A Novel Data Burst Generation Algorithm Using Hysteresis Characterstics

OBS technology, which is being actively researched as a potential option for the future backbone of the optical Internet, is already in use. This is because OBS can solve the optical buffer issue at switching nodes without the need for optical, electrical, and optical conversion. Additionally, OBS provides various classes of service without any buffering.However, implementing an OBS network comes with several challenging issues that need to be addressed. These include concerns related to the edge router, managing burst offset times, and devising a robust burst assembly mechanism. Additionally, the core router necessitates solutions for tasks such as scheduling data bursts and control header packets, establishing protection and restoration mechanisms and resolving contention. This study primarily focuses on the burst assembly mechanism aspect of OBS. In this context, a novel algorithm for generating data bursts is proposed. The hysteresis characteristics of the queuing model at the entrance edge node of optical burst switching networks are used by this approach. Simulations that include Poisson and self-similar traffic patterns show how the method may adjust data burst sizes adaptively in response to the load that is being applied. This adaptability results in a notably high average utilization of data bursts while minimizing the need for frequent timer operations. Moreover, the algorithm proficiently addresses the challenge of persistent blockage problems in bandwidth reservation appeals, minimizes peak queuing delays, and diminishes the required burst magnitude. This collective effect leads to an improved utilization of data bursts for incoming IP traffic characterized by bursts.