Talk:Pulse-Doppler radar
This is the talk page for discussing improvements to the Pulse-Doppler radar article. This is not a forum for general discussion of the article's subject. |
Article policies
|
Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL |
![]() | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
|
Major rewrite
[edit]I performed a major rewrite to describe most of the topic involving Pulse-Doppler radar. Nanoatzin (talk) 10:46, 30 January 2011 (UTC)
I performed another major rewrite to clarify the article and shift ambiguity processing into separate articles that can be linked with sonar and other types of Doppler processing systems.Nanoatzin (talk) 12:05, 3 August 2011 (UTC)
"Doppler" incorrect
[edit]It is a common misconconseption that PW-"doppler" radar uses the doppler effect to measure object velocities. The doppler effect can only be used to measure average velocity across the entire beam, using continous-wave "pulses". PW-doppler instead measures the phase-shift between succesive pulses to determine radial velocities. Doppler_radar#Doppler_radar_as_weather_radar already mentions this. --Fredrik Orderud 01:18, 30 March 2006 (UTC)
I think that you are thinking of Moving Target Indication (MTI). This technique was the only way, for many years, to discern potentially 'real' targets from ground clutter. With sufficient precision, this can be used to resolve velocity, but the resolution isn't brilliant. True Doppler radars do evaluate the apparent shift in frequency brought about by moving objects. In the case of Doppler Navigators it is the Doppler shift from ground returns which is of interest, with PD, the returned echoes are processed using an FFT (or in older systems, banks of filters), to separate the frequency spectrum into discrete elements that correspond to opening or closing velocity. User:TJC
I've added a new discussion to pick up all threads of this at 'Gathering of Threads - The Nature of Pulse-Doppler Radar' below. --Terry C 11:11, 31 December 2006 (UTC)
Some pulse-Doppler RADAR are not capable of measuring velocity and use velocity only to tell the difference between jamming and real targets. Velocity measurement is not helpful except with weather phenomenon. Most pulse-Doppler RADAR use Doppler only to improve detection by improved filtering. 29 January, 2011 Nanoatzin (talk) 12:10, 29 January 2011 (UTC)
Page move/content replacement notice
[edit]I've taken the liberty of replacing the old content of this article with the old content of Pulse-doppler. This has been done because both articles discussed exactly the same subject, but the old Pulse-doppler radar was of very low quality, containing mostly vague and noninformative information. --Fredrik Orderud 21:05, 31 March 2006 (UTC)
This article really needs some work!
[edit]The current content might be a good start, but it is vague on the theory, and does not provide much insight into the underlying principles of Pulse-"doppler" radars. Some suggestions for improvements might include:
- Clearify that the "intermediate frequency" used, is simply a complex demodulated and downsampled variant of the received RF-signal.
- Clearify that the "velocity ambiguity" problem is an aliasing problem, encountered when the phase-shift caused by radial object velocity exceeds pi/2 radians between succesive pulses.
- Move all "AWACS" specifics/examples into a separate examples-section.
--Fredrik Orderud 21:20, 31 March 2006 (UTC)
- I've now rewritten most of the article in an attempt to clearify the underlying principles, as well as adding relevant details. The article does, however, still need some figures to illustrate how target movement leads to translation of the incomming RF-signal, which in turn leads to a phase-shift in the IQ-data, which finally is interpreted as target motion. --Fredrik Orderud 00:06, 1 April 2006 (UTC)
- I've included coverage of most of the theory of operation for medium PRF pulse-Doppler RADAR. It might be nice to include an update that describes how pulse-Doppler modifies the RADAR range equation. 29 January, 2011 Nanoatzin (talk) 12:14, 29 January 2011 (UTC)
Confusing Mixture of radar and sound
[edit]The formula given seems to be from the ultrasonography area of pulse doppler in measuring blood flow. "Packet size" and "sample volume" are other terms from blood flow and possibly weather measurement. The formula given seems more appropriate for blood or weather, than for aircraft detection. It's far too generic.
I think you can stay too long in the time domain. While it's true that pulse doppler uses time domain sampling, you can better describe it post FFT. Doppler filters is what gets processed by a pulse doppler radar. Another thing to consider in describing pulse doppler, is that the time samples processed can result in many filters having some magnitude (and not just one doppler for the "volume"). Thus any scan can contain many doppler detections, and usually does. k5okc 08:42, 2 April 2006 (UTC)
- You're right in that my background is from signal processing in medical ultrasound. However, the underlying principle for blood flow measurements and pulse-"doppler" radar is exactly the same; they both compare relative movement between succesive pulses. Both my formulas are therefore correct, also for radar applications. One should exercise caution with frequency-domain descriptions, since the method basically has nothing to do with doppler-shifts. Also, FFT is the least accurate algortihm for deriving velocities. The autocorrrelation technique is MUCH more acurate, so you should be carefull when using algorithm-specific expressions, like "post FFT".
