Abstract
Sixth-generation (6G) wireless communication systems, such as Terahertz (THz) systems, are expected to provide transformative solutions for a fully connected world by offering high capacity, massive connectivity, high reliability, and low latency. However, as 6G systems move to higher frequencies, such requirements present significant challenges due to rapid attenuation and weak penetration. Intelligent Reflecting Surface (IRS) has emerged as a promising technique that can control the phase shift and reflection of the incoming signal toward the destination, achieving high spectral efficiency at a low hardware cost.While offering great potential, the IRS-assisted wireless networks also face challenges in channel modeling, system design, and network performance analysis, as the IRS-aided transmission constructs a mixed channel composed of a direct link between Base Station (BS) and User Equipment (UE), as well as a cascaded link across the BS, IRS, and UE. In this thesis, we propose a unified framework to statistically model the IRS-aided channel, determine the globally optimal locations of the IRS, and further analyze the system performance of the IRS-aided network across three types of IRS, namely passive, active, and hybrid passive/active IRS.
First, future wireless systems will operate in diverse environments, which often result in complicated propagation characteristics, including the clustering of scattered multipath contributions, shadowing caused by obstacles or human movement, and random fluctuations of received signals. To address this, we propose a novel statistical channel modeling approach based on the mixture Gamma distribution. By demonstrating its multiplicability and quadratic form, we establish its effectiveness in modeling the cascaded and mixture channels in passive IRS-aided communications, even under generalized fading distributions. In particular, it is shown that the proposed channel method is highly accurate, achieves tractable analysis, and is extremely versatile. Moreover, using tools from stochastic geometry, we propose a uniform performance analysis framework for the passive IRS-aided multi-cell networks, where the performance metrics are presented as functions of signal-to-interference-plus-noise ratio (SINR).
Second, the stochastic geometry-based framework is also applied to evaluate the active IRS-aided single-cell network performance with a customized deployment strategy for the active IRS, which is designed to further mitigate the productive path loss of the IRS-aided cascaded channel. It is shown that the deployment plays a key role in the performance of active IRS-aided transmission as there is a big performance gap between the nearest and opportunistic association policies for the IRS. Although the customized deployment of the IRS can narrow this gap, a theoretical study on the optimal locations for active IRS is crucial for the system design, especially when there is a large-scale network.
Third, we investigate the globally optimal deployment strategy for all three types of IRS, i.e., passive, active, and hybrid IRS. We employ the geometric models for integrated path loss distance (known as Cassini oval and Ellipse for product- and sum-distance path loss laws, respectively) and use them to determine the optimal locations of the hybrid IRS, where the outcomes for both passive and active IRS can be obtained as special instances of the hybrid IRS. Then, we design a novel opportunistic association policy for hybrid IRS based on the integrated path loss models, which facilitate the system performance analysis with the accurate distribution of the path loss distance in IRS-aided transmission. Furthermore, the system performance with opportunistic association for all three types of IRS is investigated. It is shown that the opportunistic association policy provides significant performance gain compared with the nearest association policy.
Fourth and last, we provide a use case for the proposed mixture Gamma distribution-based small-scale fading channel modeling method and geometric models. Based on the theoretical findings, we derive a rigorous upper bound of the signal-to-noise ratio (SNR) of the active IRS-aided communications to effectively handle the interdependence between the received signal and thermal noise generated by active parts.
| Date of Award | 29 Aug 2024 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Young Jin CHUN (Supervisor) |