Difference between revisions of "Projects:2020s1-1231 Radar Waveform Design"

From Projects
Jump to: navigation, search
(→‎Adaptive Waveform Design)
 
(19 intermediate revisions by 2 users not shown)
Line 2: Line 2:
 
[[Category:Final Year Projects]]
 
[[Category:Final Year Projects]]
 
[[Category:2020s1|1231]]
 
[[Category:2020s1|1231]]
−
 
−
Abstract here
 
−
== Introduction ==
 
−
This project will study the performance of several advanced digital binary and polyphase waveforms
 
−
and compare their performance to unmodulated pulses and linear FM pulses with comparable time-bandwidth products. Comparisons will be made in terms of key performance metrics, including sidelobe levels, Doppler tolerance, resolution, performance in the presence of noise and EM interference. This study will initially be carried out using theoretical simulations, with the findings then verified using a software-defined dual channel transceiver.
 
  
 
=== Project team ===
 
=== Project team ===
Line 18: Line 13:
 
* Dr Waddah Al-Ashwal (SRC Australia)
 
* Dr Waddah Al-Ashwal (SRC Australia)
  
−
==== Advisors ====
+
== Introduction ==
−
*  
+
This project will study the performance of several advanced digital binary and polyphase waveforms and compare their performance to unmodulated pulses and linear FM pulses with comparable time-bandwidth products. Comparisons will be made in terms of key performance metrics, including sidelobe levels, Doppler tolerance, resolution, performance in the presence of noise and EM interference. This study will initially be carried out using theoretical simulations, with the findings then verified using a software-defined dual-channel transceiver.
−
*
+
 
 +
== Background ==
 +
===Radar Theory===
 +
Radar is an acronym of Radio Detection and Ranging and the system emits EM waves. The basic principle of the radar system is to find and position a target and determine the distance between the target and transmitting antenna. The transmitting antenna would be using RF. After transmitting radio waves, the radar’s objective is to listen for any returning signals which will then be analysed for any presence of a target, this principle is also known as echolocation. A target is detected when it reflects some portions of the transmitted EM wave. There can be multiple designs of a Radar system however all Radars will have the following key features and processes in common.
 +
 
 +
[[File:Radar pic.jpg|thumb|Fig 1.Key Elements and Processes of Radar Systems<ref>M. Richards, J. Scheer and W. Holm, Principles of Modern Radar. Raleigh: SciTech Publishing, 2010.</ref>]]
 +
 
 +
Figure 1 describes the key features and processes of a monostatic radar system. It is considered as a monostatic radar because the transmitter and receiver are positioned at one location. Whereas a bistatic radar is where the transmitter and receiver are positioned differently such that the angle and ranges to the target are adequately different. Referring to figure 1, a radar system can be divided into four major subsystems which are the antenna, transmitters, receiver, and signal processor.
 +
 
 +
The radar transmitter produces short-lived, high-frequency energy (megawatts) pulses that generate EM waves which are then radiated into the propagation medium by the antenna. The T/R switch also known as the duplexer, its purpose is to provide a connection so that both the transmitter and receiver can be attached simultaneously to the antenna and provide separation between the transmitter and receiver to shield the sensitive receiver components from the high-powered transmission signal. After transmitting the signal into the propagation medium, if the signal encounters a target (e.g. aircraft, missiles, clouds, etc.), the EM waves are reradiated into the environment. However, when the target reradiates, it will produce a diffused signal due to the scattering effect. This occurs when the transmitted signal gets reflected in several different directions. Aside from the scattering effect from the desired target; there can be other surfaces on the ground and atmosphere which can reradiate the signal. Such unwanted and unnecessary but genuine signals are called clutter interference, which can be picked up by the receiver antenna.
 +
 
 +
While the EM waves propagate through the atmosphere (propagation medium), the environment could alter the intensity of the EM waves both at target and at the receiving antenna. Therefore, the receiver is designed in such a way so that it is sensitive to the frequency spectrum being transmitted and provides amplification and demodulation of the received RF signal. After amplifying the received signal, the system transforms the RF signal to an intermediate frequency (IF), and then send the signal to an analogue-to-digital converter (ADC) and then to the signal/data processor, where the demodulation will occur. To sense a presence of a target at its greatest range, the receiver antenna must be sensitive and not introduce any unnecessary noise.
 +
 
