The 2026 International Conference on Networking and Network Applications (NaNA 2026) was successfully held in Macau, China, from August 13 to 16. Hosted by City University of Macau and co‑sponsored by multiple domestic and overseas universities, the event gathered global scholars in network communications, cloud computing, artificial intelligence and cybersecurity.


The opening ceremony took place on the morning of August 14. Professor Norio Shiratori first delivered the welcome speech, extending sincere greetings to participants worldwide. Professor Wanlei Zhou, Vice Rector of City University of Macau, then gave the opening remarks.



Following the opening ceremony, keynote speeches were then delivered by three scholars.
Professor Achille Pattavina from Politecnico di Milano, Italy, presented the keynote entitled Evolution of Data Center Interconnects from Cloudscale Networking to AI Hyperscale Computing. He reviewed the technical evolution of data‑center optical interconnects over the past decade, analyzed how silicon photonics and optical transceivers support bandwidth expansion for modern data centers, and described massive bandwidth demands brought by AI hyperscale computing clusters. He also explored the optimization solutions for GPU/TPU cluster interconnections and geographically‑distributed data‑center collaboration.



Professor Hiroshi Inamura from Future University Hakodate, Japan, delivered a speech titled Availability Challenges in Large‑Scale Network Systems. Taking real‑world large‑scale service outages such as GitHub incidents as examples, he analyzed infrastructure overload risks caused by the explosive growth of AI agents and massive IoT devices. He pointed out that traditional tightly‑coupled network architectures are vulnerable to cascading failures and synchronization anomalies, and proposed availability‑oriented design principles including architecture decoupling and overload‑aware component isolation, offering new insights for reliability improvement of next‑generation IoT and distributed network systems.



Professor Wanlei Zhou gave the keynote titled Compatibility between Ethics and Security for Large Language Models. He systematically analyzed the risks of generative AI including data leakage, algorithmic bias and harmful content generation. He presented a self‑developed LLM evaluation framework integrating ethical constraints and security requirements, and put forward practical mechanisms for dynamically balancing security safeguards and ethical norms, providing feasible solutions for trustworthy and compliant large‑model systems.


In the afternoon of August 14, the Best Paper Candidate Session was held. Authors of ten shortlisted papers presented their latest research and carried out in‑depth interactive academic discussions.


The Best Paper Award ceremony was held on the evening of August 14. Through comprehensive and strict evaluation by the Technical Program Committee from dimensions of paper innovation, research quality, presentation performance and academic value, three outstanding papers were selected from ten finalists for the Best Paper Award.






Parallel sessions ran from August 15 to 16, covering six core technical tracks: Wireless Networks, Wired Networks and Data Center (DC) Networks, Network Management, Monitoring and Automation, Network Applications, AI/ML in Networks, and Industrial Control Networks Security. Participants shared cutting‑edge research progress and promoted in‑depth academic interactions and intellectual exchanges across borders and disciplines.


The TPC Meeting was held on August 16 from 14:30 to 16:30. Committee members had thorough discussions on paper review process, future conference themes and international collaboration strategies, laying a solid foundation for the sound and sustainable development of NaNA conference series.