Sep 2026· IEEE Internet of Things Journal· Vol 13, pp. 39518-39531· 0 citations· 45 references
Abstract
Terahertz (THz) communication is an essential component of sixth-generation (6G) wireless networks; however, its use is constrained by molecular absorption, ultrawideband beam squint, high- $\kappa $ Rician fading, and beam misalignment. The multiplicative fading penalty of passive RIS design can be avoided, and reliable coverage can be extended into the THz band using active, reconfigurable intelligent surfaces (A-RISs) that integrate per-element amplifiers. However, the simultaneous optimization of discrete phase shifts and continuous per-element amplification factors is a nonconvex, high-dimensional problem that is difficult to solve with conventional iterative solvers for the submillisecond coherence times of mobile THz users. A hardware-aware A-RIS-assisted THz communication framework is presented in this article, with an extensive channel model that accounts for frequency-selective molecular absorption, ultrawideband beam squint, high- $\kappa $ Rician fading ( $\kappa \in [{10,25}]$ dB), and stochastic beam misalignment. We present an unsupervised graph neural network-Transformer (GNN-Transformer) architecture to solve the resulting joint optimization problem in real time. Graph convolutional layers exploit local spatial dependencies among RIS elements, while Transformer attention layers capture global channel dependencies across the entire RIS panel. The network is trained end-to-end on channel realizations but, unlike optimal labeling, does not require optimal discrete-phase and continuous-amplification configurations, which are obtained in a single forward pass. The results of the simulations show that the proposed method, which performs an exhaustive search, achieves a higher SNR, meets the power budget without violating it, and reduces inference latency compared to successive convex approximation (SCA), thereby enabling large-scale deployments of THz A-RIS in real time.
The evolution of wireless systems toward sixth-generation (6G) networks demands data rates in the order of terabits
per second, sub-millisecond latency, and massive device connectivity that existing millimeter-wave (mmWave) infrastructure
cannot fully support. Terahertz (THz) communication, spanning the 0.1–10 THz band, has emerged as a promising enabler of
this vision because of its enormous available bandwidth. However, the practical deployment of THz networks is constrained by
severe propagation losses, high molecular absorption, and the substantial power consumption of THz-band hardware
components such as power amplifiers, mixers, and high-resolution data converters, raising serious concerns for the energy
sustainability of future networks. This paper presents a comprehensive investigation into energy-efficient THz communication
for sustainable 6G networks. We first characterize the propagation and hardware-level factors that govern THz energy
consumption, and then propose an integrated energy-efficient framework combining hybrid analog-digital beamforming,
intelligent reflecting surfaces (IRS), artificial intelligence (AI)-driven adaptive resource allocation, and renewable energy-aware
base station sleep scheduling. A system-level energy efficiency model is formulated, and simulation studies are carried out to
evaluate the proposed framework against conventional fully-digital THz architectures. Results show that the proposed scheme
achieves up to 58% improvement in energy efficiency (bits/Joule/Hz) at moderate transmit power levels and demonstrates
favourable scaling with increasing IRS array size. The findings offer practical design guidance for building THz-enabled 6G
infrastructure that balances ultra-high throughput with the sustainability goals of next-generation wireless networks.
Prof. Tejas Siddharth Nikalje· International Journal for Re...· 0 citations
The integration of non-orthogonal multiple access (NOMA) with sub-Terahertz and Terahertz (sub-THz/THz) wireless communications can be a key enabler for future 6G networks, offering enhanced spectral efficiency and massive connectivity in highly directional and bandwidth-rich frequency bands. In this work, we investigate the performance of a power-domain downlink terrestrial-to-UAV NOMA system under realistic propagation conditions and hardware impairments. We develop a comprehensive analytical framework to evaluate the outage probability (OP) and average bit error probability (ABEP) of the considered system operating over multi-cluster fluctuating two-ray (MFTR) fading channels, which provide a flexible and physically motivated model for small-scale fading in sub-THz and THz propagation environments. The proposed model jointly accounts for frequency-dependent path loss with molecular absorption, pointing error–induced beam misalignment, and imperfect successive interference cancellation (SIC), while supporting practical square $M$ -QAM modulation. An exact analytically tractable expression is derived for OP and a high-signal-to-noise ratio (SNR) approximation is obtained for ABEP, along with asymptotic results in the high-SNR regime that provide insight into the system’s coding gain and diversity behavior. The analytical results reveal two distinct operating regimes. Under strong-interference conditions, persistent multiuser interference leads to irreducible ABEP floors, whereas in mild-interference scenarios, the ABEP decays with the average SNR according to a power-law behavior, with performance limited by the dominant physical impairment, either multipath fading or beam misalignment. Monte Carlo simulations validate the analysis and highlight key design trade-offs involving power allocation, SIC imperfection, modulation order, beamwidth, carrier frequency, and channel richness.
