Research on Network Traffic Prediction and Management Based on Logarithmic Barrier Method

In order to effectively describe and manage computer network traffic, this paper proposes a new method based on logarithmic barrier and queuing method (Logarithmic Barrier and Queuing Method, LBQM). This method first analyzes the key factors that affect the network traffic by using the logarithmic barrier function, and then dynamically adjusts the network distributed packet send rate by using the logarithmic barrier method and the node queue congestion situation, so that the network can obtain the maximum aggregate utility. Finally, the simulation results show that, compared network simplex algorithm and distance-directed augmenting path algorithm, LBQM has better adaptability and robustness.