High Speed High Signal-to-Noise Ratio Antenna Measurements -- Demonstration for UAV-Based Near-Field Measurements of Modulated Terrestrial Navigation Signals

Technical Summary: High Speed High Signal-to-Noise Ratio Antenna Measurements

Problem Statement
Antenna measurements, particularly near-field (NF) scans requiring dense spatial and frequency sampling, are often limited by measurement duration. Achieving a high signal-to-noise ratio (SNR) typically necessitates long integration times or narrow intermediate frequency (IF) bandwidths in the receiver. In stepped measurement modes, this results in significant waiting times at each position; in continuously moving systems, it restricts the speed of the positioner to ensure signal stationarity. While using multi-frequency or modulated transmit signals could theoretically allow simultaneous measurement of multiple frequencies, standard vector network analyzers (VNAs) are optimized for continuous wave (CW) signals, and handling strong temporal magnitude variations in multi-frequency signals presents practical challenges. Furthermore, time-domain impulse measurements, while fast, suffer from linearity and signal magnitude generation issues.

Methodology
The paper proposes a signal processing approach to decouple measurement speed from SNR requirements. The core concept involves performing measurements with short time durations (corresponding to a wide measurement bandwidth) to capture all signal components rapidly, followed by a post-processing sequence to extract high-SNR data.

The methodology proceeds as follows:

1. Data Acquisition: The system collects measurement samples using short integration times, capturing signals that may contain single CW components, multiple discrete frequencies, or periodically modulated carriers (e.g., terrestrial navigation signals).
2. Frequency Domain Transformation: The collected time-domain data is transformed into the frequency domain via a Fourier transform.
3. Bandpass Filtering: Specific frequency components of interest are isolated using narrow bandpass (BP) filters. This step effectively removes out-of-band noise, simulating the effect of a long integration time without the associated time penalty.
4. Inverse Transformation: The filtered spectral components are transformed back into the time domain via an inverse Fourier transform.
5. Signal Extraction: The resulting time-domain signals represent the desired measurement quantities (e.g., channel transfer functions) with high SNR, derived from the narrow effective bandwidth of the BP filter, despite the original short measurement time.

This approach is applicable to both single-frequency CW signals and signals containing multiple discrete frequency components, such as periodically modulated carriers found in navigation systems.

Key Contributions and Results
The paper validates this approach through numerical simulations and real-world UAV-based measurements:

• Numerical Simulations:

  • Simulations demonstrated that the method successfully extracts channel transfer functions from noisy data.
  • For a single CW signal, the BP filtering improved the SNR by approximately 35.2 dB, consistent with the theoretical ratio of the measurement bandwidth (1 MHz) to the filter bandwidth (300 Hz).
  • In a multi-signal scenario with three superimposed CW signals, the method successfully isolated a signal component that was 20 dB weaker than others, achieving an SNR improvement of 39.0 dB (1 MHz to 120 Hz).

• UAV-Based Near-Field Measurements:

  • The approach was applied to modulated terrestrial navigation signals collected by an uninhabited aerial vehicle (UAV) at two sites: a Doppler VHF Omnidirectional Radio Range (DVOR) in Ottersberg, Germany, and an Instrument Landing System (ILS) localizer in Oberpfaffenhofen, Germany.
  • DVOR Measurements: The system processed 30 Hz AM and FM modulated carriers. The BP filtering successfully extracted CW-like NF data from the modulated signals. Inverse source solutions derived from this data accurately reconstructed the radiation centers (central element for AM, rotating ring for FM) and produced far-field (FF) patterns that matched simulations and expected physical behaviors.
  • ILS Measurements: The method was used to process the ILS localizer signal, which features a CW carrier with 90 Hz and 150 Hz AM sidebands. The processing successfully separated the carrier and sidebands, allowing for the extraction of distinct NF datasets. Inverse source solutions revealed the slightly different main beam directions of the carrier and sidebands, consistent with the system's design.

Significance and Claims
The paper claims that this approach enables high-speed antenna measurements without sacrificing SNR. By utilizing short measurement times and applying digital signal processing (Fourier transforms and BP filtering), the method overcomes the traditional trade-off between measurement duration and noise performance.

The authors emphasize that this technique is particularly valuable for:

1. Accelerating measurements where mechanical positioning speed is a limiting factor.
2. Processing modulated operational signals (such as those from navigation aids) directly, without requiring the generation of specific test signals.
3. Extracting diagnostic information (via inverse source solutions) from complex, modulated near-field data collected by mobile platforms like UAVs.

The results demonstrate that the method effectively recovers CW frequency samples with short time-step sizes but high SNR, validating its utility for both simulated scenarios and practical, field-deployed antenna diagnostics.