Reply by cpshah99 November 7, 20082008-11-07
>>Hi All >> >>I would be very thankful if you could verify the result that I have
got.
>> >>System-1: BPSK, AWGN, Reed Solomon Code (255,127) i.e. rate 0.5 >>Result: At 0 dB the BER is 0.17 and at 5.5 dB BER is 0.003 >> >>System-2: BPSK, AWGN, Reed Solomon Code (255,239) 1.e. rate 0.93 >>Result: At 0 dB the BER is 0.09 and at 5.5 dB BER is 0.006 > >Matlab has a BERtool, which gives different results from those of yours.
>It gives: > >System-1: BPSK, AWGN, Reed Solomon Code (255,127) i.e. rate 0.5 >Result: At 0 dB the BER is 0.3765 and at 5.5 dB BER is 0.00108 > >System-2: BPSK, AWGN, Reed Solomon Code (255,239) 1.e. rate 0.93 >Result: At 0 dB the BER is 0.2563 and at 5.5 dB BER is 0.0151. > >You really should be able to do this on your own, assuming you have
access
>to this tool. > >Emre > > >
%%%%%% Hi Emre Thanks a lot!! Checked this tool. May be there is some minor problem in my code. Thanks Chintan
Reply by emre November 7, 20082008-11-07
>Hi All > >I would be very thankful if you could verify the result that I have got. > >System-1: BPSK, AWGN, Reed Solomon Code (255,127) i.e. rate 0.5 >Result: At 0 dB the BER is 0.17 and at 5.5 dB BER is 0.003 > >System-2: BPSK, AWGN, Reed Solomon Code (255,239) 1.e. rate 0.93 >Result: At 0 dB the BER is 0.09 and at 5.5 dB BER is 0.006
Matlab has a BERtool, which gives different results from those of yours. It gives: System-1: BPSK, AWGN, Reed Solomon Code (255,127) i.e. rate 0.5 Result: At 0 dB the BER is 0.3765 and at 5.5 dB BER is 0.00108 System-2: BPSK, AWGN, Reed Solomon Code (255,239) 1.e. rate 0.93 Result: At 0 dB the BER is 0.2563 and at 5.5 dB BER is 0.0151. You really should be able to do this on your own, assuming you have access to this tool. Emre
Reply by cpshah99 November 6, 20082008-11-06
Hi 

Thanks for your reply.

I believe system 1 should work better because it is rate 1/2 code.

> >1. Mapping RS into BPSK in AWGN channel is bad idea.
May be, but to establish intial result, it is not that bad.
>2. Apparently you are using the hard decision decoder which is bad idea.
This is what one of the current modem is using, with which I am gng to competer
>3. BER for non-binary code is bad idea.
No idea abt that.
>3. BER ~ 1e-3 after that kind of decoder is immensely bad. Something >like 1e-7...1e-9 would be reasonable area of operation.
Defiantely.
>4. There is no point to do simulations when there is a textbook >analytical solution.
I dont have any results, that use freq selective channel with RS code. I am using RS code because they can correct burst errors. In my system I am using adaptive DFE, which creates burst errors. Now, the channel that I am using is very severe, may be 100 symbols of multipath. So if somebody can confirm my AWGN results, I can atleast be sure of my encoder and decoder. Then rest entirely depends on the performance of adaptive DFE.
>5. You stupident, go do your homework.
Thanks for the constructive comment!!!!! Regards, Chintan
Reply by Vladimir Vassilevsky November 6, 20082008-11-06

cpshah99 wrote:

> Hi All > > I would be very thankful if you could verify the result that I have got. > > System-1: BPSK, AWGN, Reed Solomon Code (255,127) i.e. rate 0.5 > Result: At 0 dB the BER is 0.17 and at 5.5 dB BER is 0.003 > > System-2: BPSK, AWGN, Reed Solomon Code (255,239) 1.e. rate 0.93 > Result: At 0 dB the BER is 0.09 and at 5.5 dB BER is 0.006 > > Can I assume that in System-1, after 5.5 dB the BER is 0 where as in > system-2 still we can expect some BER values at higher SNRs? > > Your views matters a lot.
1. Mapping RS into BPSK in AWGN channel is bad idea. 2. Apparently you are using the hard decision decoder which is bad idea. 3. BER for non-binary code is bad idea. 3. BER ~ 1e-3 after that kind of decoder is immensely bad. Something like 1e-7...1e-9 would be reasonable area of operation. 4. There is no point to do simulations when there is a textbook analytical solution. 5. You stupident, go do your homework. VLV
Reply by emre November 6, 20082008-11-06
>System-1: BPSK, AWGN, Reed Solomon Code (255,127) i.e. rate 0.5 >Result: At 0 dB the BER is 0.17 and at 5.5 dB BER is 0.003 > >System-2: BPSK, AWGN, Reed Solomon Code (255,239) 1.e. rate 0.93 >Result: At 0 dB the BER is 0.09 and at 5.5 dB BER is 0.006 > >Can I assume that in System-1, after 5.5 dB the BER is 0 where as in >system-2 still we can expect some BER values at higher SNRs?
For comms applications 0.003 is a relatively high BER. The difference between 0.006 and 0.003 is rather minor at this scale. Why do you want to believe that the first one is better, anyway? The latter code carries more information bits, and may be more desirable at high SNR. In short, you should *not* assume BER is 0 after 5.5 dB for either system, in general. Emre
Reply by cpshah99 November 6, 20082008-11-06
Hi All

I would be very thankful if you could verify the result that I have got.

System-1: BPSK, AWGN, Reed Solomon Code (255,127) i.e. rate 0.5
Result: At 0 dB the BER is 0.17 and at 5.5 dB BER is 0.003

System-2: BPSK, AWGN, Reed Solomon Code (255,239) 1.e. rate 0.93
Result: At 0 dB the BER is 0.09 and at 5.5 dB BER is 0.006

Can I assume that in System-1, after 5.5 dB the BER is 0 where as in
system-2 still we can expect some BER values at higher SNRs?

Your views matters a lot.

Thanking you.

Regards,

Chintan