Auto-routing algorithm for field-programmable photonic gate arrays

Area of Science:

• Photonics
• Integrated Circuits
• Computer Science

Background:

• Programmable photonic integrated circuits (PPICs) require sophisticated software for flexible operation.
• Field-programmable photonic gate arrays (FPPGAs) offer reconfigurable hardware but need efficient pathfinding algorithms.
• Fabrication variations and potential damage can impact FPPGA performance, necessitating robust routing solutions.

Purpose of the Study:

• To propose and demonstrate a modified tree-search algorithm for automatic optical path determination in FPPGAs.
• To enable optimization of optical paths based on user specifications, circuit architecture, and hardware imperfections.
• To introduce self-healing and fault-tolerance capabilities for damaged FPPGA regions.

Main Methods:

• Development of a modified tree-search algorithm tailored for FPPGA topology.
• Algorithm utilizes hardware topology and connection locations to define optical paths.
• Optimization considers multiple competing objectives: path length, accumulated loss, and power consumption.

Main Results:

• The algorithm successfully determines optimum optical paths in FPPGAs.
• Demonstrated ability to optimize paths considering length, loss, and power.
• Algorithm provides basic self-healing and fault-tolerance by finding alternative paths.

Conclusions:

• The proposed modified tree-search algorithm effectively automates optical path programming in FPPGAs.
• This approach enhances the usability and resilience of programmable photonic integrated circuits.
• The algorithm's multi-objective optimization and fault-tolerance capabilities are crucial for advanced photonic applications.