 +
As discussed before that when a receiver antenna receives a reflected signal there can be a presence of clutter. However the interference does not only arise from clutter, it could come in the form of internal or external noise; unintentional external EM waves created by man-made sources, also known as electromagnetic interference; and intentional jamming from electronic countermeasures (ECM) system. These interferences could be in the form of noise or false targets. Thus, the processor needs to identify the target in the presence of noise, clutter or jamming.
 +
 
 +
Previously, it was stated that due to the scattering effect, the reflected signal is diffused indicating that the received power will be weaker than the transmit power. However, in addition to that one of the major feature of the radar that can affect the intensity of the signal is the distance between the radar and the target. Assume that an isotropic antenna is emitting EM waves, the intensity of the propagating wave is calculated by the power emitted per unit area of the wave. Knowing that the radar acts as an isotropic antenna, it can be shown that the transmitted power density decreases by <math>\frac{1}{R^{2}}</math>:
 +
 
 +
{| class="wikitable"
 +
|-
 +
| <math>Q_{t}=\frac{P_{t}}{4 \pi R^{2}}</math>||
 +
where,
 +
* <math>P_{t}</math> is total radiated transmitted power
 +
* <math>R</math> is the distance from the isotropic antenna to the target
 +
|}
 +
 
 +
The range of a target is determined by multiplying the time travelled, from RADAR to the target and back to the radar in the medium (𝑡), and speed of light (𝑐).
 +
 
 +
{| class="wikitable"
 +
|-
 +
| <math>R=\frac{ct}{2}</math>||
 +
where,
 +
* 𝑅 is the slant range: the actual distance to the target and the radar
 +
* 𝑐 is the speed of light <math>\approx 3\times 10^{8}m/s</math>
 +
* 𝑡 is the total time taken for the EM wave to travel to the target and back to the radar
 +
|}
  
−
=== Objectives ===
+
Radar waveforms are generally classified under two classes: continuous wave (CW) and pulsed. The transmitter transmits a signal continuously with the CW waveform, usually without interruption, the whole time the radar transmitter is operating. The receiver also operates continuously. On the other hand, the pulsed waveform transmitter emits a sequence of pulses of finite duration, separated by the times in which the transmitter is "off." Since the transmitter is off, the receiver is on for detection of the target signals. For CW radars, it takes the bistatic configuration as the transmitter and receiver must be kept on constantly. In a CW radar, the round-trip time of the transmitted EM wave and hence the target range can be determined by alternating the composition of the wave (e.g.changing the frequency of the wave over time). This technique of FM effectively places a timing mark on the EM wave, thus allowing for the determination of the target range.
−
Creating radar waveform using Matlab to understand the basic principals.
 
  
−
== Background ==
+
Pulsed radars would usually take on the monostatic configuration as the transmitter and receiver are not utilized together and are kept isolated from each other. The radar transmits EM waves over a short period or pulse width, typically between 0.1 and 10 microseconds (𝜇𝑠), but sometimes as few as a few nanoseconds (10−9𝑛𝑠) and as long as a millisecond. As discussed previously the duplexer will provide the transition to radar to switch from transmitter to receiver, to allow the antenna to listen to echoes. The total time of transmitting (pulse width) and listening to the echoes is the completion of one pulsed radar cycle, also known as PRI.<ref>"Linear Frequency Modulated Pulse Waveforms- MATLAB & Simulink", Mathworks.com, 2020. [Online]. Available: https://www.mathworks.com/help/phased/ug/linear-frequency-modulated-pulse-waveforms.html</ref>
−
Radar is an acronym of Radio Detection and Ranging and the system emits EM waves. The basic principle of the radar system is to find and position a target and determine the distance between the target and transmitting antenna.
 
  
 
== Motivation ==
 
== Motivation ==
 
The problem of detecting small targets is challenging especially with a heavy clutter background. We aim to exploit waveform agility by tailoring the transmit waveform to have a second look on the target location to suppress the range bins with heavy clutter. An alternative method using mismatch filter was explored and implemented on the received signal to suppress sidelobes if waveform agility is not feasible.
 
The problem of detecting small targets is challenging especially with a heavy clutter background. We aim to exploit waveform agility by tailoring the transmit waveform to have a second look on the target location to suppress the range bins with heavy clutter. An alternative method using mismatch filter was explored and implemented on the received signal to suppress sidelobes if waveform agility is not feasible.
  
