Channel Measurements and Modeling based on Composite Environmental Factor for Urban Street-Canyon Intersections

Abstract: In urban environments, vehicle-to-everything (V2X) communications require accurate wireless channel characterization. This requirement is particularly critical at street-canyon intersections, where building blockage and rich multipath propagation can severely degrade link reliability. Due to its unique environmental layout, the channel characteristics in urban canyon are influenced by building distribution. However, this feature has not been well captured in existing channel models. In this paper, we propose an environment-related statistical channel model based on 5.8~GHz channel measurements. We construct a composite environmental factor to characterize environmental differences in intersections. Then, the factor is incorporated into 3GPP path-loss model and further linked to small-scale channel parameters. Finally, accuracy of the proposed model is validated using second-order channel statistics. The results show that the proposed model can effectively characterize propagation properties of urban street-canyon intersection channels with different building conditions. The proposed model provides a physically interpretable and statistically effective framework for channel simulation and performance evaluation in urban vehicular scenarios.

• The paper introduces a composite environmental factor S that links urban intersection morphology directly with large-scale and small-scale channel statistics.
• It applies rigorous 5.8 GHz measurement campaigns and regression analysis to modulate path-loss and multipath parameters for both LOS and NLOS scenarios.
• The approach significantly reduces RMSE errors compared to standard models, paving the way for more realistic urban V2X system simulations.

Environment-Related Channel Modeling for Urban Street-Canyon Intersections

Introduction and Motivation

Reliable and accurate V2X channel models are crucial for the design and optimization of intelligent transportation systems in dense urban environments. Urban street-canyon intersections, characterized by high building density and NLOS propagation, pose significant challenges for channel modeling due to severe multipath and blockage effects. Existing standard models (e.g., 3GPP TR 38.901, WINNER II, COST-231 Walfisch–Ikegami) are not parameterized with respect to the fine-grained, scenario-dependent geometry of urban intersections. Their reliance on an "average morphology" assumption leads to pronounced deviations when applied to street-canyon intersections, especially in terms of path-loss, shadowing, and multipath dispersion statistics. The paper "Channel Measurements and Modeling based on Composite Environmental Factor for Urban Street-Canyon Intersections" (2604.01767) addresses these shortcomings by developing a parameterized, environment-related channel model, derived from comprehensive 5.8 GHz channel measurements, that directly links intersection morphology to large-scale and small-scale channel statistics.

Composite Environmental Factor: Model Definition

To systematically encapsulate the environmental morphological influence, the authors introduce a composite environmental factor $S$, aggregating average building height, building-height dispersion, and building density within the intersection observation window. This factor provides a low-dimensional, physically interpretable parameterization of the local environment. The factor is quantitatively defined as

$$S = 0.5\,h_\mathrm{height} + 0.2\,h_\mathrm{std} + 0.8\,\rho,$$

where $h_\mathrm{height}$ is the mean building height, $h_\mathrm{std}$ is the height standard deviation, and $\rho$ is building footprint density. These statistics are extracted directly from geographical data for intersections.

This environmental factor $S$ is embedded into both large-scale (path-loss) and small-scale (multipath) channel model parameters, enabling joint adaptation of both domains to morphological diversity.

Large-Scale Path-Loss Modeling

The path-loss modeling framework builds on the 3GPP UMi LOS/NLOS dual-slope structure, modified to directly incorporate $S$:

• The LOS path-loss exponent and offset are modulated by $S$ (coefficients $k_A$, $k_B$),
• The NLOS path-loss exponent and intercept are likewise modulated by $$S = 0.5\,h_\mathrm{height} + 0.2\,h_\mathrm{std} + 0.8\,\rho,$$0 (coefficients $$S = 0.5\,h_\mathrm{height} + 0.2\,h_\mathrm{std} + 0.8\,\rho,$$1, $$S = 0.5\,h_\mathrm{height} + 0.2\,h_\mathrm{std} + 0.8\,\rho,$$2),
• The breakpoint distance $$S = 0.5\,h_\mathrm{height} + 0.2\,h_\mathrm{std} + 0.8\,\rho,$$3 delineates the LOS-to-NLOS region,
• All parameters are extracted by regression from extensive vehicular measurement data across intersections with a wide $$S = 0.5\,h_\mathrm{height} + 0.2\,h_\mathrm{std} + 0.8\,\rho,$$4 range.

