UBITECH is excited to announce the official launch of the CoGNETs Research and Innovation Action, a cutting-edge initiative funded by the European Commission under Grant Agreement No. 101135930. The project, which began on June 1, 2024, will span three years, running through May 2027. The kick-off meeting, held online on June 21, 2024, marked the start of a pioneering effort that aims to revolutionize how intelligent systems and IoT devices operate in a decentralized and collaborative manner.
CoGNETs is set to redefine the future of intelligent information processing by leveraging swarm intelligence to create dynamic, self-organizing IoT-to-Cloud networks. These networks will be capable of autonomously managing AI-driven tasks through innovative methods of cognitive computing. By empowering devices to recognize their own heterogeneity and make autonomous decisions, CoGNETs will ensure secure, energy-efficient, and automated processing across a wide range of domains.
At its core, CoGNETs will develop a unique Middleware architecture that incorporates decentralized, federated systems enhanced by Game-intelligent Agents. These agents will enable seamless collaboration between devices, using advanced security mechanisms like RISC-V hardware security and AI acceleration. The open-source programming models and APIs integrated into this framework will allow for highly optimized orchestration and resource management, free from vendor lock-in constraints.
UBITECH plays a pivotal role in the CoGNETs project by designing and implementing innovative Agent systems that will empower autonomous decision-making at the device level. The focus of UBITECH’s efforts includes the following:
Game-driven Pricing and Auctioning Systems: UBITECH will lead the development of new game strategies that enable devices to autonomously engage in “Pricing,” “Bidding,” and “Auctioning” processes through intelligent agents. This includes the formulation of competitive bidding frameworks, integrating diverse models such as Kelly’s, Richman’s, Poorman’s, and Taxman’s biddings. These dynamic auctioning strategies will optimize resource management at scale.
Optimization through Decentralized Game Strategies: To enhance network performance, UBITECH will develop decentralized optimization strategies based on multi-player game theory. By applying stochastic, asymmetric competitive game models—such as Blotto-type and Stackelberg-type games—UBITECH aims to resolve complex optimization challenges efficiently. The strategies will factor in the heterogeneity of devices and data, ensuring low-complexity, high-efficiency decision-making.
Agent-based Systems Development: UBITECH will translate the developed game strategies into programmable agent functions using the JAVA Agent Development (JADE) framework. This will enable the creation of advanced Java-based Agent systems, compliant with the IEEE FIBA Multi-Agent Systems (MAS) standard. These systems will empower swarm devices with autonomous, game-driven decision-making capabilities for optimized task and resource management across IoT-to-Cloud environments.
CoGNETs represents a significant step forward in the convergence of AI, IoT, and cognitive computing. By fostering an environment where devices can self-organize, collaborate, and autonomously optimize their resources, the project will open new possibilities for industries ranging from smart cities and healthcare to manufacturing and beyond.