- Feel free to replace some of my terms with more common radar-terms, but think twice before introducting "doppler" or "FFT" slang. The goal of the article should be to provide fundamental insight into the method in general, and not only a single possible algorithmic implementation. --Fredrik Orderud 10:56, 2 April 2006 (UTC)
- Real-world pulse-Doppler RADAR that are not used to evaluate weather phenomenon provide sound for the operator. 29 January, 2011Nanoatzin (talk) 12:21, 29 January 2011 (UTC)
Doppler Shmoppler
[edit]It's "Pulse-Doppler." Doppler radars like CW and FM are not pulsed.
- Some CW is pulsed to eliminate bleed-over between the transmit antenna and the receive antenna. This is called interrupted CW. Bleed-over from the transmitter into the receiver causes saturation, and that degrades range performance. Shutting off the transmitter during receiver sampling period eliminates saturation. Pure CW that is not pulsed requires an active canceler to null the bleed-over signal. Interrupted-CW and pure-CW do not attempt to resolve range by pulsing the transmitter so CW only takes one set of samples to produce a single spectrum whether or not the transmitter is pulsed.
- Pulse-Dopper uses multiple range samples to produce multiple spectra for multiple sampling intervals between transmit pulses. This allows a pulse-Doppler RADAR to resolve range. This is very different from just turning the transmitter on and off rapidly to avoid receiver saturation.
- Interrupted CW is not pulse-Doppler and pulse-Doppler is not interrupted-CW. Different kinds of RADAR.Nanoatzin (talk) 15:08, 29 January 2011 (UTC)
Merge with Doppler radar
[edit]While pulse Doppler radar works on phase shift an CW/FM work Doppler effect, both type of radars are called Doppler. Therefore it would be best to merge both article and explain completely the concepts. Pierre cb 14:02, 4 November 2006 (UTC)
I've added a new discussion to pick up all threads of this at 'Gathering of Threads - The Nature of Pulse-Doppler Radar' below.--Terry C 11:11, 31 December 2006 (UTC)
- Good luck with that merge plan. There are some cosmetic similarities between the names but the pulse-Doppler process used to identify true target range with intermediate PRF is completely different from every other type of RADAR. The RADAR equation for predicted detection range is not even the same. Merger would confuse readers that are not already familiar with RADAR.Nanoatzin (talk) 13:35, 29 January 2011 (UTC)
Working from a Negative
[edit]In the opening paragraph several people are wanting to start off with a negative. "Pulse-Doppler really isn't a Doppler Radar" or words to that effect.
This information is not "CRUCIAL" it is not very important. It's like saying Ice Cream really isn't vanilla. Who cares? 137.240.136.82 20:41, 6 December 2006 (UTC)
- No it is crucial as the name let assume what the process isn't! It is like talking about Tofu Ice Cream whitout mentioning it has no cream what so ever. One has to know that it is a misnomerPierre cb 22:11, 6 December 2006 (UTC)
Not only is this information not crucial, it is not true. This article is now completely wrong. I don't know what kind of radar it is describing, but it is not describing the PD radars that I have worked on. These are military attack radars and most definitely do use the doppler effect to determine velocity components from the returned echoes. Early PD radars (circa 1970s) had to resolve the velocity using filter banks, no mean feat where shifts from opening and closing velocities ranged from a few hundred knots to Mach 2 or above. I believe that Westinghouse achieved it first, but I may be wrong. I was told at the time that they had squeezed a 'roomfull' of filters into a 1 foot cube. User:TJC22 Dec 2006.
- Excuse-me but I think you are referring to a true Doppler radar even if it was using Pulses. I'm not in military and don't know what you describe, I am in meteorology, but if one use the variation of phase between successive pulse it is not a true Doppler. Pierre cb 15:29, 22 December 2006 (UTC)
There lies the problem; you are talking from a standpoint of working in weather radar and I am talking from a standpoint of working in military radar. However, one thing I am adamant about; in military attack radars the technology is called Pulse Doppler and does use the Doppler shift on the carrier, not the phase shift of the pulses. The term has been in use since at least the late sixties when the AWG 9, 10, 11 range of radars flew in the Phantom F4J and its decendants. Just because what I know as MTI, (see my other entry above), is known by you as Pulse Doppler, doesn't mean that all PD radars work that way. To put this into context, I followed the link to this article because I wanted to see if there was any up to date info on PD radar, since I hadn't worked on it since the late seventies. I started my search because I was trying to find out about the Italian attack radar called the Grifo S7 [1]. Others will follow my path, because they know that radars of this type are called PD radar, and may be totally confused to find that the article re-directs them to what they know as NAV radar. By all means write your article about weather radars, but please create a new one and don't distort the accuracy of this article which used to cover true PD radar. User:TJC 24 Dec 2006.
