time-based model detection technology can effectively improve the network worm detection rate.