Executive Summary : | The evolution of the particularly complex ecosystem of 5G mobile radio communications towards 6G will require, besides the allocation of new spectrum bands, smart, dynamic, distributed, and safe spectrum management schemes in order to accommodate the growing service requirements of the enhanced mobile broadband (eMBB), massive machine-type communication (mMTC), and ultra-reliable and low-latency communication (URLLC) domains. Such networks will have to manage effectively the use of licensed, shared, and unlicensed spectrum in the FR1 (sub-6 GHz) and FR2 (millimeter-wave) frequency regions to improve the overall spectral efficiency (SE), satisfy the requirements of the corresponding heterogeneous services, and ensure spectrum access security while staying energy efficient. In summary, the licensed spectrum in the sub-6 GHz segment constitutes an expensive and scarce resource that is often either inaccessible or insufficient for certain players/services, while at the same time underused by others. The FR2 spectrum, on the other hand, promises substantial amounts of available bandwidth. However, the FR2 spectrum is not a panacea: some of the considered frequency bands are already utilized to some extent by other applications, whereas others are suitable only for very short-range applications. Given these and the lessons learned from the FR2 experience, today’s wireless community seems more open to designing spectrally efficient mmWave networks upfront and promoting more flexible spectrum usage paradigms. To this end, this focused research will explore recent advances in artificial intelligence (AI), such as advanced convolutional neural network (CNN) methods based on long short-term memory (LSTM), and generative adversarial networks (GAN) as well as deep reinforcement learning (DRL) towards in-depth spectrum discovery and AI-based communication and resource allocation in a real-time radio-aware spectrum sharing framework. In addition, blockchain technology will be explored as a distributed database to verify and secure spectrum-sharing contracts between mobile nodes. The demonstration will target key automotive and industrial Internet-of-Things (IoT) scenarios involving eMBB and URLLC, while a system-level simulator will be built for the large-scale evaluation of the developed techniques. These will also help avoid congestion in the foreseeable future and achieve a faster return of investment (RoI) given the higher cost of equipment and additional spectrum in the FR1 and FR2 bands. This focused research aspires to be one of the first concerted initiatives in India that will elevate spectrum sharing to the next level, coupling it with relevant emerging breakthroughs in the field of wireless communications using AI and blockchain technologies. In doing so, the project's outcomes will lead to the development of new telecom industry products and services designed in India and licensed internationally. |