- In our discussion lies the confusion about the whole matter. Mesuring velocities with pulse or continuous wave radar is a usage, a principle, not a specific field of use. By what you say, different users name their radar different ways and that is very bad. For instance, originally the author of Doppler radar article was wrongly assuming that it was a weather radar article, a common misuse of the term in the American medias, and I was glad that someone eliminated that part. I personally wrote, a long time ago, a weather radar article which explains specifically about such radars, their use, the different types of data they produce, etc...
- The two articles, Doppler radar and this Pulse-Doppler, are describing the principles as I understand them and not a specific type of use. I did not create them, just added on to them. However, I think that the confusion will persist as long as there is a split among those articles. If you look at the top of the article, I suggested to merge the PD radar article with Doppler radar into a single article and elaborate more. That way, one would see about the different ways of using radar to mesure speed. In such an article, ALL variants from continuous to pulse emission and from mesuring Doppler shift to phase differential between pulses, could be discussed with referal to related articles for specific use. Pierre cb 12:46, 24 December 2006 (UTC)
I personally wrote, a long time ago, a weather radar article which explains specifically about such radars, their use, the different types of data they produce, etc...
Please think about this. If we wanted an article about the Doppler effect we would go and look at it. These articles, including yours, are about the application of a principle, not the principle itself. In the case of your weather radar article, you describe the techniques used in that application, referring to other articles to illustrate the details. The Doppler radar article does the same for CW radars and the Pulse Doppler article should also do that. This is how Wikipedia is supposed to work, using structured hyperlinks to build up a full understanding of the subject. Fredrik Orderud started trying to do that, but you have effectively obfuscated his work and created confusion with your erroneous statement that PD radars don't use the Doppler effect.
I wanted to learn about the evolution of Pulse Doppler radars since the 1970s. This is a specific application of the Doppler effect, has nothing to do with Weather radar or MTI, but is an extremely important area of radar technology. There are special issues with PD radar that do not occur with other radars, and for this reason alone, the subject deserves its own topic. What I or any other user shouldn't have to do is wade through pages of irrelevant information that is only applicable to NAV and weather radars in order to find out about modern attack radars. To do that is to miss the whole point of Wikipedia.
Please, I beg of you; restore this article to the way it was before you totally changed its message. You have acknowledged that you know nothing about military attack radars, so how can you assert that you are right about this? I have also acknowledged that I know very little about Weather radar, and wouldn't dream of trying to update your article. I'm not even an expert on radar technology as such, my experience lies more in the maintenance of such systems and the specification and implementation of test systems for radar components. This is why I am reluctant to try to update it myself. (After all, I originally came here to learn, not contribute to this particular area.)
So, please put the article back to the way it was and then let the peer group decide (over time) if it contains useful information or not. After all that is how Wikipedia is supposed to work. If more expert people than I agree with you, then I'll shut up. User:TJC 25 Dec 2006.
- As I said, I did not create the article, I just added on it! Look at the versions before november 2006, for instance the one from 31st of march 2006, by Orderud who says that he redid the intro:
- "It is a common misconception that pulse-"doppler" radars utilize the doppler effect to calculate velocities, since it is considered impossible to measure the tiny frequency shift originated from targets moving at sub-sonic speeds when excited by very brief radar pulses. Velocity measurements are instead made possible by transmitting several radar pulses towards each target over a very short period of time, and measuring relative target movement between each pulse. The number of pulses used are usually referred to as packet size and the frequency in which they are emitted as pulse repetition frequency (PRF)."
- As I said, I did not create the article, I just added on it! Look at the versions before november 2006, for instance the one from 31st of march 2006, by Orderud who says that he redid the intro:
- As you see, this article was created saying the same thing as I rewrote: PD using the phase shift not the Doppler shift. I did not change the premise. However, there was a merge just before that between two articles PD and PD radar, look at the history, the latter was more your idea on the subject. The discussion we have does not lead anywhere as I am not the one who created it the way it is now. If you feel so strongly about your point, contact Orderud and discuss with him where he got a different interpretation than yours of what is a PD, or CHANGE THE ARTICLE yourself. I cannot say more than that.Pierre cb 20:55, 25 December 2006 (UTC)
This is a mess. I originally responded to your comment that it was important to make it clear that PD didn't use Doppler (6th Dec). At the time, I was making no judgments about who originally changed it. In the ensuing exchange, you argued so strongly for the non-Doppler theory, I assumed that the change was yours. Re-reading the trail, I see that Fredrik Orderud also says that he knows nothing about this particular topic; his expertise lying in medical ultrasound, but asserts that makes no difference. However, if you look carefully at the maths that underlies the theory, you will see that the carrier frequency is needed to resolve the velocity. This would be irrelevant if the phenomena relied solely on pulse arrival times. I will restate, despite my own limited expertise, that I am extremely confident that the Doppler shift on the carrier is the key parameter.
