Forums

Negative Input values to RS encoder

Started by khalloud70 October 15, 2008
I want to feed negative values to the input of reed solomon encoder??
is this means that i have to use Binary RS encoder ,and if so ,how can i
convert the negative values to a  BIT assignment.??


On Oct 15, 10:46&#2013266080;pm, "khalloud70" <khallou...@gmail.com> wrote:
> I want to feed negative values to the input of reed solomon encoder?? > is this means that i have to use Binary RS encoder ,and if so ,how can i > convert the negative values to a &#2013266080;BIT assignment.??
The RS encoder does not care. You can represent the data using twos complement: 0 to 127 for non-negative values; 255 to 128 for negative ones. Or you can make up your own scheme; say for example 0,2,4,..254 are non-negative and 1,3,5,...255 are negative. John
On Oct 15, 10:46&#2013266080;pm, "khalloud70" <khallou...@gmail.com> wrote:
> I want to feed negative values to the input of reed solomon encoder?? > is this means that i have to use Binary RS encoder ,and if so ,how can i > convert the negative values to a &#2013266080;BIT assignment.??
Antipodal mapping??? a "1" is a positive value and a "0" is a similar magnitude negative value. But you may map the bits anyway you want. Clay
On Wed, 15 Oct 2008 22:13:17 -0700 (PDT), John <sampson164@gmail.com>
wrote:

>On Oct 15, 10:46&#2013266080;pm, "khalloud70" <khallou...@gmail.com> wrote: >> I want to feed negative values to the input of reed solomon encoder?? >> is this means that i have to use Binary RS encoder ,and if so ,how can i >> convert the negative values to a &#2013266080;BIT assignment.?? > >The RS encoder does not care. You can represent the data using twos >complement: 0 to 127 for non-negative values; 255 to 128 for negative >ones. Or you can make up your own scheme; say for example 0,2,4,..254 >are non-negative and 1,3,5,...255 are negative. > >John
Another way to look at it is that the RS system works with "symbols" which can be multi-bit. The bit patterns that make up the "symbols" can be interpreted elsewhere however they wish, i.e., ASCII, two's complement data, binary streams, etc., etc. It's like writing down a string of ones and zeros and asking what format it's in...it doesn't really matter. Eric Jacobsen Minister of Algorithms Abineau Communications http://www.ericjacobsen.org Blog: http://www.dsprelated.com/blogs-1/hf/Eric_Jacobsen.php
On Thu, 16 Oct 2008 09:26:16 -0700, Eric Jacobsen
<eric.jacobsen@ieee.org> wrote:
>Another way to look at it is that the RS system works with "symbols" >which can be multi-bit. The bit patterns that make up the "symbols" >can be interpreted elsewhere however they wish, i.e., ASCII, two's >complement data, binary streams, etc., etc. It's like writing down >a string of ones and zeros and asking what format it's in...it doesn't >really matter.
I can think of one reason why it matters but I'm not sure if it's relevant in the context of RS decoders. If the operator (RS decoder in this case) needs to do any sorting on the symbols ie an operation of comparison other than straight equality then it needs to know which symbol is greater than another ie if the operator needs a "field" (if I am remembering my algebra correctly) than representation is important. If all one needs is equality checks then there is no such requirement. Muzaffer Kal ASIC/FPGA Design Services DSPIA INC. http://www.dspia.com
On Thu, 16 Oct 2008 10:03:15 -0700, Muzaffer Kal <kal@dspia.com>
wrote:

>On Thu, 16 Oct 2008 09:26:16 -0700, Eric Jacobsen ><eric.jacobsen@ieee.org> wrote: >>Another way to look at it is that the RS system works with "symbols" >>which can be multi-bit. The bit patterns that make up the "symbols" >>can be interpreted elsewhere however they wish, i.e., ASCII, two's >>complement data, binary streams, etc., etc. It's like writing down >>a string of ones and zeros and asking what format it's in...it doesn't >>really matter. > >I can think of one reason why it matters but I'm not sure if it's >relevant in the context of RS decoders. If the operator (RS decoder in >this case) needs to do any sorting on the symbols ie an operation of >comparison other than straight equality then it needs to know which >symbol is greater than another ie if the operator needs a "field" (if >I am remembering my algebra correctly) than representation is >important. If all one needs is equality checks then there is no such >requirement. > >Muzaffer Kal > >ASIC/FPGA Design Services >DSPIA INC. >http://www.dspia.com
Most decoding of the sort that Reed-Solomon decoders use is analyzed from the standpoint of "symbols" in an "alphabet". The symbols are not greater than or less than each other, they're just different. Likewise, the codewords in a code space are not greater than or less than each other (except in the sense of something like Hamming weight, but that's not compared the same as a number system). Soft-decision decoders do compare the "values" of symbols, but that's only in the sense that the reliability of the correct reception of that symbol is evaluated, not the "value" of the symbol itself. Eric Jacobsen Minister of Algorithms Abineau Communications http://www.ericjacobsen.org Blog: http://www.dsprelated.com/blogs-1/hf/Eric_Jacobsen.php