Tokenized Intelligence: Redefine Network Optimization in Softwarized Networks

Tokenized Intelligence (TI) aims to enhance network performance by combining machine learning and blockchain-based tokenomics in software-defined networks (5G/6G). This paper examines the potential for incentivizing decentralized edge computing through economic rewards, which can facilitate the efficient training of AI models. TI enables the coordination of a decentralized network of nodes, which allows for collaborative sharing of resources and decreases the expenses linked to centralized cloud-based solutions. TI is utilized to design customized logical networks that effectively optimize the use of computing, storage, and networking resources within a common physical infrastructure. This framework, which is decentralized and distributed, speeds up data analysis and reduces delays. It also provides economic rewards for edge nodes that participate in model training activities. The core of TI lies in its ability to improve network slicing strategies, such as best effort, high availability, and low latency, while also incentivizing edge nodes based on their contributions to the training process of the machine learning model. This economic incentive mechanism guarantees the efficient allocation of resources according to the specific needs of machine learning applications, thereby enhancing overall network performance.