Needless to say, I am sorry for sounding off at you about changing the article; I was targeting the wrong individual as you have pointed out. I will have to look at this a bit more to find out where the reality left off and how much of what remains is accurate. The trouble is, I am not a radar designer, as I stated above, so feel unqualified to update the article with any confidence that I won't drop yet more clangers. User:TJC 25 Dec 2006.
- Estimation of target velocities, using pulse-doppler radar as currently presented in this article does NOT depend on any doppler-shift in the received signal. Instead it is based on changes in target distance between pulses, which translates into phase-shifts between returning pulses. Utilization of the carrier frequency in the formula to determine target velocity is still required, since it relates target movements to phase shifts; but this has nothing to do with (real) doppler-shifts. --Fredrik Orderud 02:54, 30 December 2006 (UTC)
I've added a new discussion to pick up all threads of this at 'Gathering of Threads - The Nature of Pulse-Doppler Radar' in the next discussion. --Terry C 11:13, 31 December 2006 (UTC)
Gathering of Threads - The Nature of Pulse-Doppler Radar
[edit]Introduction & Summary I have started this new discussion as a continuation of three others about this article; namely '"Doppler" incorrect', 'Merge with Doppler radar' and 'Working from a Negative', (particulary the latter). I have done this because the main thread of the discussion in 'Working from a Negative' moved away from its original point towards '"Doppler" incorrect' (my fault), and also the request to merge with Doppler radar is relevant to the overall point. Before I pickup the discussion, I will provide a brief summary of the current situation.
Several people have asserted that Pulse Doppler radar does not use the Doppler Effect to resolve velocity. In the Introduction to the main article Frederik Orderud wrote earlier this year that:
- Pulse-Doppler radar does not use the doppler effect on each returning pulse to determine velocity, contrary to a non-pulsed Doppler radar. Instead velocity measurements are made possible by transmitting many radar pulses (due to a high PRF), and measuring the relative phase shift between each pulse returning from the target
If true, then the suggestion to merge the article with Doppler radar makes no sense, if untrue, then a merge might be in order, although I still feel that the topic deserves its own article for reasons that I hope to show later.
- Kindly read my additions describing how pulse-Dopper systems resolve range before considering merger with pure-Doppler RADAR.Nanoatzin (talk) 14:36, 29 January 2011 (UTC)
Moving on to the underlying principle of Pulse Doppler (PD) radar; I happened across this article when I was trying to bring my own knowledge of the subject up to date. Having spotted Frederik's statement regarding how PD radar works I felt that it was incorrect, based on my experience with PD in the seventies. To date however, I seem to be in a minority of one, but still believe that I am not wrong.
Doppler Effect or not? In the last statement made on this topic on 30 December, Fredrik said:
- Estimation of target velocities, using pulse-doppler radar as currently presented in this article does NOT depend on any doppler-shift in the received signal. Instead it is based on changes in target distance between pulses, which translates into phase-shifts between returning pulses. Utilization of the carrier frequency in the formula to determine target velocity is still required, since it relates target movements to phase shifts; but this has nothing to do with (real) doppler-shifts.
I will state straight away that when I read the original statement in the Introduction to the article, I interpreted it to mean that the 'relative phase shift between pulses' refered to the pulse waveforms not the carrier, eg movements in the leading edges of the returned echoes. I believe that most people would, because it is valid way to determine velocity and is used in Moving Target Indication (MTI) radars. However, that isn't how PD radar works.
Returning to Fredrik's most recent statement, he restates the point, but then says that 'Utilization of the carrier frequency in the formula to determine target velocity is still required, since it relates target movements to phase shifts'. This is how PD radar works and the phenomena is caused by the Doppler effect
Frederik has also added some references to the main article; a presentation and a handout from the University of Iowa on Doppler radar principles. I think that examination of these items shows that the phenomena is consistently refered to as Doppler, and PD radars are specifically mentioned without any qualification regarding the name being a misnomer. I do note some ambiguities however, when discussing the phase shift, which could lead to misunderstandings.
Being a practical engineer and not a mathematician, I cannot prove this mathematically. However, I hope to be able to show it theoretically. If you read the excellent description of the principle in Doppler effect, you will see that a tennis ball analogy is used to explain the apparent change in frequency. In a Doppler radar the principle can be visualised by imagining that the echoes from a closing target are the tennis balls being thrown at the receiver. However, remember that in a radar echo the pulses do not actually exist; they are simply the envelope of the modulated carrier which is extracted in the receiver. I think of the tennis balls as phase angle zero of each carrier cycle, (they can be any point on the waveform, as long as the same point is imagined each time). It can be seen that each zero crossing point will apparently arrive a little earlier than expected, due to the relative movement of the target and receiver.
