Overview
Future softwarized networks are evolving toward highly decentralized and multi-domain operation. While Software-Defined Networking (SDN), Network Functions Virtualization (NFV), and cloud-native platforms have enabled programmability, centralized orchestration is increasingly limited in terms of scalability, responsiveness, resilience, and trust.
ADeCoS 2026 aims to shape the next generation of AI-native and decentralized network control, where agentic AI systems autonomously coordinate decisions across domains, stakeholders, and layers of the network. These systems go beyond traditional automation by incorporating goal-driven behavior, adaptive communication, and context-aware decision-making.
The workshop focuses on fundamental and practical challenges in enabling such systems, including multi-agent coordination, intent-to-action translation, learning and planning under uncertainty, and policy-aware control with built-in safety and verification mechanisms.
Despite recent advances in AI-driven network management, scalable and trustworthy decentralized coordination across domains remains largely unresolved. ADeCoS 2026 aims to address this gap by bringing together researchers and practitioners from networking, distributed systems, and AI.
The workshop emphasizes realistic deployments, experimental validation, and reproducibility, fostering new directions toward trustworthy, scalable, and autonomous network infrastructures.
Topics of interest
Non-exhaustive list
- Agentic architectures for SDN/NFV control
- Intent-to-action and goal decomposition
- Planning and decision-making under uncertainty
- Multi-agent coordination, negotiation, and communication
- Decentralized and federated control planes
- Network slicing and multi-domain orchestration
- LLM-assisted reasoning and decision-making
- Policy compliance, verification, and safety
- Trust, governance, and security
- Deterministic networking (TSN / DetNet)
- Energy-aware and sustainable orchestration
- Benchmarks, datasets, and experimental testbeds
Call for Papers
We invite original and unpublished contributions on agentic AI and decentralized coordination for softwarized networks. Contributions may include theoretical advances, system designs, algorithms, prototypes, and experimental evaluations. We particularly encourage submissions demonstrating practical relevance and real-world applicability.
Submission guidelines
- Format: IEEE two-column conference format
- Length: Up to 7 pages (including references)
- Submission: Submit via EDAS
- Publication: IEEE NetSoft 2026 Workshop Proceedings (IEEE Xplore, subject to policy)
Review process
- Each submission will receive at least 3 independent reviews
- Evaluation criteria include originality, technical quality, clarity, and relevance
Important dates
How to submit
- Prepare your manuscript in IEEE format (maximum 7 pages)
- Submit via EDAS: https://edas.info/N34949
- Select the workshop: ADeCoS 2026
Program
The workshop program will be announced soon.
Organizing Committee
- Dr. Rituparna Datta (Cognizant) — rituparna.datta@cognizant.com
- Dr. Nakjung Choi (Nokia Bell Labs) — nakjung.choi@nokia-bell-labs.com
- Dr. Fetia Bannour (ensiie, SAMOVAR – Télécom SudParis) — fetia.bannour@telecom-sudparis.eu
- Dr. Risto Miikkulainen (Cognizant / UT Austin) — risto@cognizant.com
- Dr. Praveen Tanguturi (Cognizant) — praveen.tanguturi@cognizant.com
- Dr. Kurdman Rasol (CEIT) — kurdman.rasol@ceit.es
- Dr. Chrysa Papagianni (University of Amsterdam) — c.papagianni@uva.nl
Technical Program Committee
To be announced