Abstract
Omni-directional pathloss, which refers to the pathloss when omni-directional antennas are used at the link ends, is essential for system design and evaluation. In the millimeter-wave (mm-Wave) and beyond bands, high-gain directional antennas are widely used for channel measurements due to significant signal attenuation. Conventional methods for omni-directional pathloss estimation are based on directional scanning sounding (DSS) systems, where a single directional antenna is rotated to capture signals from different angles. However, these methods are often inaccurate due to relatively wide main beams and high side-lobes. This paper implements a directional antenna-based virtual antenna array (VAA) system for omni-directional pathloss estimation. The VAA scheme offers high angular resolution and low side-lobes, which are essential for achieving accurate multipath detection in power angular delay profiles (PADPs). A measurement campaign conducted in an indoor corridor at 28-30 GHz verifies the effectiveness of the proposed method.
Introduction
In the development of 5G and 6G networks operating in mmWave and sub-THz bands, accurate omni-directional pathloss models are critical for network design. However, physical constraints make high-quality omni-directional antennas difficult to implement at these frequencies, often resulting in poor link budgets. Consequently, researchers typically employ Directional Scanning Sounding (DSS), where a high-gain directional antenna is rotated 360 degrees to estimate total power.
Traditional DSS methods suffer from significant inaccuracies. Because physical directional antennas have relatively wide beams and high side-lobes, they often "double-count" signal energy across multiple rotation steps. This leads to an overestimation of received power and an underestimation of pathloss. Conversely, attempting to isolate individual paths with a "blurry" wide beam can cause weaker signals to be missed entirely.
Methodology
The proposed approach utilizes a Virtual Antenna Array (VAA) to overcome the physical limitations of the antenna hardware. By offsetting a corrugated horn antenna by 0.15 meters from the center of a mechanical rotator, the system creates a Virtual Uniform Circular Array (UCA) as it rotates. In this study, 240 elements were synthesized using 1.5° rotation steps.
By applying a modified classical beamforming algorithm to the data from these 240 virtual positions, the researchers synthesize a virtual antenna pattern that is significantly narrower and has lower side-lobes than the physical horn. This "sharper" vision allows for the generation of high-resolution Power Angular Delay Profiles (PADPs), where individual propagation paths can be identified as local maxima and summed accurately to calculate pathloss.
Experimental Setup
To validate the VAA method, a measurement campaign was conducted in an indoor corridor at Aalborg University. The setup utilized a Vector Network Analyzer (VNA) operating between 28 and 30 GHz. Measurements were taken across both Line-of-Sight (LOS) and Non-Line-of-Sight (NLOS) scenarios at distances ranging from 8 to 33.4 meters. The VAA results were benchmarked against two standard DSS reference methods and theoretical free-space pathloss models.
Results and Discussion
The core advantage of the VAA method lies in its superior angular resolution. While traditional DSS methods produce "blurry" PADPs where paths overlap, the VAA approach resolves distinct multipath components.
Key findings include:
- Accuracy: In LOS scenarios, VAA estimates closely matched theoretical free-space pathloss.
- Error Reduction: Traditional "sum-all-power" DSS methods underestimated pathloss by up to 3.9 dB due to double-counting.
- Resolution: Reference methods attempting path identification via standard DSS showed errors up to 2.8 dB (LOS) and 4 dB (NLOS) because the wide physical beam smeared weak paths.
Conclusion
The Virtual Antenna Array approach significantly improves the accuracy of omni-directional pathloss measurements in mmWave bands by providing the angular resolution necessary to separate multipath components. While the method requires longer measurement times and assumes a static environment, it offers a robust alternative to traditional scanning techniques. Future work will explore extending this methodology to 3-D (azimuth and elevation) and conducting broader measurement campaigns in diverse environments.
Summary updated