Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
source encoders: it is techinques to encode the signal to be sent to the write distination like 3gp, mpeg, amr, evrc, smv.
it has some types like vedio encoders, vocoders,speach encoders
these pdf will explain to you some types of source and channel encoders
purpose of the source coding is to eliminate redundant data from a signal. For example, if your digital data contains many '0' as in the sequence "000010100010000011", then it is much more efficient to code that long data into "52461". "5" means that there are 4 '0' and one '1', "2" means there is one '0' and one '1', etc. In reality, indeed many physical signal such speech and image contain many redundancies, so source coding is very important. By using source coding, people could design many data compression algorithms to save data efficiently, or reduce the bandwidth for transmitting communication signal via a channel.
The purpose of channel coding is to introduce redundancies into information data. This seeems like a contradiction with that of source coding. But you must be aware that channel coding introduces redundancies in a controlled manner, whereas redundancies inherent in physical signals are random (uncontrolled). This redundancies introduced by channel coding are very important to reduce (or in some cases eliminated) the errors which may occur when you transmit or save/record information data.
coding by introducing parity bit(s) is the simplest way to do channel coding. Parity bits only can be used to detect error, but not to correct error. This is enough for example in computer systems, but not in communication systems.
Other channel coding such as convolutional coding and reed-solomon coding could also correct errors. So, generally there are two kind of channel coding: error detection and error correction.
ther channel coding such as convolutional coding and reed-solomon coding could also correct errors. So, generally there are two kind of channel coding: error detection and error correction.
Hello,
I guess all the above replies have given you an idea of source and channel coding.
Very simply put, source coding is about compression...'00000' may be represented as 5 '0', as mentioned in one of the replies above.
Channel coding is about adding redundancies...These redundant bits are added to detect and correct errors.
There seems to be contradiction. Why do we need both souce and channel coding??
One adds redundancies and one removes them.
For video transmission, there need to be a trade-off between source and channel coding.
Please refer to the attached paper on trade off.
Cheers,
ans2006.
as i said in my previous post, the different between redundacies at the source and redundancies introduced by channel coding is that, the first one is uncontrolled, and the second is controlled. So if one gives more attention to both, it is no contradiction at all.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