Measurement results reveal that the standard 3GPP UMi models significantly underestimate LOS path-loss (with errors increasing with $$S = 0.5\,h_\mathrm{height} + 0.2\,h_\mathrm{std} + 0.8\,\rho,$$5) in dense intersections and fail to track the NLOS breakpoint dynamics. The environment-related model reduces RMSE by approximately 8 dB (LOS) and 3 dB (NLOS) versus 3GPP, demonstrating robust adaptation to intersection morphology.

Small-Scale Multipath Parameterization

The small-scale model parameterizes multipath power ($$S = 0.5\,h_\mathrm{height} + 0.2\,h_\mathrm{std} + 0.8\,\rho,$$6), delay ($$S = 0.5\,h_\mathrm{height} + 0.2\,h_\mathrm{std} + 0.8\,\rho,$$7), AoA ($$S = 0.5\,h_\mathrm{height} + 0.2\,h_\mathrm{std} + 0.8\,\rho,$$8), EOA ($$S = 0.5\,h_\mathrm{height} + 0.2\,h_\mathrm{std} + 0.8\,\rho,$$9), and cluster structure ($h_\mathrm{height}$0, $h_\mathrm{height}$1) as random variables conditioned on $h_\mathrm{height}$2:

• Power, cluster number, and MPC count are modeled via normal distributions,
• Delay is lognormal in LOS, Laplace in NLOS,
• Angular spreads are Laplace distributed,
• All distribution parameters (location, scale) are explicit functions of normalized $h_\mathrm{height}$3.

The SAGE algorithm is utilized for delay and angular estimation; results establish that increasing $h_\mathrm{height}$4 (i.e., more complex, denser intersection) systematically increases delay and angular spreads and the number of significant multipath clusters, especially in NLOS. The effect magnifies as LOS conditions transition to NLOS around the corner.

Measurement Campaign and Model Validation

The dataset comprises wideband 5.8 GHz vehicular channel measurements in Changsha, China, covering three intersections with varying building density, height, and street complexity. The measurement system uses a multi-carrier sounding signal, omnidirectional Tx, and a $h_\mathrm{height}$5 Rx array, with full 3D radiation pattern calibration. Both LOS and NLOS trajectories are sampled by coordinated vehicle movement.

Model validation is performed via comparison of generated channels (for given $h_\mathrm{height}$6, $h_\mathrm{height}$7, $h_\mathrm{height}$8) with measured statistics in an independent test intersection. Key second-order and third-order metrics, notably RMS Delay Spread (DS), Azimuth Spread of Arrival (ASA), and Elevation Spread of Arrival (ESA), match measured distributions closely, confirming the statistical fidelity of the environment-related model.

Implications and Future Work

By directly coupling environment geometry with channel statistics, the presented framework provides a path toward environment-aware, scenario-adaptive vehicular channel modeling, bridging the gap between geometric, deterministic, and parameterized stochastic models. The ability to synthesize channel responses for arbitrary intersection geometries using directly measurable urban features is especially pertinent for:

• Urban V2X system-level simulation and performance evaluation (e.g., for adaptive beam management, interference analysis),
• Integrated sensing and communication (ISAC) studies requiring physically interpretable channel state control,
• Data-efficient site-specific channel prediction and generalization in virtual/augmented environments.

However, the model is empirically anchored at 5.8 GHz and urban Chinese morphologies; extension and revalidation for other frequency bands, propagation mechanisms (e.g., mmWave, sub-THz), and urban typologies remain open. The single-factor $h_\mathrm{height}$9 descriptor, while concise, may omit higher-order geometric features (corner shape, street furniture), necessitating future semantic-aware generative models (cf. [Zhang2026semantic]) and possible fusion with AI-based prediction frameworks [He2026AI],[He2025INTERACT].

Conclusion

This work establishes a unified environment-dependent statistical model for vehicular channels at urban street-canyon intersections, using a composite environmental factor to link physical morphology with large- and small-scale propagation metrics. The model outperforms standard baseline channel models, providing traceable, scenario-dependent path-loss and multipath statistics that match real intersection measurements. This approach advances the realism and adaptability of statistical channel models for next-generation urban V2X communications and provides a foundation for environment- and semantic-aware propagation modeling.