The IEEE Signal Processing Society is inviting submissions for a forthcoming special issue of the IEEE Journal of Selected Topics in Signal Processing (JSTSP), focusing on what it describes as a growing field: Autonomous and Evolutive Optimization in Networked AI.
According to the call for papers, the topic “represents a transformative paradigm for Signal Processing and Artificial Intelligence (AI) communities”, combining traditional signal processing techniques with modern deep learning approaches.
The organizers say the approach integrates “traditional knowledge-based adaptive signal processing techniques and data-centric deep-neural network models”, enabling systems to “dynamically acquire high-quality data in the continuous inferences of networked AI models”.
The concept centers on networked AI systems capable of self-optimization through adaptive feedback mechanisms. These systems can “optimize every individual model by adaptively generating corresponding rewards and pseudo-labels online”, a process the organizers say mirrors how complex organizations evolve over time.
The call for papers also highlights the potential to unify different machine learning paradigms, noting that such systems “can unify supervised and reinforcement learning in the networking systems of AI, by the adaptive signal processing”.
A key focus is on multi-agent systems, where distributed AI models interact dynamically. These interactions are said to enable “autonomous self-optimization and evolution of networked AI, ensuring robust performance in time-varying environments without human interventions”.
The scope of the special issue spans multiple disciplines, including signal processing, communications, and industrial automation. Suggested application areas include large language models, autonomous driving systems, and real-time 3D reconstruction.
The organizers state that the issue aims “to consolidate and expand the foundational principles of the adaptive and online optimization for networked AI models, and foster its advancements in intelligent signal processing systems”.
Submissions are open until June 15, 2026, with publication scheduled for January 2027. The special issue is being led by Liang Song of Fudan University, alongside guest editors from institutions in Canada, Israel, Greece, and China.
Further details, including submission guidelines, are available via the IEEE Signal Processing Society.
