A model reduction method for traffic described by MMPP with unknown rate limit

The main hindrance to apply Markovian modulated Poisson process to network traffic modeling is the state-space explosion problem in traffic aggregation. The existing rate limit algorithm cannot be directly used to reduce the model of the aggregated traffic and capture the most represented time constants (RTCs) of the traffic. In this letter, we extend the rate limit algorithm to the model reduction in terms of RTCs with unknown rate limit. We obtain a more concise form of the algorithm than the existing one. The effectiveness of the proposed method is demonstrated via a numerical example.