Queuing Delay Reduction based on Network Traffic Patterns: A Predictive QoS Framework For Point-To-Point Communications

Ensuring optimal quality of service (QoS) in computer networks requires a detailed assessment of performance metrics, with data network queuing delay within intermediate devices being critical parameters. This paper presents a predictive Quality of Service (QoS) model designed to reduce queuing delays by analyzing traffic patterns in intermediate devices in point-to-point network connections. The proposed novel Length Packet Queuing (LPQ) model leverages packet length analysis to predict and manage queuing delays without relying on traditional packet marking mechanisms. Through Poisson distribution and polynomial regression models, network traffic patterns and queuing delays are estimated, respectively, demonstrating significant improvements of conventional QoS models. Simulations and experimental scenarios validated the LPQ model’s effectiveness, showing lower delays through various network loads and traffic conditions. The results of this research highlight the potential of the novel LPQ model for enhancing QoS in hybrid networks, where user applications generate diverse packets.