−
The agile waveform would suppress range bins with high clutter while the mismatch filter would instead be suppressing overall sidelobes. From both implementations, we would aim to decrease the false alarm rate and improve the detection performance.
+
The agile waveform would suppress range bins with high clutter while the mismatched filter would instead be suppressing overall sidelobes. From both implementations, we would aim to decrease the false alarm rate and improve detection performance.
  
 
== Adaptive Waveform Design ==
 
== Adaptive Waveform Design ==
Line 38: Line 65:
 
The diagrams below show the match filter output of the transmit waveform post suppression.
 
The diagrams below show the match filter output of the transmit waveform post suppression.
  
 +
{| class="wikitable"
 +
|-
 +
! Discription!! Results
 +
|-
 +
| By increasing the number of chips, the degree of freedom increases. Therefore, the suppression performance increases.||
 +
[[File:The effect of number of chips to the suppression level.png|thumb]]
 +
|-
 +
|  By shifting the suppression zone away from the mainlobe, the suppression performance increases as the initial energy of the range bins further from the mainlobe is lower.
 +
||
 +
[[File:The effect of the distance from the mainlobe to the suppression level.png|thumb]]
 +
|-
 +
| By increasing the suppression width, the number of range bins where the energy could escape to decreases, causing the suppression performance to decrease.
 +
||  [[File:The effect of the number of suppressed range bins to the suppression level.png|thumb|The effect of the number of suppressed range bins to the suppression level.png]]
 +
|-
 +
|  The effect of having multiple suppression zones does not affect the performance. ||
 +
[[File:The effect of having multiple suppression zones .png|thumb]]
 +
|}
 +
 +
== Mismatched Filter Design ==
 +
Phase coded waveforms generally produce high range sidelobes and to suppress the sidelobes mismatch filters are generally used. However, with the suppression of the sidelobes, it comes with a cost of loss to the main peak sensitivity. The following diagrams show the normalised results of the mismatched filter with changes in the parameters and compared to the matched filter outputs.
  
−
[[File:The effect of the number of suppressed range bins to the suppression level.png|frameless|The effect of the number of suppressed range bins to the suppression level.png]]
+
{| class="wikitable"
−
By increasing the number of chips, the degree of freedom increases. Therefore, the suppression performance increases.
+
|-
 +
! Description!! Results
 +
|-
 +
| Creating a mismatched filter of the same length, it can be shown that in order to suppress the sidelobes ( 15dB), there is significant loss in the main peak.||[[File: Matched and Mismatch filter same length.jpg|thumb]]
 +
|-
 +
| Increasing the length of the mismatch filter three times the original length. There is significant improvement in the suppression of the sidelobes (20dB) and there is negligible loss in the main peak.|| [[File:Matched and Mismatch filter different length.jpg|thumb]]
 +
|-
 +
| With the previous conditions and allowing a loss in the main peak, it can be shown that it can suppress the sidelobes even greater ( 25dB) compared when there is no loss in the main peak.||  [[File:Matched and Mismatch filter different length loss 0.3.jpg|thumb|]]
 +
|}
  
 
== Conclusion ==
 
== Conclusion ==
Line 48: Line 103:
  
 
== References ==
 
== References ==
−
[1] a, b, c, "Simple page", In Proceedings of the Conference of Simpleness, 2010.
 

Latest revision as of 17:04, 21 October 2020


Project team

Project students

  • Wong Ming Eer
  • Shalin Shah

Supervisors

  • Brian Ng
  • Dr Waddah Al-Ashwal (SRC Australia)

Introduction

This project will study the performance of several advanced digital binary and polyphase waveforms and compare their performance to unmodulated pulses and linear FM pulses with comparable time-bandwidth products. Comparisons will be made in terms of key performance metrics, including sidelobe levels, Doppler tolerance, resolution, performance in the presence of noise and EM interference. This study will initially be carried out using theoretical simulations, with the findings then verified using a software-defined dual-channel transceiver.

Background

Radar Theory

Radar is an acronym of Radio Detection and Ranging and the system emits EM waves. The basic principle of the radar system is to find and position a target and determine the distance between the target and transmitting antenna. The transmitting antenna would be using RF. After transmitting radio waves, the radar’s objective is to listen for any returning signals which will then be analysed for any presence of a target, this principle is also known as echolocation. A target is detected when it reflects some portions of the transmitted EM wave. There can be multiple designs of a Radar system however all Radars will have the following key features and processes in common.

