Urban street canyons—those corridors formed by closely packed, often tall buildings lining both sides of city streets—pose a unique and pressing challenge for next-generation wireless communication, especially as we move into the sub-terahertz (sub-THz) frequency range. Sub-THz wireless technology, which typically refers to frequencies between 100 GHz and 1 THz, promises ultra-high data rates and massive connectivity. Yet, these frequencies behave very differently from those used in today’s cellular and Wi-Fi systems, particularly in complex urban environments. How, then, do researchers and engineers model these sub-THz channels to predict, plan, and optimize network performance in the dense, canyon-like streetscapes that typify major cities?
Short answer: Sub-THz wireless channels in urban street canyon environments are modeled using a blend of deterministic ray-tracing and empirical/statistical methods. These approaches incorporate the unique propagation characteristics of sub-THz waves—such as high free-space path loss, pronounced sensitivity to obstacles and atmospheric effects, and the dominant influence of reflections, diffractions, and scattering from building surfaces and street furniture. Models are validated and refined through measurement campaigns in real urban canyons, capturing how signals behave under varied line-of-sight (LOS), non-line-of-sight (NLOS), and multipath conditions. The resulting models are far more complex than those at lower frequencies, requiring detailed 3D representations of the urban environment and careful consideration of frequency-dependent phenomena.
The Complex Playground of Sub-THz Propagation
To appreciate why modeling sub-THz channels in urban canyons is so intricate, it helps to understand the physical realities at play. Sub-THz signals have “short wavelengths and high frequency,” which means they can support extremely high data rates but are also “highly susceptible to blockage and absorption” by common urban features, as highlighted in research referenced by ieeexplore.ieee.org. Buildings, vehicles, trees, and even atmospheric molecules can significantly attenuate or scatter sub-THz waves. Unlike lower-frequency signals, which can bend around corners or penetrate walls to some extent, sub-THz signals are far more likely to be stopped or diverted, making the environment’s physical details critically important.
The most widely used approach for modeling these channels is deterministic ray-tracing. In this method, the urban canyon is mapped in detail—down to the shapes and materials of buildings, the presence of street lamps, signage, and even foliage. Then, the model simulates how electromagnetic waves would propagate from the transmitter to the receiver, accounting for direct paths, reflections off building facades and windows, diffractions around corners, and scattering from rough surfaces. Ray-tracing models can predict the received signal strength and multipath characteristics at specific locations, allowing for highly granular analysis.
According to several studies cited by ieeexplore.ieee.org, these ray-tracing models are “calibrated with real-world measurements” to ensure accuracy. This calibration is crucial because the reflectivity and absorptivity of building materials at sub-THz frequencies can differ significantly from those at lower frequencies, and empirical data is needed to fine-tune the simulations.
Statistical and Hybrid Models: Capturing the Unpredictable
While ray-tracing provides precision, it also demands exhaustive detail about the environment and significant computational resources. To balance realism with practicality, researchers often supplement deterministic modeling with statistical or hybrid approaches. For example, sciencedirect.com describes how “empirical path loss models” are developed by collecting measurement data in representative urban canyons and fitting mathematical functions to the observed signal attenuation as a function of distance, frequency, and environmental factors.
A well-known statistical model is the “multi-slope path loss model,” where the rate at which signal strength decreases with distance changes depending on whether the signal is in LOS, NLOS, or obstructed NLOS (OLOS) conditions. In urban canyons, this means modeling how a signal behaves differently if it travels straight down a street (LOS), bounces around a corner (NLOS), or is partially blocked by vehicles or foliage (OLOS). According to sciencedirect.com, these models often include “shadow fading” terms to account for random variations in signal strength caused by moving objects and small-scale features that are too complex to model deterministically.
Yet, statistical models alone cannot capture the full complexity of sub-THz propagation, especially when it comes to predicting multipath effects—where multiple signal copies arrive at the receiver after bouncing off various surfaces. Therefore, hybrid models, which combine deterministic and empirical/statistical elements, are gaining popularity. These might use ray-tracing for the most significant paths and supplement with statistical descriptions for weaker or more diffuse signal components.
Key Parameters and Environmental Considerations
Several concrete details must be incorporated into urban canyon channel models at sub-THz frequencies. First, the “free-space path loss” is much higher than at lower frequencies, so even in clear LOS conditions, signal strength drops off rapidly with distance. For example, at 300 GHz, the path loss over 100 meters can be more than 20 dB higher than at 28 GHz, as reported by studies referenced in ieeexplore.ieee.org.
