Publication

Abstract : VoIP, or voice over internet protocol, has gained popularity as a real-time communication technology because it is flexible and reasonably priced. However, maintaining a good level of quality of service over a variety of dynamic network conditions is necessary to guarantee a satisfactory user experience for VoIP applications. To optimize voice traffic delivery and guarantee a reliable and excellent communication experience, adaptive QoS dynamically modifies network parameters. This paper introduces the concept of Adaptive QoS, a proactive strategy leveraging Call Sequencing, Machine Learning Regression Analysis, and two pivotal algorithms-Silence Suppression and Adaptive Jitter. The Silence Suppression Algorithm efficiently manages silent periods, optimizing network resources, while the Adaptive Jitter Algorithm mitigates jitter impact for smoother voice transmission. Through comprehensive assessments of latency, jitter, bandwidth, packet loss, and network stability, approach dynamically adjusts network parameters. Results indicate that the Adaptive Jitter Buffer has a latency of 0.02 milliseconds, while Silence Suppression (CNG) exhibits a significantly lower latency of 0.001 milliseconds. In terms of jitter, the Adaptive Jitter Buffer introduces a jitter value of 0.0639375 milliseconds, whereas Silence Suppression achieves a much lower jitter of 0.001 milliseconds. These findings demonstrate that Silence Suppression outperforms the Adaptive Jitter Buffer in both latency and jitter, contributing to a smoother and more reliable real-time communication experience.