摘要
Current worm detection methods are unable to detect multi-vector polymorphic worms effectively. Based on negative selection mechanism of the immune system, a local network worm detection system that detects worms was proposed. Normal network service requests were represented by self-strings, and the detection system used self-strings to monitor the network for anomaly. According to the properties of worm propagation, a control center correlated the anomalies detected in the form of binary trees to ensure the accuracy of worm detection. Experiments show the system to be effective in detecting the traditional as well as multi-vector polymorphic worms.
Current worm detection methods are unable to detect multi-vector polymorphic worms effectively. Based on negative selection mechanism of the immune system, a local network worm detection system that detects worms was proposed. Normal network service requests were represented by self-strings, and the detection system used self-strings to monitor the network for anomaly. According to the properties of worm propagation, a control center correlated the anomalies detected in the form of binary trees to ensure the accuracy of worm detection. Experiments show the system to be effective in detecting the traditional as well as multi-vector polymorphic worms.
基金
the National"863"Program Project (2002AA141090 ,2004AA147070)
