Kitz Forum

Internet => Interesting Websites => Topic started by: Weaver on May 27, 2019, 03:55:30 PM

Title: Communications Mathematics Tutorials Website
Post by: Weaver on May 27, 2019, 03:55:30 PM
A fantastic resource for serious, thorough tutorials about communications theory and practice in analog and digital - waveforms, signal processing and coding/decoding. It has a particularly good explanation of trellis coded modulation, lattices and cosets.

Tutorials on Digital Communications Engineering (http://complextoreal.com/tutorials/)

The author has done a huge amount of work. Valuable references too.
Title: Re: Communications Mathematics Tutorials Website
Post by: burakkucat on May 27, 2019, 04:36:36 PM
Thank you for the link. It now looks as if some reading is required, by me.  :)
Title: Re: Communications Mathematics Tutorials Website
Post by: Weaver on May 27, 2019, 05:45:37 PM
I’m actually starting to partly get some of the trellis coded modulation stuff of Gottfried Ungerboeck (Ungerböck ? No umlauts in the original’s systems?) but I haven’t read the higher order stuff ‘four dimensional’ stuff of Wei which is used in ADSL.

I can see why it has got a chance of working, but I haven’t managed to work out the numbers to show why the gains are so much more than any losses. He is a very very clever fellow indeed.

The sequence coding aspect makes you think of language. Words of say English that are corrupted are recognisable to some extent by virtue of the impossibility of certain letter combinations and the likelihood of others in context, indeed with some texts using the English lexicon only, you can stick to the dictionary and require that any possible candidate correction be a dictionary word. It would help your with ‘at’ vs ‘an’ vs ‘am’ vs ‘it’, or ‘in’ vs ‘on’ as there is not a long enough sequence and there are too few constraints in those particular examples. But going further, syntactic context would save you if you widened the capabilities of sequence recognition.

Finding the constellation points that are farthest apart and then finding a practical way to exploit it like that is an impressive achievement of Ungerboeck’s.
Title: Re: Communications Mathematics Tutorials Website
Post by: 22over7 on May 28, 2019, 07:20:10 AM
Thanks from me too for the link.  Anything that bridges the (rather large) gap between abstract mathematics and concrete application (to "broadband") is really valuable.
Can you recommend anything else of that kind?

 
Title: Re: Communications Mathematics Tutorials Website
Post by: Weaver on May 28, 2019, 07:54:33 AM
This article is superbly done and very amusing. Someone put a vast amount of work into it: An Interactive Guide To The Fourier Transform (https://betterexplained.com/articles/an-interactive-guide-to-the-fourier-transform/)

I will dig around a bit for material for other topics. Anyone have any specific questions ideas for areas that they would like illuminated?
Title: Re: Communications Mathematics Tutorials Website
Post by: 22over7 on May 28, 2019, 08:29:46 PM
Quote
Anyone have any specific questions ideas for areas that they would like illuminated?

Not specifically!  What I found delightful about your link is that there are various areas of pure mathematics, like analysis of analogue signals,
and information theory, of discrete codes transmitted over sampled noisy channels, that are bridged to a unforgiving real-world application.
Beauty and The Beast.  It can be difficult to cross the bridge in either direction.

A book that I once found very helpful when writing signal processing code was by Ronald N. Bracewell: "The Fourier Transform and
its Applications" (2nd edition 1978).  https://en.wikipedia.org/wiki/Ronald_N._Bracewell (https://en.wikipedia.org/wiki/Ronald_N._Bracewell).
 
A fascinating (and colourful) bridge-straddling character is Oliver Heaviside  https://en.wikipedia.org/wiki/Oliver_Heaviside (https://en.wikipedia.org/wiki/Oliver_Heaviside).