This is the Doppler Shift and it applies whether the radar is CW or pulsed. In order for this to work, it is critical that the returned echo is phase related with the transmitted signal. In the case of CW radars, this is relatively easy; the returned signal is mixed with a pick-off from the transmitter and as long as the phase stability of the transmitting device is reasonable, (eg stability over a few milliseconds), the result is the Doppler Shift. With PD radars the transmitted signal is pulsed and it isn't so easy to obtain a stable transmitter sample. To achieve this, a stable local oscillator (STALO) and special modulators are needed and the radar is known as coherent.
So should this article be moved or not? As I have already stated, if my assertion remains unproven, then clearly moving the article makes no sense. However, if the consensus is to accept that the Doppler effect is central to PD radar, then there might be some argument for moving the article. I would oppose that too and here are my reasons:
- The two systems, (CW and PD) are used for totally different applications (I touched on this in an earlier discussion (probably over at Doppler radar).
- The problems of Velocity ambiguity (aliasing) and Range ambiguity do not occur with CW radar. (There is some discussion about this in the article).
- CW radars do not have the same problems achieving coherency.
Sorry for being long winded, but I feel that it is important that this is fully understood by everyone. User:TJC 11:01, 31 December 2006 (UTC).
Argument against combining
I think the theory we want to project, is that a PD radar does not measure a Doppler shift in real-time. A PD radar sends out a pulse spectrum but cannot process that spectrum in real-time. A PD radar works in the time-domain and cannot measure doppler frequencies pulse-to-pulse. The radar converts the time-domain into the frequency-domain using an FFT for multiple targets, or autocorrelation for averaged targets over a "look."
Pulse-Doppler was coined to signify a velocity capable coherent pulse radar. The term Doppler here isn't how the radar determines velocity, but rather that it outputs information that contains velocity information.
While the first PD radars used analog filter banks to sort velocity information, they were quickly replaced by digital doppler processors, or what is known as a 1D FFT. The E-3 for example used a non-real-time 128 filter FFT by 50 range gates. So it took 128 (or more) transmitter pulses, and 50 range gates per pulse before it output a Doppler velocity. This Doppler velocity was a transform of the time-domain that is causing the confusion here. PD radars do not measure Doppler directly. That is the bottom line we would like to have in the end. 137.240.136.81 16:36, 10 January 2007 (UTC)
- Thank you. That was another reason for not combining that I hadn't thought of. There now seems to be no dissenters, so I think we can assume that we have a consensus for retaining this article on its own page, so I will remove the merge proposals both here and at Doppler radar sometime during this weekend. We've also had no new arguments to support the 'PD is not Doppler' theory, so I propose to remove that reference from the introduction. If there is no-one able to develop the article further, I'll see if I can obtain to reference material and do it myself, although I still feel a little under qualified to do it. --Terry C 13:23, 12 January 2007 (UTC)
- It's hard to develop the article, because the weather and health experts demand equal time. If you use the word FFT they will come out of the woodworks and demand equal time and delete your contribution. 137.240.136.82 14:39, 18 January 2007 (UTC)
Actually
[edit]We are talking about a mathematical process. The phase shift would not exist if the doppler effect did not exist for photons—any frequency. The phase shift is a physical reality and evident on a per packet basis. It is wrong "physics" to say that it is based on target distances, because you do not compare the returning pulses to one another. The original signal is compared to a returning pulse-pair (unfortunately creates a small amount of error but gets the job done). Also, we are talking about extreme-sub-light-speed motion that is causing the phase shift. However, you cannot get a full spectral shift from pulsed, which is where the Nyquist interval comes in handy....agreed? Great! — HRS IAM 01:41, 12 August 2007 (UTC)
- It is just plain wrong to assert that phase shift would not exist if the doppler effect did not exist! The phase shift observed between successive pulses originates from the fact that the distance to the target (measured in wavelengths) has changed (by a fraction of a wavelength), and not from the fact that the target's motion affects the reflected pulse in any way. The exact same phase shift would in fact be observed if the target was stationary at the time of both pulses, and only moved momentarily in the time-slot between the pulses.--Fredrik Orderud 13:25, 12 August 2007 (UTC)
- Actually, The phase shift observed between successive pulses originates from the fact that the distance to the target (measured in wavelengths) has changed (by a fraction of a wavelength) is the Doppler effect!--Terry C 08:17, 16 August 2007 (UTC)
- Nope. The doppler effect has to do with changes to the pulse frequency, which might occur independently of any phase changes. If, as in my example above, the target is stationary at the time of both pulses, then no doppler effect will occur what so ever when the pulse is reflected by the target. A phase shift will nevertheless occur if the target has moved in the time-slot between the pulses. --Fredrik Orderud 14:49, 16 August 2007 (UTC)
- A frequency shift is just another way of expressing a phase shift. When a target is moving at a velocity greater than that which would result in a 360 degree phase shift, (the norm in radar situations), then the phase shift is also a frequency shift. At lower velocities in the radar situation, then only a sub-Hz change is resolved and the velocity would probably be considered too small to concern anyone. However, in medical ultrasound, or other applications where both the carrier frequency and the velocity is low, then the Doppler shift is correspondingly reduced. In this case it is observed as a phase shift and is this that must be resolved, if the velocity is to be determined. The applications may be different, but the mechanism used is still the Doppler effect. --Terry C 18:37, 16 August 2007 (UTC)
- If you look at the figure below, I've tried to illustrate how phase-shifts between successive pulses occurs independently of any change in pulse frequency due to the doppler effect.
