An integrated regulation and scheduling scheme for real-time traffic management

Typical rate-based traffic management schemes for real-time applications attempt to allocate resources by controlling the packet delivery to the resource arbitrator (scheduler). This control is typically based only on the characteristics of the particular (tagged) traffic stream and would fail to optimally adjust to non-nominal network conditions such as overload. An integrated regulation and scheduling (dynamic-R&S) scheme is proposed whose regulation function is modulated by both the tagged stream's characteristics and some information capturing the state of the co-existing applications as provided by the scheduler. The performance of the proposed scheme-as well as that of the equivalent static one-is investigated for both under-load and overload traffic conditions-analytically and through simulation-and the substantially better throughput/jitter characteristics of the dynamic-R&S scheme are established.