based on Adaptive Spectral Shaping for LFM Waveforms Van Khanh - - PowerPoint PPT Presentation

1

A Range Sidelobe Suppression Technique based on Adaptive Spectral Shaping for LFM Waveforms

Van Khanh NGUYEN and Mike TURLEY High Frequency Radar Branch NSI Division, Defence Science and Technology Group 30 August 2018

2

Outline

• Introduction – range sidelobes
• Existing range sidelobe suppression techniques
• Low complexity techniques: spectral shaping

– Conventional non-adaptive technique – Proposed adaptive technique

• Complexity analysis
• Results
• Summary

3

Introduction – Range sidelobes

Range (km) Doppler (Hz) Range (km) Doppler (Hz)

4

Existing techniques

• Techniques based on CLEAN

– Strong echoes are estimated and subtracted from the received signal – Perform well for discrete point targets but fail in cases of contiguous scattering sources.

• Adaptive pulse compression techniques

– An optimal filter for each range cell – One inversion of a structured covariance matrix for each filter. – Not suitable for radar systems with a high-time bandwidth product.

• Bandwidth extrapolation techniques

– Extrapolate the bandwidth and apply a non-adaptive window over the extrapolated spectral region. – Require estimation of a set of auto-regressive models.

• Spectral shaping based techniques

– Low complexity

5

Signal model

Transmitted waveform Received signal # targets # paths # sweeps Target return Clutter return Transmitted waveform range delay Doppler shift

Frequency Time

6

LFM waveform

Waveform Bandwidth

Pulse compression: Spectral response:

7

Non-adaptive spectral shaping

Reference waveform: Spectral response: Output:

8

Non-adaptive spectral shaping

Spectral windowing: Various range sidelobe suppression levels

Range Doppler Range Doppler Range Doppler Range Doppler

With windowing Without windowing

Without windowing Spectral windowing Power lost Conventional range processing 13dB suppression Sidelobe suppression level

9

Adaptive reference waveform

Reference waveform: Spectral response: One reference waveform for each evaluated Doppler bin

10

Proposed adaptive spectral shaping

Reference waveform: Spectral response: Output:

11

Proposed adaptive spectral shaping

Proposed approach Non-adaptive approach

12

Complexity analysis

number of samples per pulse period

FFT and IFFT can be efficiently implemented by arithmetic operations

Non-adaptive spectral shaping: Proposed adaptive spectral shaping: Adaptive reference waveform: Pulse compression: 2 FFT & 1 IFFT of length NrP Doppler processing: Nr FFT of length P requires additional 2 FFT & 1 IFFT of length Nr for each Doppler bin requires additional 1 FFT & 1 IFFT of length NrP for each Doppler bin 1 FFT of the data of length NrP

13

Results – Simulated data

1.6 dB gain 1.7 dB gain 0.4 dB gain 0.9 dB gain

14

Results – Real data

JORN Transmitter antenna array JORN Receiver antenna array

15

Results – Real data

1.6 dB gain

16

Summary

• Problem addressed:

– How to adaptively control range sidelobes of target returns for radar systems employing LFM waveforms.

• Existing techniques:

– CLEAN based techniques – Adaptive pulse compression

• Spectral shaping:

– Conventional non-adaptive spectral shaping – single reference waveform – Adaptive reference waveform – one reference waveform for each evaluated Doppler bin. – Proposed adaptive spectral shaping – on reference waveform for all evaluated Doppler bins with the spectral shaping being applied after range-Doppler processing

• Complexity of the proposed method: less than 2 times that of the

conventional non-adaptive spectral shaping.

• Performance: SNR gain of around 1.7 dB for weak targets