Performance Analysis of Data Detection in the THz-Band under Channel-Correlated Noise

Almutasem Bellah Enad, Hadi Sarieddeen, Jihad Fahs, Hakim Jemaa, Tareq Y. Al-Naffouri · 2025

Read Performance Analysis of Data Detection in the THz-Band under Channel-Correlated Noise by Almutasem Bellah Enad, Hadi Sarieddeen, Jihad Fahs, Hakim Jemaa, Tareq Y. Al-Naffouri (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-03702.txt.

Introduction

Almutasem Bellah Enad and four co-authors present a 2025 paper on error rates for data detection in terahertz wireless links. Terahertz bands promise very fast short range links, but their channels fade in complex ways and hardware adds its own distortion. Simple receiver math that assumes tidy, independent noise can mislead design badly.

The team builds an error rate framework for a common linear receiver. They first work out the error rate under the usual independent assumptions. Then they redo it with channel and noise linked, using a coupling method that tracks the dependence. A third case adds distortion noise from flawed hardware. Simulations show the independence shortcut can skew error rates by several decibels.

Catalog record

CREATORAlmutasem Bellah Enad, Hadi Sarieddeen, Jihad Fahs, Hakim Jemaa, Tareq Y. Al-Naffouri
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE26 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-03702

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Performance Analysis of Data Detection in the THz-Band under Channel-Correlated Noise Almutasem Bellah Enad, Hadi Sarieddeen, Jihad Fahs, Hakim Jemaa, Tareq Y. Al-Naffouri Abstract We present a comprehensive symbol error rate (SER) analysis framework for link-level terahertz (THz)-band communication systems under linear zero-forcing (ZF) data detection. First, we derive the mismatched SER for indoor THz systems under independent channel and noise assumptions, calculating the probability density function of the ratio of Gaussian noise to α \alpha - μ \mu channels resulting from ZF filtering. Next, we derive the precise SER under correlated channel and noise conditions, modeling dependencies using the copula method. Finally, we evaluate the SER for THz channels with correlated distortion noise from hardware impairments. Simulations demonstrate that the proposed framework corrects for multi-dB SERs resulting from the channel-noise independence assumption. I introduction Terahertz (THz) communications will play a central role in next-generation wireless systems. Spanning 0.3 to 3 THz, the THz band promises terabit-per-second (Tbps) data rates, supporting diverse communication and sensing applications [ 1 , 2 ] . However, achieving ultra-reliability, low latency, and high energy efficiency remains challenging [ 3 ] , and performance analysis frameworks are still not mature. At the link level, linear data detection methods are favored for their computational efficiency, essential under stringent Tbps constraints. Yet, their performance under THz-specific channels and noise remains underexplored. Zero-forcing (ZF) detection, which reduces interference by applying the pseudo-inverse of the channel, generally yields higher symbol error rates (SERs) than minimum mean squared error (MMSE) detection, particularly at low signal-to-noise ratios (SNRs) [ 4 ] . However, ZF offers lower complexity and does not require SNR knowledge. Although most linear detection SER studies assume Gaussian noise post-filtering, optimizing THz systems demands accurate channel and noise modeling [ 5 , 6 ] . In practical THz systems, channel-noise correlation can arise from hardware imperfections causing transmit and receive noise distortions or from channel-induced molecular absorption noise [ 5 , 2 ] . Traditional independent noise models fail to capture such effects [ 7 , 8 ] . Incorporating channel-correlated noise is more physically consistent and increasingly relevant in multi-antenna deployments, addressing a key gap in the literature. Small-scale fading in indoor THz channels can be effectively modeled by the \alpha - \mu distribution with flexible parameters [ 6 , 9 ] , which has been used in frameworks incorporating misalignment fading [ 5 , 10 ] and random fog effects [ 11 ] . Copula theory [ 12 , 13 ] has been widely adopted in wireless communications to model complex dependencies, including Nakagami- m fading with tail dependence [ 14 ] , and non-standard inter-band fading [ 15 ] . Copula-based methods have also been used to model correlation in the context of multiple access channels [ 16 ] , intelligent-surface-aided systems [ 17 ] , and physical layer security in wiretap channels [ 18 ] . In this paper, we propose a SER analysis framework for single-input single-output (SISO) THz links, addressing both correlated and independent channel and noise scenarios. While SER analysis is well-established in the literature [ 7 , 11 , 5 ] , our work extends it by incorporating THz-specific challenges such as \alpha - \mu fading, hardware impairments, and channel-correlated noise. These factors are often overlooked in conventional analyses but are essential for a realistic assessment of THz system performance. Our contributions are