H. Mora, Fernando Moya Caceres, N. V. Garzón et al.· IEEE Open Journal of the Com...· 0 citations
The transition to terahertz (THz) and millimeter-wave (mmWave) frequencies is essential for meeting the extreme data-rate demands of 6G wireless networks. However, the associated ultra-wide bandwidth induces severe beam squint—a frequency-dependent beam misalignment that drastically reduces array gain and spectral efficiency. Conventional approaches face an irreconcilable trade-off: fully digital processing offers high accuracy but with prohibitive computational complexity, while hardware-based true-time-delay (TTD) networks incur excessive power consumption and insertion loss. In this paper, we propose a novel fully-digital architecture termed Spectrum-Aligned DFT (SADFT), which eliminates beam squint purely in the digital domain via mathematical reconstruction, designed for a fully-digital transmit architecture. The SADFT organizes the computation as a single-path pruned radix-2 recursive signal flow graph—a computational structure that enables regular in-place memory access, pipeline-friendly stage-wise multiplier reuse, and direct mapping to existing FFT hardware accelerators by reconfiguring only the twiddle factor generation logic—that directly targets the frequency-variant spatial frequencies, achieving mathematically exact alignment (gain loss bounded by <inline-formula> <tex-math notation="LaTeX">$10^{-4}$ </tex-math></inline-formula> dB for the simulated configurations) through virtual dense-grid mapping and path pruning. We rigorously extend the framework to two-dimensional uniform planar arrays (UPAs), proving mathematically exact separability into cascaded 1D operations that achieve the theoretical lower bound of <inline-formula> <tex-math notation="LaTeX">$\mathcal {O}(M)$ </tex-math></inline-formula> complexity for an <inline-formula> <tex-math notation="LaTeX">$M$ </tex-math></inline-formula>-element array. Extensive simulations using THz-specific Saleh–Valenzuela multipath models and the 3GPP CDL-A clustered channel model at 300 GHz demonstrate that SADFT closely approaches the Shannon capacity, delivering over <inline-formula> <tex-math notation="LaTeX">$3\times $ </tex-math></inline-formula> capacity improvement over standard FFT-based beamforming in ultra-wideband scenarios and effectively removing the baseband computational barrier that has historically prevented fully-digital wideband beamforming from being computationally feasible. The proposed architecture offers a scalable, low-complexity solution for future extremely large-scale antenna arrays (ELAAs) in 6G systems.
Henry H. Liao, Kaiyuan Wang, Y. Wan· IEEE Open Journal of the Com...· 0 citations
Wideband terahertz (THz) communication systems assisted by active reconfigurable intelligent surfaces (RIS) are susceptible to frequency dependent beam squint, which degrades the array gain at band edge subcarriers and reduces the achievable rate. In active RIS architectures, mitigating this effect requires the insertion of true time delay (TTD) units, along with the joint optimization of their delays and the amplification and phase of all active RIS elements, while accounting for practical hardware constraints including amplifier noise, output power limitations, and finite resolution delay elements. In this work, we propose a hardware aware wideband active RIS design that jointly optimizes the per element amplitude, phase, and TTD responses to mitigate beam squint under realistic hardware impairments. The resulting non convex optimization problem captures the coupling among amplification, hardware induced noise, and frequency dependent beamforming, and is efficiently solved using a projected gradient ascent algorithm with element wise feasibility projections. Numerical results demonstrate that the proposed design significantly improves the signal to interference plus noise ratio (SINR) uniformity across subcarriers and consistently achieves the highest achievable rate among the considered benchmark schemes, including a frequency flat max min array gain active RIS design that optimizes the worst case array gain without employing TTD elements. Furthermore, the proposed scheme consistently outperforms phase only passive RIS, active RIS without TTD elements, and quantized delay active RIS baselines across different transmit power levels and active RIS noise conditions, demonstrating its effectiveness and robustness for practical wideband THz communications.
Faran Awais Butt, A. A. Nasir, A. Muqaibel· IEEE Open Journal of the Com...· 0 citations
The pinching-antenna system (PASS) uses dielectric particles along a low-loss waveguide as reconfigurable passive radiators. Existing analyses conclude that the in-waveguide attenuation is negligible at low frequencies and millimeter wave bands; we show this fails at terahertz (THz), where realizable waveguide losses are dramatically larger. We develop a unified wideband THz-PASS propagation model integrating in-waveguide attenuation, atmospheric absorption, molecular re-radiation noise, and beam squint. Closed-form results follow: a band-averaged coherence factor; a cluster-center placement satisfying a band-edge SINR equalization condition; an associated placement-inversion threshold; and a proposed \emph{Switched-Feed PASS} (SF-PASS) architecture in which a centrally located radio-frequency switch routes the signal among multiple waveguide segments, with a closed-form insertion-loss payoff threshold. Numerical evaluation at the best PASS-compatible THz operating point shows that SF-PASS substantially outperforms single-feed PASS in spectral efficiency and is competitive with a large-scale antenna array at much lower hardware costs.
Terahertz (THz) communication offers extreme data rates for next generation communication systems, yet it suffers from severe path loss. Though multiple-input multiple-output communication technology is a promising solution to overcome this, it may bring an excessive hardware complexity. Hence, index modulation stands as an alternative to exploit the multiple antenna structure and to reduce the hardware complexity. In this paper, a THz system employing dual-polarized variable generalized spatial modulation (DP-VGSM) is proposed, and its performance is examined in terms of bit error rate under free space loss, absorption loss, α-μ fading, and pointing errors. The performance of DP-VGSM is compared to spatial modulation (SM), generalized SM (GSM), variable GSM (VGSM), and DP-GSM. Computer simulations demonstrate that dual-polarized systems provide a significant advantage over single-polarized systems by reducing the correlation between channels and that DP-VGSM outperforms others.
Batu Burgu, Kerem Karadeniz, Ayşegül Zahter et al.· Signal Processing and Commun...· 0 citations
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduJul 15, 2026
Assistant Professor Pat Pataranutaporn describes a new interface that lets everyday users glimpse inside an AI's neural network before their chatbot ever says a word.
MIT News · Artificial Intelligence· news.mit.eduJul 6, 2026
PhD student Rachel Sava, winner of the Envisioning the Future of Computing Prize, explores transformative improvements and dystopian risks of neural technology.
