Algorithmic Analysis of an Efficient Packet Scheduler for Optimizing the QoS of VoIP Networks

Abstract

Owing to the emerging realities in the Information and Communication Technology (ICT) industry, where data traffic is growing much faster than traditional voice traffic, there is now a global growing desire to migrate to digital form of communication. The trend is focused in transporting voice over data networks rather than the traditional data over voice networks. Voice traffic carried over a system originally designed for data creates technical challenges. This is primarily due to limitations that have resulted from the nature of the Internet and its bandwidth, transmission impairments and voice compression technology, which degrade voice quality. In this paper, the design of a hybrid architecture of an efficient packet scheduler for optimizing the QoS of VoIP networks have been presented. The design addresses the transmission impairment factors of delay (or latency), jitter (or delay variation) and packet loss. Mechanisms for: accommodating the demands of the expected rapid increase in the volume of voice traffic as PSTN progressively migrates to VoIP; according due precedence to the delay-sensitive voice and business/mission critical data (B/MCD) traffic flows; ensuring fair resource sharing among all traffic flows (realtime and non real-time) and adaptively maintaining optimal voice quality without over provisioning the users, have been incorporated. In the analysis, the algorithm defining the different levels of services was developed and explained. The interaction and integration of the designed functional pieces were used to develop a structured signal flowchart. The work therefore provides a theoretical framework that guarantees a graceful tradeoff between priority and fairness to all traffics flows running on the network.

Effect of Period of Day on Mobile Broadband Network Performance in a High-Mobility Suburban Environment

“…Thus, it is expected that network administrators should be able to ensure their networks meet the stringent requirements for QoS and QoE delivered to their customers. [1,2]. The provision of readily available network performance data sets at every point and time of service by administrators is therefore critical.…”