Development of statistical traffic models and their application to information security

点击翻译按钮获取中文摘要

The project developed statistical models for network traffic analysis with bot hidden Markov models (HMMs) and Markovian arrival processes (MAPs) by revealing the mathematical relationship between HMMs and MAPs. We also considered parallelization of the parameter estimation algorithm for the proposed HMM-MAP-based models to handle a long-term observation for network packet arrivals. Moreover, we proposed the applicable example of the developed parameter estimation algorithm towards the information security.