Optimization of coded signals based on wavelet neural network
Pulse compression technique is used in many modern radar signal processing systems to achieve the range accuracy and resolution of a narrow pulse while retaining the detection capability of a long pulse. It is important for improving range resolution for target. Matched filtering of binary phase...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English English |
| Published: |
2015
|
| Subjects: | |
| Online Access: | http://eprints.uthm.edu.my/1386/2/MUSTAFA%20SAMI%20AHMED%20COPYRIGHT%20DECLARATION.pdf http://eprints.uthm.edu.my/1386/1/24p%20MUSTAFA%20SAMI%20AHMED.pdf http://eprints.uthm.edu.my/1386/3/MUSTAFA%20SAMI%20AHMED%20WATERMARK.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Pulse compression technique is used in many modern radar signal processing
systems to achieve the range accuracy and resolution of a narrow pulse while
retaining the detection capability of a long pulse. It is important for improving range
resolution for target. Matched filtering of binary phase coded radar signals create
undesirable sidelobes, which may mask important information. The application of
neural networks for pulse compression has been explored in the past. Nonetheless,
there is still need for improvement in pulse compression to improve the range
resolution for target. A novel approach for pulse compression using Feed-forward
Wavelet Neural Network (WNN) was proposed, using one input layer and output
layer and one hidden layer that consists three neurons. Each hidden layer uses Morlet
function as activation function. WNN is a new class of network that combines the
classic sigmoid neural network and wavelet analysis. We performed a simulation to
evaluate the effectiveness of the proposed method. The simulation results
demonstrated great approximation ability of WNN and its ability in prediction and
system modeling. We performed evaluation using 13-bit, 35-bit and 69-bit Barker
codes as signal codes to WNN. When compared with other existing methods, WNN
yields better PSR, low Mean Square Error (MSE), less noise, range resolution ability
and Doppler shift performance than the previous and some traditional algorithms like
auto correlation function (ACF) algorithm. |
|---|