ECE PhD Proposal Review: Lorenzo Bertizzolo

PhD Proposal Review: Software-Defined Wireless Networking for 5G and Beyond: From Indoor Cells to Distributed Aerial Swarms

Lorenzo Bertizzolo

Location: MS Teams Link

Abstract: While Software-defined Networking is a consolidated and widely adopted concept in fixed infrastructure, its adoption to the wireless domain has been limited by some fundamental challenges. Different from wired deployments, wireless stacks are characterized by tight inter-dependencies among their protocol stack layers (known as vertical coupling) and the nodes sharing the wireless channel (horizontal coupling). These effects combined undermine the implementation of i) Control plane / data plane separation, and ii) Control of multiple data planes from a separate controller; the two founding principles of SDN. Recent developments in spectrum access technology, however, made it possible to reconfigure multiple layers of the wireless stack at once. This paved the way for the development of cross-layer algorithms toward the implementation of control plane / data plane separation for wireless. Moreover, cross-layer algorithms can be employed together with distributed control theory to implement distributed and scalable control for wireless networks, this way overcoming the difficulties of implementing separate control for multiple data planes.
This proposal exploits full-stack programmability to propose, design, and implement new cross-layer algorithms that reconfigure the wireless stack at multiple layers and in real-time. Through the systematic use of cross-layer optimization, closed-loop control, and dynamic network adaptation, this proposal contributes to the development of a wide range of technological innovations for spectrum access, to bring the benefits of Software-defined networking to the wireless domain. We present a closed-loop PHY/MAC cross-layer control algorithm to enable spectrally-efficient OFDM spectrum access in Wi-Fi populated bands. Then, we exploit the technological innovations of a 5G Open-RAN infrastructure and propose a control system that enables broadband 5G connectivity for aerial cellular users that dynamically adapts to the changing network conditions like the time-changing distribution of pedestrian users in the surrounding. At millimeter-wave frequencies, we propose a cross-domain control algorithm that reduces the initial access latency in standalone high-frequency systems and obtains higher spectral efficiency for aerial links. Finally, we empower the SDN paradigm to bring network management to distributed aerial swarms. Through full-stack software programmability and programmable motion control, we implement scalable wireless network management for distributed aerial swarms.
We conclude the proposal with an overview of the requirements and design principles for next-generation wireless testing platforms to support software-programmable spectrum access.