Spectrum Allocation for Terahertz Band Communication Systems

Abstract

Terahertz (THz) band communication (THzCom) has been envisaged as a highly promising paradigm for the sixth-generation (6G) and beyond era. In particular, the ultra-wide spectrum ranging from 0.1 to 10 THz and the short wavelengths at the THz band offer enormous potential to realize new applications that demand ultra-high-data rates. The exploration of novel and efficient spectrum allocation strategies is of paramount significance to harness the potential of the THz band. When such strategies are to be devised, the unique characteristics of the THz band, such as the frequency- and distance-dependent molecular absorption loss, pose new and pressing challenges that have never been seen at lower frequencies. Against this background, this thesis aims to design and analyze novel and efficient spectrum allocation strategies for multiuser THzCom systems that can wisely tackle the challenges imposed by the characteristics at the THz band.

First, in Chapter 2, we, for the first time, propose multi-band-based spectrum allocation with adaptive sub-band bandwidth (ASB) to improve the spectral efficiency of multiuser THzCom systems, by allowing to divide the to-be-allocated spectrum into sub-bands with unequal bandwidth. To study the impact of ASB, we formulate a resource allocation problem, while focusing on spectrum allocation. We then propose reasonable approximations and transformations and arrive at a solvable approximate convex problem. Aided by numerical results, we show that considering ASB leads to a significantly higher data rate as compared to considering equal sub-band bandwidth.

Second, in Chapter 3, we propose a novel spectrum allocation strategy for multiuser THzCom systems when the to-be-allocated spectrum is composed of multiple transmission windows (TWs). Specifically, we explore the benefits and designs of (i) sub-band assignment, (ii) avoiding using some spectra that exist at the edges of TWs where molecular absorption loss is very high, and (iii) ASB. With these considerations in mind, we formulate a resource allocation problem and solve it using a convex-optimization-based approach. Aided by numerical results, we show that data rate gains, as well as improvement in the feasibility region of the resource allocation problem, can be obtained by adopting ASB and optimally determining the unused spectra at the edges of TWs.

Third, in Chapter 4, we propose an unsupervised learning-based approach to obtaining the solution to the multi-band-based spectrum allocation problem with ASB. In the proposed approach, we first train a deep neural network (DNN) while utilizing a loss function that is inspired by the Lagrangian of the formulated problem, and then approximate the near-optimal sub-band bandwidths using the trained DNN. Aided by numerical results, we show that when the values of the molecular absorption coefficient within the to-be-allocated spectrum vary highly non-linearly, the proposed unsupervised learning-based approach outperforms the existing approximate approaches.

Finally, in Chapter 5, we investigate (i) the modeling and analysis of different types of blockages in a THzCom system and (ii) the coverage probability of an indoor three-dimensional (3D) THzCom system when frequency reuse is considered. We first characterize the joint impact of blockages caused by the user itself, moving humans, and wall blockers in a THzCom environment, and then derive the blockage probabilities associated with a point-to-point THzCom link. We then develop a tractable analytical framework, using stochastic geometry, to evaluate the coverage probability by characterizing the regions where dominant interferers (i.e., those that can cause outage by themselves) can exist, and the average number of interferers that exist in these regions. Using numerical results, we show the significance of blockage characterization and coverage analysis in a 3D THzCom system while considering the unique THz band propagation properties. Show more