# Overview Noise Reduction Techniques

I am trying to learn signal processing and when I research about ways to reduce noise in different situations it's so easy to get lost in one solution or go down the wrong path

for example, is your noise correlated? Is it not? Is it from multiple sources? Is it not?

I need to get some intuition on common solutions to these scenarios, but I too easily get lost in the massive amount of really specific research papers

Can someone point me to an overview of these techniques and when to use them etc...? Any help would be useful at this point

• Note that suggesting a suitable technique requires us to know what kind of signals you are trying to denoise. A denoising algorithm that works well for images will have almost nothing in common with one to denoise audio signals or RF data. – Florian May 14 '19 at 12:49
• Yeah, this is exactly the problem, thats another degree of freedom, how can I get an overview of the whole topic rather than getting lost?. How did you learn this topic I suppose is really my question? – Stephen Jackson May 14 '19 at 13:03
• One learns best by doing. Intuition comes from experience. The concept of noise is deeply tied to the concept of probability. You might find some value from books like Savage or Jayne. – user28715 May 14 '19 at 14:08
• @StephenJ: one could go so far to say that it is not the right approach trying to learn this in general. Instead, it would be better to take a few concrete examples and get your hands dirty. Otherwise it's like trying to learn to paint and never touching a brush. – Florian May 15 '19 at 13:57

## 1 Answer

The frequency domain conjugate multiplication (correlation) of the received signal with the reference signal followed by the power delay profile will provide you the overall signal to noise ratio. As explained in paper titled "SNR Estimation based on Sounding Reference Signal in LTE Uplink".

There are various noise reduction algorithm exist. Which will reduce the noise and will enhance the main expected signal component. There are adaptive filter which will converge to the expected value and depend on there performance, complexity is dependent. For example LMS, NLMS and RLS are the type of adaptive filters. Which will reduce the noise but the accuracy is directly proportional to the complexity.

By filtering we can reduce the noise, but we cant remove it totally for example AWGN noise.