Signaling Protocol for Session-Aware Popularity-Based Resource Allocation

The Differentiated Services model (DS) maps traffic into services that offer different quality levels. However, flows are treated unfairly in each service, since the DS model lacks a policy to distribute service bandwidth between flows that form the service aggregate traffic. We present a signaling protocol called Session-Aware Popularity-based Resource Allocation (SAPRA) that fairly distributes resources along the path of each session, with an amount of resources proportional to their receiver population. We assume that scalable sessions are layered hierarchically, with each layer sent to a different multicast group. We evaluate the efficiency of SAPRA using theoretical analysis and simulation. 1