Differentiated service in OBS networks using a dynamic FDL bank partitioning algorithm

Area of Science:

• Computer Science
• Telecommunications Engineering
• Network Engineering

Background:

• Optical Burst Switching (OBS) networks face challenges in providing differentiated services.
• Existing solutions may not efficiently handle varying traffic demands and quality of service requirements.

Purpose of the Study:

• To propose a novel dynamic Fiber Delay Line (FDL) bank partitioning algorithm for OBS networks.
• To address the differentiated service problem by enabling loss differentiation.

Main Methods:

• A feed-forward output buffering architecture is utilized.
• An FDL bank is dynamically partitioned into multiple groups, with each group assigned to a specific traffic class.
• Analysis considers three traffic classes and corresponding FDL groups.

Main Results:

• The proposed dynamic FDL bank partitioning algorithm facilitates effective loss differentiation in OBS networks.
• The algorithm demonstrates success in Poisson traffic environments.

Conclusions:

• Dynamic FDL bank partitioning is a viable strategy for achieving differentiated services in OBS.
• The algorithm simplifies the implementation of loss differentiation for multiple traffic classes.