2

I want to understand what Signal Processing experts do in their professional work, in essence.

I was thinking about:

  • Sampling signals
  • Receiving signals by machines
  • Broadcasting signals by machines
  • Creating mathematical models based on signals

What do Signal Processing experts do, in essence?

Semo
  • 29
  • 3
  • This is a very broad question, can you narrow down exactly what you’re asking? And what do you mean by “limits of the field?” – Ryan Aug 18 '21 at 02:54
  • @Ryan I have edited the question to further narrow it down and I have also removed that phrasing you mentioned. – Semo Aug 18 '21 at 05:01
  • I upvoted the answer, which is far better than the vague listing. – Ed V Aug 18 '21 at 12:14

1 Answers1

6

Signal processing is often at a junction between applied statistics, numerical math, analytic math, physics, electronics and programming. Big data, visualization and machine learning may be overlapping skills. Depending on interest or requirements of your employer, you could end up designing new modem (algorithms) destined for an fpga, noise reduction algorithms for speech running on smartphones, or a new radar system for looking into the ice cover of one of the moons of Jupiter.

You could design and work with the low level circuitry of asic design, write assembly code for a quirky micro controller, spend your days writing high level numpy code or do calculations and white board design of systems for others to implement.

Knut Inge
  • 3,384
  • 1
  • 8
  • 13
  • Thank you. Did you find something essentially wrong in what I wrote in the body of my question (in the 4 paragraphs)? – Semo Aug 18 '21 at 07:38
  • @semo what you wrote is at best a drop in the ocean. It's not wrong, these are some things you do in signal processing, but if you'd use these four points only to describe what signal processing is, then you'd exclude a majority of things that are signal processing. Knut's answer is closer to a sensible description. – mmmm Aug 18 '21 at 13:41
  • 1
    This is a really good answer. I would say most signal processing engineers do math and write code, in MATLAB, R, python (numpy), C, C++, and the assembly languages of various DSPs and embedded processors. Some code in FPGAs which is getting closer to hardware design. – robert bristow-johnson Aug 18 '21 at 16:09
  • @semo I would not say that «Sampling signals» is descriptive of what I (believe) that most dsp people do at work. Yet I would say that Nyquist sampling theorem and time/frequency analysis are perhaps the most useful bits of knowledge that I got in university. – Knut Inge Aug 18 '21 at 20:53
  • @mmmm thank you for confirming that what I wrote isn't wrong; I just comment on the phrasing "a drop in the ocean" --- I find it very exaggerated; I ought to find what are the most fundamental things any SP expert did (in the limits of the field) any time, even before programming languages as we know them today existed; I am well aware I missed some practical actions Knut described in his answer. – Semo Aug 19 '21 at 00:57
  • @KnutInge the main point I gather from your answer (which was very nice for me to read) is that an SP expert, at least nowadays, is pretty much an algorithms programmer. – Semo Aug 19 '21 at 05:05
  • 1
    @Semo, I think that is a decent summary for what many DSP experts do. Using and programming computers are basic to many lines of work today. There are many algorithms out there, and DSP people may be interested in a subset of them. While I have had the honour of working with CompSci people that knew a lot more than me about sorting algorithms and databases, I may be better at explaining the reverse wagon-wheel effect seen in cowboy movies, how to demodulate a 5GHz-centered radio signal, how exponential smoothing relates to an iir biquad filter and why filterbanks are used. – Knut Inge Aug 19 '21 at 06:47