Iterative available bandwidth estimation for mobile transport networks

Abstract

Available bandwidth estimation has lately been proposed to be used for end-to-end resource management in existing and emerging mobile communication systems, whose transport networks could end up being the bottleneck rather than the air interface. Algorithms for admission control, handover and adaptive QoS management could take into account such estimations, which may lead to an overall system performance improvement. This paper introduces two new iterative algorithms (Zoom and Adaptive pathChirp) for network load monitoring that can be combined with a linear least squares fitting achieving reliable estimations without causing congestion collapse. Their performance is evaluated in a simulated UTRAN Long Term Evolution backhaul.