Signal and Graph Processing for Autonomous Agents
Siwei Zhang (email@example.com), German Aerospace Center, Germany
Anna Guerra (firstname.lastname@example.org), National Laboratory of Wireless Communications (WiLAB), CNIT, Italy
Francesco Guidi (email@example.com), National Research Council of Italy, Italy
Davide Dardari (firstname.lastname@example.org), University of Bologna, Italy
Petar Djurić (email@example.com), Stony Brook University, USA
Autonomous agents that are connected through a network have become more popular in applications that involve sensing, positioning, and exploration. When working together, these agents improve their awareness of their surroundings and greatly enhance communication, navigation, sensing, learning, and decision-making capabilities of the system. The agents are required to accomplish these sequence of inter-connected tasks and eventually making decisions on next actions in time. A joint investigation of these tasks, emphasizing the cooperation among agents and interaction among tasks, is an important and timely relevant research topic. Both fundamental understanding and efficient realization of the cooperative, and often distributed, signal processing and machine learning in such networks are of a great interest to the community.
This special session is motivated by the great success of a special session organized for ICASSP’23 in Rhodes. For ICASSP’24, we organize a follow-up special session with an advancement of the focus on the potential of applying graph signal processing techniques for autonomous agents. We aim at bringing together researchers to discuss the possibility of empowering autonomous agents by advanced signal and graph processing techniques. It is a self-contained topic which interacts with many topics of interest at ICASSP, e.g., sensing, learning, array processing, communications and navigation. However, the multi-agent autonomy emphasis and the interdisciplinary flavor make it well distinguished from the main tracks of ICASSP.
Topics of interest:
- Fundamentals of multi-agent signal processing and machine learning;
- Signal and graph processing for autonomous networks;
- Integrated multi-agent communication, sensing, navigation, control and path planning;
- Decentralized estimation and optimization;
- Advanced machine learning tools for individual and team decision making;
- Hybrid data- and model-based multi-agent learning;
- Multi-agent signal processing and learning techniques for specific applications, like exploration and UAVs;
- Signal Processing for Future Smart Radio Environments.
We thank our Senior Advisors, Davide Dardari and Petar Djuric, for their continuous support in shaping the scope of the session. In addition, we are honored to have the Autonomous Systems Initiative (ASI) of IEEE SPS as a technical co-sponsor.