Noise-modulated GPR

Noise-modulation is a technique rarely mentioned in the GPR literature. Nonetheless, variants of the approach it have implemented in commercial 3D GPR systems for more than 15 years in Australia and 5 years in South-East Asia.

General approach

Noise-Modulated GPR systems operate by emitting coded signals into the ground. Like conventional Impulse GPRs, the signal emitted is an ultrabroadband time domain signal. Unlike those systems, however, the code adds phase adjustments that spreads the signal over time. Return signals measured by the GPR equipment are then cross-correlated with the original code (or other deconvolution methods are used) to produce the familiar GPR A-scan.  A general overview of noise-modulation in the context of GPR can be found in Ground Penetrating Radar 2nd Edition, 2004 (Daniels, Ed.) (link)

NM-GPR

NM-GPR is not merely an implementation of noise-modulation techniques with standard hardware. Rather, it is a complete reconsideration of the type of signals emitted by GPR systems and the hardware used to maximise collection efficiency, ease regulatory compliance and achieve superior 3D operational performance.

Limitations of conventional methods

Most GPR systems use a high-fidelity receiver, such as a 16-bit Analog to Digital Converter (ADC), to measure return reflections. These devices cannot measure the entire return signal in one operation, so instead combine many incremental measurements.

Impulse GPR systems do this by sampling the magnitude of the return for brief instant(s) after each broadband pulse. A varying delay is applied between pulse generation and sampling, and responses from many pulses are then combined to form the equivalent of one complete return (A-scan). Although simple, this approach is also highly inefficient because the ADC is only “listening” to a very small fraction of each return signal.

Step frequency GPR systems, on the other hand, emit a series of narrowband signals at discrete frequency steps. The amplitude and phase of the return signal is measured continuously at each step. An Inverse Fourier Transform (IFT) is then used to combine measurements from the complete set of frequency steps and produce the A-scan. While this is a far more efficient use of receiver hardware, it is also complex and difficult to scale. Furthermore, the narrowband signals are problematic in terms of regulatory compliance.