Re: [Discuss-gnuradio] bpsk lost some packet when transmit data discontinuously

[Tom Rondeau]

On Tue, Mar 9, 2010 at 3:03 AM, Fisheep <address@hidden> wrote:
>
> Hi:
>
> I have some problem about dbpsk and bpsk using benchmark_tx/rx.py with some
> modification.
>
> My experiment set up was that the transmitter waits for a time delay (0.5 s)
> after sending a packet.
> It means that we transmit data discontinuously.
>
> If we using dbpsk modulation, we can successfully receive every packet.
> But if we removed the differential encoder/decoder (equals to bpsk
> modulation), we uaually
> lost some packet (almost half of the data lost).
>
> However, when we transmit data continuously,the bpsk modulation seems work
> fine and we can successfully receive every packet.
>
> My problem is that what reason causing the bpsk modulation can not
> successfully receive packet when
> transmit data discontinuously.
>
> Could anyone have any idea about it, please let me know.
> I am deeply appreciative.
>
>                             Fisheep :)


Fisheep,

I'm wondering how you are compensating for the phase ambiguity of the
BPSK signal in the receiver? In the current DBPSK, there is no method
to properly rotate the constellation because the differential encoding
doesn't care. If you are just transmitting BPSK, you'll have a 180
degree phase ambiguity in the receiver. If that's the case, then for
every burst of a packet, when the receiver locks to the constellation,
it will be in one of two phase rotations, one which works and one
which inverts every symbol. They should be distributed with equal
probability, so half of the time you'll have the correct rotation and
the other half you won't.

When you transmit continuously, I would expect that you'll also see
the same behavior. If you've got the right rotation, you'll receive
every packet. Otherwise, you'll see nothing. Does that happen
sometimes?

Tom