Fast end-to-end available bandwidth estimation algorithm in BcN environment

Abstract

We present an end-to-end available bandwidth estimation algorithm using self-induced congestion approach, one-way delay and measured traffic load. Broadband Convergence Network(BcN) is an integrated realization of Next Generation Network(NGN) where one can make use of quality guaranteed multimedia service such as broadcasting, telecommunication and internet at any time, any place. When customers are connected with each other through the best-effort network as well as BcN, it is hard to guarantee Quality of Service (QoS) over the best-effort network. Therefore, it is necessary to estimate accurate end-to-end available bandwidth to support end-to-end QoS. The available bandwidth measurement has been an important part in traffic management. Knowing the amount of available bandwidth is useful for many purposes. For example, end-to-end admission control, server selection, optimal route selection, congestion control and service level agreement verification are performed based on the result of available bandwidth estimation. To address this issue, we use measurement algorithm, which are performed on the terminal equipment, and delay model. In addition, the concept of the proportional share of link capacity is used for estimating end-to-end available bandwidth. For evaluating our method in terms of measurement latency and accuracy, we compare simulation results of our proposed algorithm with the one of conventional method.