- Phase-shifts does, however, lead to frequency shifts when performing frequency-analysis on the phase of successive pulses, but this frequency shift has absolutely nothing to do with the doppler effect known from physics (although they can both be used to estimate target velocities). It is therefore important to separate changes in radar pulse frequency (due to the doppler effect) from changes detected in frequency spectrum when comparing the phase of successive received pulses. --Fredrik Orderud 19:40, 16 August 2007 (UTC)
- Fredrik; I'm not convinced that we're ever going to agree on this point. I've tried to expand my viewpoint in the discussion below under your diagram. I am now more than convinced that this article still contains some fundamental errors, which are contributing to the confusion. I've only come to realise this as I thought about your arguments below.
Phase shift vs. Doppler effect
[edit]The above figure tries to illustrate how target motion induces phase-shifts in the received pulses that are reflected by a moving target. This phase-shift is not dependent of any shifts in pulse frequency, and is therefore not to be confused with the doppler effect.
A doppler effect might also be present if the target is moving at the time of the pulses, but this effect is believed to be negligible due to the small velocity of airplanes compared to the speed of light. Any change in pulse frequency due to the doppler effect will also be canceled out when comparing successive pulses, rendering frequency shifts undetectable. --Fredrik Orderud 15:41, 16 August 2007 (UTC)
- As I've tried to explain above, it doesn't actually matter if the target movement changes the phase or the frequency, since the physical result is the same. There are two things to consider here. First the pulse frequency is irrelevant. (Here I think the formula given in the article is wrong, the PRF only affects the maximum velocity that may be resolved, not the velocity calculation. I will start a new discussion on this.) Remember too that the pulse doesn't actually exist in the transmitted signal; only bursts of carrier are sent and the pulse envelope is extracted from the received single to form the pulse so it's the carrier phase that is of interest. The second point to consider is that it doesn't matter whether the target is moving at the time that the pulse impinges, or simply moves during the interval between pulses. It is the perceived change that is resolved at the receiver; there is no actual change in frequency (or phase shift).--Terry C 10:21, 17 August 2007 (UTC)
- My point is that you should not use the term doppler effect in this article, since PW-radars does not depend on the doppler effect known from physics. The doppler effect has to do solely with changes in pulse frequency upon target reflection, which you self claim is irrelevant for PW-radars. Encyclopedic articles should strive towards correctness, and not misuse laws known from physics in inappropriate places. Instead of "doppler effect" one can use "perceived doppler effect" or similar to illustrate that target motion does indeed induce a frequency shift when frequency-analysing the phase of successive pulses, that often is wrongfully mistaken to be caused by the doppler effect. --Fredrik Orderud 12:46, 17 August 2007 (UTC)
- I'm sorry. I cannot agree with you. The technology is called Pulse-Doppler because it uses the Doppler effect not some pseudo equivalent. Please refer me to an independent source that supports your assertion. Every reference that I have ever found; including the ones I believe you yourself supplied for this article some months ago, refer to the Doppler effect and none mention that it is a misnomer. The phenomena that you have spent many hours describing is the Doppler effect; I believe that you are simply applying the wrong interpretation to the facts.--Terry C 13:01, 17 August 2007 (UTC)
- Quote from Marzetta, Martinsen & Plum "Fast pulse Doppler radar processing accounting for range bin migration", IEEE National Radar Conference 1993: "It has been pointed out [7] that this true Doppler effect plays no role in the function of either a pulse Doppler radar or a SAR; if the target came to a halt during the reflection of the pulse, the discontinuous motion of the target would still give rise to a phase shift from pulse to pulse, and either type of radar would still function perfectly" (ieee doi, pdf)
- Usage of "apparent Doppler" for pulsed-radar: [2]. --Fredrik Orderud 19:54, 19 August 2007 (UTC)
- I still can't agree with you. Unless I've misunderstood the paper, you are misinterpreting statements made to justify a way of evaluating Doppler velocity as evidence of your assertion that the effect isn't the Doppler effect at all. If you look at my page Radar signal characteristics, you will see that I attempt to explain unambiguous velocity and I also reproduce a spectral diagram from one of the references I give which clearly shows the Doppler shift in the targets. Unambiguous velocity only occurs in a pulsed system and is exacerbated by the problem of unambiguous range forcing radar designers to resort to very high PRFs. The net result of this is that echoes from any particular target cannot be received before the next pulse is transmitted and even if they were, they wouldn't be a continuous Doppler signal (what your reference calls true Doppler). They are in fact a burst of carrier which exhibits the same compression or expansion of wavelength as if they had been part of a CW signal. It really doesn't matter whether the carrier is CW or only a few nanoseconds long, Doppler shift will occur on reflection (and also at the radar antenna if it is also moving, once on transmit and once on receive).
