Dynamic network control with QoS and resource priority monitor based on active "eM" for commercial VoIP

• Main Authors: Grah, M., Radcliffe, P.
• Format: Conference Proceeding
• Language: English
• Subjects: Codecs; Control system synthesis; Control systems; Costs; Current measurement; Equations; Monitoring; QoS; Quality management; Quality of service; Telecommunication traffic; VoIP
• DOI: 10.1109/ICIT.2009.4939740

Reduction in the cost of the network infrastructure will always be a relevant and commercially important topic. There is a balance between cost and Quality of Service (QoS) that has been traditionally managed in a static manner. This is no longer an ideal situation as network traffic and configuration can change in a dynamic manner. This paper examines methods by which VoIP can be managed using a dynamic method to balance cost and quality. In VoIP systems, quality has a real commercial value as clients will leave a service provider that does not deliver an adequate quality of service. Nearly as important is the network monitoring and prediction function which can tell a carrier of impending problems before they become an annoyance to paying clients. We propose to design a new dynamic network control system that is based on a superior measurement of QoS with the VoIP-eM equation [1] and the CODEC selection and priority adjustment model. The new system will monitor the quality with an active indicator of the quality, one we propose to call ldquoActive -eMrdquo value, and then actively control the priority levels based on the measured ldquoeMrdquo value. The inputs to the QoS equations can be easily measured in a real network. The system will also incorporate the network reporting mechanisms. Both models, CODEC selection (incorporating CODEC G.711, G.729, AMR and GSM-FR) and priority allocation model, are based on the output of the measured quality metrics, and indicated by the total eMvalue that then serves as an output factor and reports the measured metrics that depicts the current status of the network. The system will report back the total quality score in a circular fashion, where eM-value acts as an output and it is then fed back into the system to act as an input. The dynamic estimation of the quality will furthermore use a new eM-score system that is a modification of the current MOS score scale. We propose 10 levels of eM-MOS scores and consequently 10 levels of dynamic control. Our model will be validated and integrated into a middleware-based commercial solution using CORBA system framework for Enterprise management in Heterogeneous VoIP networks.