A technique is disclosed that enables latency-tolerant and latency-intolerant applications to intelligently share a shared-communications channel in a manner that seeks to satisfy the needs of all of the applications. In particular, the illustrative embodiment enables each application to be associated with a different class of service, wherein each class of service is associated with one or more quality-of-service parameters (e.g., minimum throughput, maximum latency, etc.). The illustrative embodiment then effectively apportions access to the shared-communications channel by regulating different degrees of bursting (i.e., the transmission of multiple frames at a single transmission opportunity) based on the class of service associated with the application.
展开▼