- I'm beginning to question your assertion that a PD radar would see a movement between pulses as a velocity. I don't think it would, but I don't think it matters, because the carrier frequencies and PRFs in use would make that particular case of no interest to the radar operator. I'm also beginning to suspect that the problem here isn't related to the phenomena, but to the way that the radar receiver perceives the returns. If the radar had been a CW system getting returns from a target with the stop/go velocity pattern, then the instantaneous velocity values would show up the erratic nature of the target. In fact though, the user would be unlikely to see anything other than an average velocity because in any practical system the displayed velocity is integrated over a period of time to reduce errors. In the PD system, there is no choice; the returns have to be integrated because only a few tens to a few thousand cycles of carrier are available to work with from each individual echo. As a result the velocity is computed from successive returns from the same target. To achieve this the radar system must be coherent and use correlation techniques to ensure that the right pulses are integrated to give the target velocity. That is I think, what your first reference is talking about; there a lots of ways of achieving correlation. The radar signal processing circuits, whether they be analogue filter banks or FFT plus mathematical computations have to sort the Doppler shift into closing and opening velocity. In a practical system, the jittered PRF necessitated by the high PRF simply adds complexity to the problem.
- If possible check the references that I give on that page (unfortunately had to get the dead tree versions from the IET Library), and look at your own references again. Those extracts do still refer to the phenomena as Doppler shift and the formulae given for range rate and/or Doppler shift do not contain any component of PRF. The standard formula for Range Rate in a PD system still only contains frequency shift and wavelength as variables. In the paper, they are using a correlation technique that ties the echo to its delay from the main transmitter (not the PRF), but there are plenty of other ways to do it. I will say that their method is probably convenient where signal processing circuits are being implemented anyway. I would also like to add that with the PD radar that I worked on in the 1970s, a typical Doppler shift for a rapidly closing target, eg an incoming aircraft, was of the order of 5kHz. —The preceding unsigned comment was added by TJC (talk • contribs) 20:03, August 20, 2007 (UTC).
Is the formula in this article wrong?
[edit]The formula given under the heading 'Underlying Principles' contains PRF as one of its terms. Having been involved in the discussion with Fredrik on the nature of the Doppler shift I've begun to be concerned that this is the wrong formula. I'm a practical engineer and not a mathematician, so do not feel confident to comment on the maths, but I had never understood that PRF affected the velocity calculation.
- The formula to find velocity using frequency shift is Va = - 0.5 x Fd x C / Fo, and this can be derived from the information provided in the article on Doppler effect. Va is the ambiguous velocity. Fd is the Doppler frequency. C is the speed of light. Fo is the transmit frequency.
- The issue with this formula is that it provides the ambiguous velocity, where other possible velocities are above or below that one by Vb = 0.5 x PRF x C / Fo (Vb is the first blind velocity).
- The true target velocity is Va + N x Vp, where N is a positive or negative integer that must be found during the detection process.Nanoatzin (talk) 15:45, 29 January 2011 (UTC)
Looking at the references given in the article, this Doppler radar presentation would indicate that the formula for velocity is much simpler, but again, my mathematical abilities leave me uncertain about what is being proved.
I think that the formula given in the article is actually the one for maximum unambiguous velocity, which is dependent on PRF.
Can anyone else help?--Terry C 10:42, 17 August 2007 (UTC)
- I don't think the formula is wrong. I'll try to derive the formula for you below:
- Relations known in advance:
- , where Ts is the time between successive pulses
- , which is the relationship between wavelength, speed of light and pulse frequency.
- The phase shift can be expressed as a fraction of a distance in wavelength based on target movement length L (see figure above):
- .
- By solving with regards to target movement length (L), we get:
- Target velocity can then be calculated by dividing the movement length (L) by the amount of time passed (Ts):
- Which is the same formula as the one in the article. Please note that none of the equations above depend on any frequency shifts in the pulse frequency caused by the doppler effect. --Fredrik Orderud 13:15, 17 August 2007 (UTC)
- OK. As I said, I'm no mathematician, so can only take your word for it, but I cannot see how you can assert that Ts, the time between successive pulses, is needed to resolve velocity. That is simply based on your own statement that this is not a Doppler effect. I cannot say how modern FFT based radars achieve their result, because I haven't worked with them, but the old analogue filter bank systems were simply that; a bank of filters overlapping over the spectral equivalent of the radar's velocity range. The only effect of PRF was to stimulate the filter regularly and thus perform a form of integration. The PRF was not a part of the velocity resolution. I would have thought that the modern FFT simply performs the same function.
