World's First Integration of Full-Length Optical Network Monitoring into Communication DSP Chips

Visualizing Optical Networks While Communicating

NTT Corporation (President and CEO: Akira Shimada) has successfully integrated, for the first time in the world, a capability into a communication DSP chip that visualizes the state of the entire optical network using only small optical transceivers at the receiving end. This achievement was presented as a post-deadline paper at the international conference OFC2026 held in March 2026.

Background and Challenges

With the expansion of AI demand, the importance of high-capacity and wide-area optical networks has intensified. For the IOWN APN, NTT's next-generation infrastructure, identifying abnormal loss points early is crucial for stable operation. Traditionally, dedicated measurement equipment such as OTDRs was required, making it difficult to monitor the entire network continuously while communicating, due to costs and operational burdens. Although NTT previously developed visualization techniques without measurement equipment, the massive computational resources required were a bottleneck for commercial integration.

World's First Technical Breakthrough

NTT developed proprietary technology that reduced the required computational processing for visualization by 100 times compared to conventional methods. This enabled integration into communication DSP chips, where power consumption and space are strictly constrained. The result was implemented in an 800G coherent DSP manufactured by NTT Innovative Devices Corporation, successfully identifying network abnormalities over a distance of up to 1,005 km using compact pluggable optical transceivers (OSFP). Measurement accuracy was found to be consistent with dedicated OTDR results.

Future Prospects

This technology enables optical transceivers to 'detect their own abnormalities,' bringing revolutionary efficiency to optical network operation and maintenance. NTT will continue to promote the implementation of this technology into optical networks, including the IOWN APN, to realize constant monitoring and autonomous operation of high-capacity networks for the AI era.

This research was partially supported by the National Institute of Information and Communications Technology (NICT) project (JPJ012368G60201).