Doctoral theses of the School of Electrical Engineering at Aaltodoc (external link)
Doctoral theses of the School of Electrical Engineering are available in the open access repository maintained by Aalto, Aaltodoc.
In our events, we are committed to Aalto University’s principles for a safer space.
Title of the thesis: Tensor and AI for Intelligent 6G Wireless Systems
Thesis defender: Tingting Zhang
Opponents: Prof. Martin Haardt, Ilmenau University of Technology, Germany and Prof. André L. F. de Almeida, Federal University of Ceara, Brazil
Custos: Prof. Sergiy Vorobyov, Aalto University School of Electrical Engineering
Future wireless networks are expected to do more than communication. They are envisioned to support an increasing number of smart devices and applications with growing levels of environmental awareness, interactivity, and intelligence, such as fully autonomous driving, unmanned aerial vehicles, and large-scale industrial automation. To realize these capabilities, a range of novel network paradigms and enabling technologies are being actively explored. Among them, this thesis focuses on two key areas: integrated sensing and communications (ISAC) and cell-free massive MIMO (CFmMIMO).
ISAC is an emerging design paradigm that enables wireless systems to communicate and sense their surroundings simultaneously. This capability is particularly relevant to connected and automated vehicles, which require accurate, wide-area sensing and high-rate communication at the same time. To address these requirements, the thesis develops a new multibeam method for accurate sensing over a wide area and introduces a time-division scheme to balance the competing performance requirements of sensing and communication. It also develops techniques to address velocity estimation challenges, particularly for high-speed objects. CFmMIMO is a novel network architecture in which a number of distributed access points (APs) cooperate to serve user equipments (UEs). Such networks can provide more uniform quality of service across different user locations and improve resource efficiency through effective coordination among APs and UEs. However, the large numbers of APs and UEs in CFmMIMO can lead to exponentially increasing computational complexity and high latency. The thesis adopts a learning-based approach to power allocation. An attention-enhanced graph neural network is proposed that achieves superior performance, adapts to dynamic wireless network topologies, and significantly reduces computational complexity.
This thesis provides theoretical insights and develops practical methods to address key challenges in next-generation wireless networks. These advances collectively aim to drive a transition toward environmentally aware, interactive, intelligent wireless ecosystems that support the evolving needs of digital society.
Key words: tensor, artificial intelligence, transmit beamspace design, integrated sensing and communications, power control in cell-free massive MIMO
Thesis available for public display 7 days prior to the defence at Aalto University's public display page.
Doctoral theses of the School of Electrical Engineering are available in the open access repository maintained by Aalto, Aaltodoc.