VoIP Quality of Service (QoS) Testing

Network switching equipment vendors have become increasingly concerned with Quality of Service (QoS) issues. This is in response to the introduction of different types of network applications. Voice over IP (VoIP), for example, requires a fundamentally different approach to the way in which data is delivered. Historically, IP was designed to deliver data on a "Best Effort" basis; whereas, VoIP must deliver its data (voice) on a more reliable, low-latency basis. Supporting both types of services requires that a level of intelligence be built into the network since each packet (or data flow) must be handled differently, depending on the type of service being provisioned. With this in mind, QoS protocols have been designed to operate within the domain of current network protocols and to provide for a degree of control over latency and delivery issues that was not previously necessary. Though these protocols do not, by themselves, increase bandwidth, they do provide for the intelligent allocation of bandwidth such that latency-sensitive traffic can co-exist with "Best Effort" services in an economically feasible solution.

Consequently, as network equipment vendors incorporate such QoS protocols, there exists a need to qualify their equipment before and during deployment. This is especially important because the Internet is made up of equipment from many different vendors, each of which will incorporate its own proprietary QoS algorithms. This white paper focuses on the issue of testing and verifying network equipment in light of QoS issues. Specifically, it will describe testing algorithms that model mixed VoIP and "Best Effort" traffic under worst-case stress conditions.