Not sure if this is the proper forum to ask this question, but still I am trying my luck.

I am looking for an application (in any area, medical, defence, or any other area) if exist, such that we want to dectect something by sending a waveform and get a retun on a linear array and determine the exact location of the target from the return signal. The catch is due to the situation or the physical limitation we cannot use wideband waveform. (We know that linear FM or wavefrom with higher bandwidth will help in spatial resolution, but we cannot use it due to some constraint.) In addition, the sampling rate cannot be increased so that we cannot form correlation matrix for better spatial resolution. This may also mean that target changes fast and we must rely only on time snapshot.

It appears in such case we stuck with Fourier beamforming.

The question is: Is there any such application where the above constaint really exists.

Any ideas will be highly appreciated.


These constraints absolutely exist. There are the norm! We could only wish in our wildest dreams to use as wide as bandwidth as we like. There are many areas in a radar system that place limitations on how wide the bandwidth can be and we'll go over a few straight forward ones. Mainly we're talking about limitations due to the antenna and waveguide as well as signal processing requirements.

Antenna and Waveguide

With antennas, especially arrays, their geometries can be defined in terms of wavelengths. Given that wavelength, and therefore frequency, we can yield the antenna's directivity pattern. It's important to stress that when you typically look at an antenna pattern, it is for a single RF frequency only.

Even simple pulses have some bandwidth around the RF frequency but are usually not too large. This allows us to make the narrowband approximation that the antenna pattern is the same for frequencies near the RF frequency, which is how the pattern was defined.

With an exceptionally wideband waveform the frequencies transmitted are now farther apart and so the antenna pattern itself can change significantly intra-pulse. These affects are more extreme at lower RF frequencies where desired bandwidths are well within the order of the transmitted frequency.

These effects are undesirable because (including but not limited to):

  1. The effective radiated power is no longer the same during the pulse.
  2. Because the antenna pattern itself is changing during the pulse, it distorts the waveform itself and you incur SNR degradation as well as undesired phase behavior which may lead to other issues, such as errors in angle measurements.

The waveguide itself, as well as other receiver/exciter components, need to support the bandwidth of interest and thus will require the use of certain technologies and processes to support it. This is a nice way of saying that you're going to be spending a lot of money and making sacrifices in size, weight, and power (SWaP).

Signal Processing

Assuming that your RF hardware is taken care of, there are issues with data collection and front-end signal processing. High bandwidth waveforms require high sampling rates. Introducing higher sampling rates has the following overarching issues

  1. High sampling rate analog-to-digital converters (ADCs) can be prohibitively expensive...for a good one.
  2. Depending on the bandwidths and pulse widths desired, you may have to deal with a large number of samples. In critical applications such as a fire control radar your computing resource (usually FPGAs/CPUs) need to be able to take care of common tasks such as pulse compression, FFTs, memory copies, etc. within a very strict time schedule.

These are just a few examples. I hope it's clear that the narrowband restriction is many times the norm when designing a system. Thankfully, technology is advancing where we can get away with higher bandwidths easier than we could in the past.

| improve this answer | |
  • $\begingroup$ Thank you for you answer. Actually I am interested to know if the application you mentioned require to detect target? Does this aplication use Fourier Beaforming to detect target in spatial direction? $\endgroup$ – Creator Jul 23 at 23:00
  • $\begingroup$ @Creator Radar systems are used because they are trying to detect something, so yes. What I've described are general considerations and are independent of digital beamforming via FFT. Using the FFT for beamforming gives you other options, especially when using wideband waveforms since the result can be used to detect a peak angle of arrival at the cost of calculating the FFT. $\endgroup$ – Envidia Jul 24 at 0:37
  • $\begingroup$ I am interested in the specific applications as mentioned in the question, rather than possible scenerios. Where these waveguide as you mentioned is used, what kind of product etc. Any comment please. $\endgroup$ – Creator Jul 25 at 6:56
  • $\begingroup$ @ Creator First let me comment that when I say "waveguide" I am also including any other type of transmission line. These can include microstrip, stripline, cables, etc., which are all used to transfer electromagnetic energy. These are ubiquitous and are used in virtually all RF applications. $\endgroup$ – Envidia Jul 25 at 20:37
  • $\begingroup$ Thank you for your replies. The issue is, I am from signal processing area and I want to know where people use Fourier beamforming specificaly in product. By product I mean ultrasound machince which scans human body etc. When you say transmission of energy it is a "work" not a product. Sorry for my questions but I really want to know. $\endgroup$ – Creator Jul 25 at 20:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.