2nd Symposium on Networking and Artificial Intelligence (SNAI 2025)
Oct 24 - Oct 26, 2025
Singapore Management University, Singapore
About the Conference
The 2nd Symposium on Networking and Artificial Intelligence (SNAI 2025) was held in Singapore. The second edition grew substantially over the inaugural one, adding tracks on programmable networks and data-plane machine learning, and on generative AI and foundation models for networking. It also introduced an author rebuttal phase and artifact evaluation.
SNAI 2025 received 143 submissions and accepted 38 papers, an acceptance rate of 26.6%, with authors from 28 countries and roughly 260 attendees.
This edition is closed and archived. Please see SNAI 2026 for the current call for papers.
Submission Guidelines
Papers must be submitted as PDF in the two-column IEEE conference format. Full papers are limited to 8 pages of technical content plus unlimited pages of references; short papers are limited to 4 pages plus references.
SNAI uses double-blind review. Submissions must not reveal author names or affiliations, and self-citations must be phrased in the third person. Papers that are not properly anonymized may be desk-rejected.
Submitted work must be original and must not be under review at another venue at the time of submission. Authors are encouraged to submit artifacts (code, datasets, measurement traces) to support reproducibility.
Review Guidelines
Evaluate each submission on originality, technical soundness, experimental rigour, clarity of presentation, and relevance to the networking/AI intersection that defines SNAI.
Every paper receives at least three independent reviews. Please justify your score with specific, actionable comments the authors can respond to, and flag any reproducibility or ethical concerns in the confidential comments to the chairs.
Do not attempt to deanonymize authors, and declare any conflict of interest to the program chairs immediately.
Tracks
AI-Driven Network Management and Automation
Self-driving networks, intent-based networking, and closed-loop control informed by learned models.
Edge and In-Network Intelligence
Inference at the edge and inside the network, model partitioning, and latency-aware serving.
Distributed and Federated Learning over Networks
Communication-efficient and privacy-preserving training across bandwidth-constrained links.
Machine Learning for Network Security
Detection, attribution, and defence, including robustness against adaptive adversaries.
Programmable Networks and Data-Plane ML
P4 and SmartNIC-based inference, line-rate feature extraction, and hardware-aware model design.
Generative AI and Foundation Models for Networking
Large models applied to configuration synthesis, troubleshooting, and network telemetry summarization.
Important Dates
Past Edition
This edition has concluded and is kept here as a permanent record. Its program, proceedings and committees remain available on the conference website.
Past EditionGet CFP Updates
Subscribe to receive email updates about this call for papers (deadlines, announcements).