Second, atmospheric absorption—especially by water vapor and oxygen—becomes non-negligible. At certain frequencies, absorption peaks can add several dB of loss per kilometer, even in clear air. This effect must be explicitly modeled, especially for longer-range links or when considering weather variability.
Third, the angular and delay spreads, which describe how much the arriving signal is spread out in direction and time, are typically lower in sub-THz channels due to the dominance of a few strong paths (primarily reflections from building facades) and the lack of strong diffraction around obstacles. According to sciencedirect.com, “multipath richness is reduced compared to microwave bands,” which impacts how well advanced antenna techniques like beamforming can exploit spatial diversity.
Fourth, the impact of mobile obstacles—like vehicles and pedestrians—is much more pronounced. A single bus or truck can completely block a sub-THz link, causing deep and sudden fades. Models must therefore include dynamic blockage, often using stochastic processes or time-varying ray-tracing to simulate the effects of moving objects.
Finally, material properties matter greatly. Glass, concrete, metal, and vegetation all reflect, absorb, and scatter sub-THz waves in different ways, and these effects are frequency-dependent. For instance, “the reflection coefficient of standard glass can drop by more than 10 dB” from 100 GHz to 300 GHz, as shown in measurement campaigns cited by ieeexplore.ieee.org.
Measurement Campaigns: The Ground Truth
To ensure that these models reflect real-world conditions, extensive measurement campaigns are conducted in actual urban street canyons. Researchers deploy transmitters and receivers at various heights and distances, systematically recording received signal strength, delay profiles, and angular spreads under different environmental conditions. For example, sciencedirect.com discusses measurement campaigns in cities like New York and Tokyo, where the “unique geometry of high-rise canyons” is specifically captured.
These campaigns often reveal practical insights that challenge or refine existing models. For instance, it has been observed that while LOS paths dominate, “specular reflections from windows and metallic surfaces can create significant secondary paths” that must be included in the model (as detailed on ieeexplore.ieee.org). Moreover, the presence of street-level clutter—cars, kiosks, foliage—can introduce rapid, unpredictable variations in received power, leading to the inclusion of time-varying and spatially correlated shadowing in advanced models.
From Models to Network Design
The ultimate goal of channel modeling is to inform the design and deployment of sub-THz wireless systems—everything from antenna placement to beamforming strategy to handover and routing protocols. Because sub-THz signals are so sensitive to environmental factors, accurate channel models are the foundation for reliable network performance prediction.
For example, in planning a sub-THz backhaul network for 6G in a city center, engineers use these models to determine “where links are feasible, what data rates can be supported, and how robust the connections will be to everyday variations” (as described on sciencedirect.com). They might simulate hundreds of possible transmitter-receiver placements, testing how well the system performs under different weather conditions, rush hour traffic, or urban development scenarios.
A Living Model: The Need for Continuous Refinement
Urban environments are never static, and neither are the channel models that describe them. As new materials are used in construction, as cities add greenery or smart infrastructure, and as measurement technologies improve, models are constantly updated. Collaboration between academia, industry, and standardization bodies—such as IEEE, as noted by ieeexplore.ieee.org—is crucial to ensure that models stay relevant and useful for real-world deployment.
Moreover, with the advent of machine learning, there is growing interest in using data-driven approaches to supplement or even partially replace traditional modeling. By training algorithms on large datasets collected in cities worldwide, researchers hope to develop models that can generalize across different urban morphologies and adapt in real time to environmental changes.
Summary and Outlook
In summary, modeling sub-THz wireless channels in urban street canyons is a sophisticated endeavor that blends deterministic ray-tracing, empirical/statistical modeling, and real-world measurement. The process must account for “high path loss, strong sensitivity to blockage, frequency-dependent material properties, and dynamic urban clutter,” as found on ieeexplore.ieee.org and sciencedirect.com. Models are validated and refined through extensive measurement campaigns, enabling accurate prediction of network performance in the uniquely challenging environment of the city street canyon. As we look toward the future of 6G and beyond, these models will only grow in complexity and importance, serving as the bedrock for the ultra-fast, ultra-reliable wireless networks that will define tomorrow’s cities.
