Intelligent offset time algorithm for optical burst switching networks

Abstract

Optical Burst Switching (OBS) is an optical network that is efficient and reliable. However, there are some critical issues in OBS such as burst loss ratio, which occur due to contending bursts as optical buffers are lacked at the core node. Many techniques have been proposed to tackle the problem such as using offset time. Offset time is managed to allow Burst Header Packet (BHP) to reserve the required resources necessary for a successful transmission process. If the added extra time is long, it reserves the resources for a long time. In contrast, setting low offset time results in burst loss. As a consequence, the techniques that manages offset time need to be developed. This paper proposes an Intelligent Offset Time (IOT) algorithm that adapts offset time based on the condition of the network and the traffic. Three parameters are used as IOT’s fuzzy input that are burst size, destination and burst queuing delay. The proposed algorithm is evaluated against the conventional offset time and adaptive offset time algorithms. The results of IOT’s simulation show that it has effectively reduced burst loss ratio and end-to-end delay in comparison with the conventional and adaptive algorithms.