threefold. First, we analyze the SER under independent channel and noise assumptions by deriving the probability density function (PDF) of the ratio between Gaussian noise and the \alpha - \mu channel. Second, we evaluate the SER under correlated channel and noise conditions using the copula method, which captures statistical dependencies for more realistic error analysis. Finally, we adopt a practical THz communication model, derive the corresponding SERs accounting for channel–distortion noise correlation due to hardware impairments, and complement the analysis with asymptotic results. Throughout the paper, \left|\cdot\right| represents the absolute value, \mathbb{E}[\cdot] is the expectation operator, and \mathcal{CN}(0,2\sigma^{2}) denotes circularly symmetric complex Gaussian distributions with independent and identically distributed real and imaginary components, each with variance \sigma^{2} . Q(x)\!=\!\frac{1}{\sqrt{2\pi}}\int_{x}^{\infty}e^{\left(-\frac{u^{2}}{2}\right)}du is the Q(.) function [ 19 ] , \Gamma(.) is the gamma function, and \gamma(\zeta,u)\!=\!\int_{0}^{u}t^{\zeta-1}e^{-t}dt is the lower incomplete gamma function. G_{p,q}^{m,n}\left[z\left\lvert\begin{smallmatrix}a_{1},\ldots,a_{p}\\ b_{1},\ldots,b_{q}\end{smallmatrix}\right.\right] is the Meijer G-function [ 19 , eq. (9.301)] . II system Model and Problem Formulation We analyze the impact of key THz-specific factors on system performance, providing insights into the robustness of THz communications. Specifically, we compare two models: a standard model capturing fast fading via the \alpha - \mu distribution [ 5 , 11 ] , and a specialized model accounting for hardware impairments as distortion noise [ 5 ] . In both cases, \alpha - \mu fading effectively models small-scale THz channel variations. II-A Generic THz system model We consider a system of baseband input-output relation, r=\sqrt{p_{t}G_{t}G_{r}}h_{p}h_{f}s+n=\nu h_{f}s+n=hs+n, (1) where r is the received symbol, p_{t} is the average transmit power, G_{t} / G_{r} are the transmit/receive antenna gains, and s is the transmitted quadrature amplitude modulation (QAM) symbol. The additive noise, n\sim\mathcal{CN}(0,2\sigma^{2}) , has a Rayleigh-distributed amplitude with a PDF \displaystyle f_{|n|}(x) \displaystyle=\frac{x}{\sigma^{2}}\exp\left(-\frac{x^{2}}{2\sigma^{2}}\right),\quad\text{for}\ \,x\geq 0. (2) The THz channel is characterized by h_{p} and h_{f} . Here, h_{p} represents free-space path loss due to spreading and molecular absorption, given by h_{p}=\left(c/(4\pi fd)\right)^{\frac{\varrho}{2}}e^{-\frac{1}{2}K_{\text{abs}}d}, where c is the speed of light, f is the operating frequency, d is the communication distance, and K_{\text{abs}} is the molecular absorption coefficient (more details in [ 20 ] ). In measurement-based sub-THz/THz works [ 20 , 7 ] , the path loss exponent, \varrho , is best-fit to 2 . Finally, h_{f} represents complex small-scale fading of \alpha - \mu -distributed magnitude [ 8 ] . Defining \nu\!\triangleq\!\sqrt{p_{t}G_{t}G_{r}}h_{p} , the PDF and CDF of h\!=\!\nu h_{f} can be expressed as \displaystyle f_{|h|}(y) \displaystyle=\!\frac{\alpha\mu^{\mu}y^{\alpha\mu-1}}{(\hat{Z}\nu)^{\alpha\mu}\Gamma(\mu)}\exp\left(-\mu y^{\alpha}/(\nu\hat{Z})^{\alpha}\right), (3) \displaystyle F_{|h|}(y) \displaystyle=\frac{\gamma{(\mu,\mu y^{\alpha}/(\hat{Z}\nu)^{\alpha})}}{\Gamma(\mu)}, (4) where \alpha\!>\!0 is a fading parameter, \mu is the normalized variance of the fading channel, and \hat{Z}=\sqrt[\alpha]{\mathbb{E}[|h_{f}|^{\alpha}]} is the \alpha root mean value of the fading channel. II-B Specific THz system model We consider a model that incorporates THz distortion noise from hardware impairments, capturing the dependency between channel fading and receiver noise [ 5 , 21 ] , r=\sqrt{p_{t}G_{t}G_{r}}h_{p}h_{f}(s+n_{t})+n_{r}+n=h(s+n_{t})+n_{r}+n, (5) where \kappa_{t},\kappa_{r}>0 represent hardware imperfections, n_{t}\!\sim\!\mathcal{CN}(0,\kappa_{t}^{2}) and n_{r}\!\sim\!\mathcal{CN}(0,\kappa_{r}^{2}|h|^{2}) are distortion noise components from transmit and receive impairments, respectively, and n\!\sim\!\mathcal{CN}(0,2\sigma^{2}) is independent of h . Here, \left|h\right| also follows an \alpha - \mu distribution, but only n_{r} is correlated with h . II-C Problem formulation At the receiver, a linear ZF detector equalizes (inverts) the channel. The post-filtering generic system model ( 1 ) is thus \tilde{r}=h^{-1}r=s+n/h=s+z, (6) where z\!=\!n/h is the resultant effective noise. Similarly, the post-filtering system model in ( 5 ) is \tilde{r}=h^{-1}r=s+\frac{hn_{t}+n_{r}+n}{h}=s+n_{t}+n_{r}^{\prime}+\frac{n}{h}=s+z, (7) where n_{r}^{\prime}\!=\!n_{r}/h\!\sim\!\mathcal{CN}(0,\kappa_{r}^{2}) . Our aim is to develop a generic framework for calculating the SER for these system models without necessarily assuming a channel-independent noise. We distinguish three scenarios for which we derive the PDFs of the corresponding effective noise variables and leverage them in finding the SERs. III Proposed SER Analyses III-A Independent channel and noise model We compute the SER by analyzing the statistics of the ratio between the additive noise and the channel. For ZF detection

Read the complete text