Fig 1.Key Elements and Processes of Radar Systems[1]

Figure 1 describes the key features and processes of a monostatic radar system. It is considered as a monostatic radar because the transmitter and receiver are positioned at one location. Whereas a bistatic radar is where the transmitter and receiver are positioned differently such that the angle and ranges to the target are adequately different. Referring to figure 1, a radar system can be divided into four major subsystems which are the antenna, transmitters, receiver, and signal processor.

The radar transmitter produces short-lived, high-frequency energy (megawatts) pulses that generate EM waves which are then radiated into the propagation medium by the antenna. The T/R switch also known as the duplexer, its purpose is to provide a connection so that both the transmitter and receiver can be attached simultaneously to the antenna and provide separation between the transmitter and receiver to shield the sensitive receiver components from the high-powered transmission signal. After transmitting the signal into the propagation medium, if the signal encounters a target (e.g. aircraft, missiles, clouds, etc.), the EM waves are reradiated into the environment. However, when the target reradiates, it will produce a diffused signal due to the scattering effect. This occurs when the transmitted signal gets reflected in several different directions. Aside from the scattering effect from the desired target; there can be other surfaces on the ground and atmosphere which can reradiate the signal. Such unwanted and unnecessary but genuine signals are called clutter interference, which can be picked up by the receiver antenna.

While the EM waves propagate through the atmosphere (propagation medium), the environment could alter the intensity of the EM waves both at target and at the receiving antenna. Therefore, the receiver is designed in such a way so that it is sensitive to the frequency spectrum being transmitted and provides amplification and demodulation of the received RF signal. After amplifying the received signal, the system transforms the RF signal to an intermediate frequency (IF), and then send the signal to an analogue-to-digital converter (ADC) and then to the signal/data processor, where the demodulation will occur. To sense a presence of a target at its greatest range, the receiver antenna must be sensitive and not introduce any unnecessary noise.

As discussed before that when a receiver antenna receives a reflected signal there can be a presence of clutter. However the interference does not only arise from clutter, it could come in the form of internal or external noise; unintentional external EM waves created by man-made sources, also known as electromagnetic interference; and intentional jamming from electronic countermeasures (ECM) system. These interferences could be in the form of noise or false targets. Thus, the processor needs to identify the target in the presence of noise, clutter or jamming.

Previously, it was stated that due to the scattering effect, the reflected signal is diffused indicating that the received power will be weaker than the transmit power. However, in addition to that one of the major feature of the radar that can affect the intensity of the signal is the distance between the radar and the target. Assume that an isotropic antenna is emitting EM waves, the intensity of the propagating wave is calculated by the power emitted per unit area of the wave. Knowing that the radar acts as an isotropic antenna, it can be shown that the transmitted power density decreases by Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://en.wikipedia.org/api/rest_v1/":): {\displaystyle \frac{1}{R^{2}}} :

Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://en.wikipedia.org/api/rest_v1/":): {\displaystyle Q_{t}=\frac{P_{t}}{4 \pi R^{2}}}

where,

  • Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://en.wikipedia.org/api/rest_v1/":): {\displaystyle P_{t}} is total radiated transmitted power
  • Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://en.wikipedia.org/api/rest_v1/":): {\displaystyle R} is the distance from the isotropic antenna to the target

The range of a target is determined by multiplying the time travelled, from RADAR to the target and back to the radar in the medium (𝑡), and speed of light (𝑐).

Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://en.wikipedia.org/api/rest_v1/":): {\displaystyle R=\frac{ct}{2}}

where,

  • 𝑅 is the slant range: the actual distance to the target and the radar
  • 𝑐 is the speed of light Failed to parse (MathML with SVG or PNG fallback (recommended for modern browsers and accessibility tools): Invalid response ("Math extension cannot connect to Restbase.") from server "https://en.wikipedia.org/api/rest_v1/":): {\displaystyle \approx 3\times 10^{8}m/s}
  • 𝑡 is the total time taken for the EM wave to travel to the target and back to the radar

Radar waveforms are generally classified under two classes: continuous wave (CW) and pulsed. The transmitter transmits a signal continuously with the CW waveform, usually without interruption, the whole time the radar transmitter is operating. The receiver also operates continuously. On the other hand, the pulsed waveform transmitter emits a sequence of pulses of finite duration, separated by the times in which the transmitter is "off." Since the transmitter is off, the receiver is on for detection of the target signals. For CW radars, it takes the bistatic configuration as the transmitter and receiver must be kept on constantly. In a CW radar, the round-trip time of the transmitted EM wave and hence the target range can be determined by alternating the composition of the wave (e.g.changing the frequency of the wave over time). This technique of FM effectively places a timing mark on the EM wave, thus allowing for the determination of the target range.

