The delay before the decoder might be throwing off the alignment at the
output. Might consider setting to 0 or removing.

Generally If you use a test input vector like the block of 0xff and 0x00 I
mentioned before, you should be able to look at the relative delay between
input and output by putting the source output and decoder output into a
time sink and measuring the delay between the source signal being non 0 and
the decoder output being non 0.

Jared.

On Wed, Aug 23, 2023, 4:58 AM Jiya Johnson <jiyajohnso...@gmail.com> wrote:

> I am getting constant BER through our the SNR ranges
>
>
>
>
> On Tue, Aug 22, 2023, 7:15 PM Jiya Johnson <jiyajohnso...@gmail.com>
> wrote:
>
>> Ok, I will check.
>> I was confused with this stack input (
>> https://stackoverflow.com/questions/54946638/punctured-convolutional-codes-in-gnu-radio
>> )
>> I will try with 1011 and delay blocks and will update you.
>>
>> On Tue, Aug 22, 2023, 7:12 PM U L <jaredd...@gmail.com> wrote:
>>
>>> The extended en/decoder does puncturing and zero stuffing internally so
>>> you don't need an extra puncturing block. Like I mentioned, the puncturing
>>> string can only have 0s and 1s so 13 is invalid. I'd try '1011' instead.
>>>
>>> Finally I think you still need some delay between your source and the
>>> ber block to align your input bytes with those from your output. The output
>>> does not start with the first decoded bit, but some number of 0s. In one of
>>> your prior flowgraphs you did have a delay. What happened to it?
>>>
>>> Jared.
>>>
>>>
>>> On Tue, Aug 22, 2023, 4:13 AM Jiya Johnson <jiyajohnso...@gmail.com>
>>> wrote:
>>>
>>>> Hi Jared,
>>>>      Completed the flowgraph based on the inputs from you like i have
>>>> taken code rate :2/3 and pattern 1101(0xD)-puncture size-4 as reference
>>>> from (
>>>> https://stackoverflow.com/questions/54946638/punctured-convolutional-codes-in-gnu-radio)
>>>> bt still i am facing the issue.Always i am getting 10^-0.3 BER FROM 0 to 15
>>>> dB.
>>>>
>>>> [image: Screenshot from 2023-08-22 15-41-43.png]
>>>>
>>>>

Reply via email to