Goals

The project focuses on enhancing the performance of Cisco's proprietary Path Computation Element (PCE) algorithm, which is responsible for computing optimal network paths and enabling multiplexing of optical paths across complex network topologies. By integrating advanced AI techniques, including deep reinforcement learning and machine learning, the initiative aims to augment the existing PCE logic to more effectively determine feasible network paths in challenging scenarios. These scenarios involve managing conflicting demands arising from limited network resources, diverse service requirements, and stringent routing constraints—conditions that often exceed the capabilities of default path placement algorithms. The enhanced PCE will enable more intelligent, adaptive, and efficient path computation to optimize network utilization and service quality.