Pulsed radars would usually take on the monostatic configuration as the transmitter and receiver are not utilized together and are kept isolated from each other. The radar transmits EM waves over a short period or pulse width, typically between 0.1 and 10 microseconds (𝜇𝑠), but sometimes as few as a few nanoseconds (10−9𝑛𝑠) and as long as a millisecond. As discussed previously the duplexer will provide the transition to radar to switch from transmitter to receiver, to allow the antenna to listen to echoes. The total time of transmitting (pulse width) and listening to the echoes is the completion of one pulsed radar cycle, also known as PRI.[2]

Motivation

The problem of detecting small targets is challenging especially with a heavy clutter background. We aim to exploit waveform agility by tailoring the transmit waveform to have a second look on the target location to suppress the range bins with heavy clutter. An alternative method using mismatch filter was explored and implemented on the received signal to suppress sidelobes if waveform agility is not feasible.

The agile waveform would suppress range bins with high clutter while the mismatched filter would instead be suppressing overall sidelobes. From both implementations, we would aim to decrease the false alarm rate and improve detection performance.

Adaptive Waveform Design

With a high scanning rate and very short duration pulses, we assume the clutter returns are stationary over timescales of hundreds of milliseconds. With an ocean surface with a small target, we would be able to identify range bins of the strong sea wave corresponding to the target and perform suppression.

The diagrams below show the match filter output of the transmit waveform post suppression.

Discription Results
By increasing the number of chips, the degree of freedom increases. Therefore, the suppression performance increases.
The effect of number of chips to the suppression level.png
By shifting the suppression zone away from the mainlobe, the suppression performance increases as the initial energy of the range bins further from the mainlobe is lower.
The effect of the distance from the mainlobe to the suppression level.png
By increasing the suppression width, the number of range bins where the energy could escape to decreases, causing the suppression performance to decrease.
The effect of the number of suppressed range bins to the suppression level.png
The effect of having multiple suppression zones does not affect the performance.
The effect of having multiple suppression zones .png

Mismatched Filter Design

Phase coded waveforms generally produce high range sidelobes and to suppress the sidelobes mismatch filters are generally used. However, with the suppression of the sidelobes, it comes with a cost of loss to the main peak sensitivity. The following diagrams show the normalised results of the mismatched filter with changes in the parameters and compared to the matched filter outputs.

Description Results
Creating a mismatched filter of the same length, it can be shown that in order to suppress the sidelobes ( 15dB), there is significant loss in the main peak.
Matched and Mismatch filter same length.jpg
Increasing the length of the mismatch filter three times the original length. There is significant improvement in the suppression of the sidelobes (20dB) and there is negligible loss in the main peak.
Matched and Mismatch filter different length.jpg
With the previous conditions and allowing a loss in the main peak, it can be shown that it can suppress the sidelobes even greater ( 25dB) compared when there is no loss in the main peak.
Matched and Mismatch filter different length loss 0.3.jpg

Conclusion

In this project, the agile waveform allows the suppressed energy to be transferred to the other sidelobes which has weaker clutter response while the mismatch filter averages out the sidelobes at the cost of the main peak. The mismatch filter performs well when its length is greater than the length of the transmit waveform. We could identify and filter out specific clutter patterns using the agile waveform. However, the mismatch filter would be more suitable in conditions of distributed clutters with weak targets.

More work would be put into computing the computational costs and investigating the performance of both implementations in simulation and real life.

References

  1. ↑ M. Richards, J. Scheer and W. Holm, Principles of Modern Radar. Raleigh: SciTech Publishing, 2010.
  2. ↑ "Linear Frequency Modulated Pulse Waveforms- MATLAB & Simulink", Mathworks.com, 2020. [Online]. Available: https://www.mathworks.com/help/phased/ug/linear-frequency-modulated-pulse-waveforms.html
Retrieved from "https://projectswiki.eleceng.adelaide.edu.au/projects/index.php?title=Projects:2020s1-1231_Radar_Waveform_Design&oldid=15864"