A Random Forest-based Operating System Recognition Algorithm for Network Security

With the development and popularization of the Internet, network security has become the focus of public attention. Operating system detection and identification is an important part of network security work, which has very important significance for its research. This thesis designs an algorithm of operating system identification based on random forest. Based on the third-party fingerprint database, this method constructs simulation data and maps it into a vector that can be learned by the algorithm. In order to construct a highly reliable and generalized resource fingerprint structure, a random forest algorithm is used to combine multiple weak fingerprints into Strong fingerprints to improve the accuracy of operating system recognition, and design a layered training architecture to achieve long-term expansion and maintenance of operating system fingerprint models based on random forest construction. The fingerprint similarity algorithm is adopted to measure the similarity between fingerprints, which provides a basis for the operating system to mark the card. Experiments show that this method can accurately identify the remote operating system, and has a better accuracy effect than the traditional static matching method.