Algorithms for the design of WDM translucent optical networks

Area of Science:

• Telecommunications Engineering
• Computer Networking
• Optical Communications

Background:

• Optical fiber transmission faces impairments over long distances, necessitating regeneration in translucent optical networks.
• Translucent optical networks utilize spare transceivers in intermediate nodes for signal regeneration.
• Maximizing the number of established connections (NEC) is a critical challenge in these networks.

Purpose of the Study:

• To address the computational intractability of mixed-integer linear programming (MILP) for large-scale translucent optical networks.
• To develop an effective heuristic algorithm for maximizing network connections.
• To validate the heuristic algorithm's performance against MILP for smaller networks.

Main Methods:

• Formulation of the NEC problem as a mixed-integer linear programming (MILP) problem for small networks.
• Development of a heuristic algorithm based on K-least-wavelength-weight-path routing and statistical methods for large networks.
• Comparative analysis of the heuristic algorithm against MILP for performance evaluation.

Main Results:

• The proposed heuristic algorithm achieves optimal or near-optimal solutions in most cases, even for small networks solvable by MILP.
• The heuristic algorithm demonstrates scalability and effectiveness for large networks and large traffic matrices where MILP is intractable.
• The algorithm successfully maximizes the number of established connections in translucent optical networks.

Conclusions:

• The heuristic algorithm offers a computationally efficient and scalable solution for maximizing connections in translucent optical networks.
• This algorithm is suitable for practical implementation in route engines and network design tools for large-scale optical networks.
• The study highlights the viability of heuristic approaches for complex network optimization problems.