- Is there anyone other than myself and Frederik who can add to this?
Pulse-Doppler processing implies one is transmitting and receiving a pulse train within a dwell, cpi, etc. Doppler shift can indeed be determined by analyzing the delta phase changes across the SLOW time of the received pulses (i.e., all samples at range bin XYZ where matched filter detection occurred). Pulse-Doppler processing does not work without per pulse delta phase changes within a pulse train. Performing FFTs in the slow time direction allows one to estimate the doppler shift to the nearest bin.
http://users.ece.gatech.edu/~mrichard/Doppler%20and%20StopHop.pdf gives a great, quick and concise overview.
I would also highly recommend the book, "Principles of Modern Radar: Basic Principles" by Richards, Scheer and Hold. — Preceding unsigned comment added by Thrakkor (talk • contribs) 23:22, 23 November 2016 (UTC)
Typography, etc.
[edit]Allways use \text{} to wrap any text in LaTEX equations. If you just type it without, LaTEX math mode will think that the letters of every word are separate variables and
- typeset them in (math-)italics, where upright roman should be used
- increase the spacing ("multiplication space") between the characters ("variables") evenly
Please use the proper symbol for the "speed of light", i.e., a lower case c.
Range and velocity resolution equations seem incorrect
[edit]I agree that range resolution, conceptually, is related to the spacing of time samples, but fundamentally I don't think it has anything to do with the PRF. According to this, if you have a really low PRF, then your range resolution increases (worsens)? But that's totally missing the point. You can have a low PRF but also have a really short pulse (low duty cycle on the radar transmitter). The short pulse is ultimately what determines range resolution, not the number of pulses you sample. You can have a really large number of time (range) samples, but if you're using long pulses, then returns from objects that are closer together than are just going to smear together. You can't resolve them. Radar range resolution is fundamentally related to a short pulse, , with range resolution equal to . You can use a short pulse in a high-PRF radar or a low-PRF radar.
Furthermore, almost any pulse Doppler radar is also going to be performing pulse compression (time compression) with a matched filter so that you can use longer pulses to get higher peak power while still maintaining fine range resolution. After pulse compression, the profile has a width where is the bandwidth of, for example, a linear FM chirp pulse. When doing pulse compression, range resolution is now defined by .
For velocity resolution, it looks like the equation shown is just taking the filter size from a "filter bank" and converting that to a velocity using the Doppler equation. It's almost correct, but is not described well, nor is it related to the important aspects of pulse Doppler. After the FFT operation of pulse Doppler processing, you get a peak in Doppler frequency with a width of , where is usually the number of pulses in your coherent processing interval (CPI), but sometimes can be larger if you zero-pad the FFT to estimate finer Doppler resolution. To convert to velocity, this equation listed is missing a factor of two. . So your velocity resolution is .
Willkeim (talk) 12:04, 14 November 2015 (UTC)willkeim
External links modified
[edit]Hello fellow Wikipedians,
I have just modified one external link on Pulse-Doppler radar. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
- Added archive https://web.archive.org/web/20090227021024/http://www.es.northropgrumman.com/solutions/awacs/assets/AWACS.pdf to http://www.es.northropgrumman.com/solutions/awacs/assets/AWACS.pdf
When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.
This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}}
(last update: 5 June 2024).
- If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
- If you found an error with any archives or the URLs themselves, you can fix them with this tool.
Cheers.—InternetArchiveBot (Report bug) 03:58, 22 May 2017 (UTC)
External links modified (January 2018)
[edit]Hello fellow Wikipedians,
I have just modified one external link on Pulse-Doppler radar. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
- Added archive https://web.archive.org/web/20110101200238/http://www.radartutorial.eu/18.explanations/ex17.en.html to http://www.radartutorial.eu/18.explanations/ex17.en.html
When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.
This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}}
(last update: 5 June 2024).
- If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
- If you found an error with any archives or the URLs themselves, you can fix them with this tool.
Cheers.—InternetArchiveBot (Report bug) 10:50, 23 January 2018 (UTC)
Lock criterion or lock criteria?
[edit]The recent edit made by Cyfal to change 'lock criteria' to 'lock criterion' is grammatically correct, but doesn't appear to accord with common usage on the subject. The few links that I've found appear to use 'lock criteria' as per the original text and the linked reference within the section provides a formula to determine the 'Lock criteria'. As it stands, the term would still be wrong since that formula only seems to be based on that single calculation. However, I'm wondering if that formula may in fact be an over simplification of what goes on in practical systems and in fact there may be several criteria and that any or all have to be satisfied.
My experience of Pulse-Doppler radar dates back to the early 70s, so I do not feel qualified to do anything more than ask the question. I do however feel that the plural version should be used if that is in common usage in the Pulse-Doppler radar community - even if it is grammatically incorrect.
Anyone able to clarify this? Terry C (talk) 09:22, 17 October 2021 (UTC)