discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] Eventstream source latency


From: Jacob Gilbert
Subject: Re: [Discuss-gnuradio] Eventstream source latency
Date: Sat, 17 Aug 2019 16:26:05 -0600

Hi Daniel

You might be interested in the gr-pdu_utils and gr-timing_utils modules. They are about to get a significant update but there is a decent writeup on the concept of operations in github.

Sub-ms latency is possible depending on host hardware and system configuration.

Jacob



On Fri, Aug 16, 2019, 10:08 Lundberg, Daniel via Discuss-gnuradio <address@hidden> wrote:
I have a flowgraph with a simple FDMA implementation with two channels.  One channel has a constant stream.  The other channel has complex sample bursts generated in a custom python function, which I package as a PDU.  I apply the frequency modulation within the python file to place it in the correct frequency channel.  I send the PDU to an eventstream source and then add the two channels in an add block before sending the merged signal to a USRP sink.  I am seeing latencies of over a second from the PDU generation to when the eventstream source sends them on.  My sample rate is ~5 Ms/s and my PDU is about 1.25 MS, so about a quarter of a second.  My eventstream source block is using ASAP event placement and MEMCPY settings.

Is this level of latency expected?  I do not fully understand the merge behavior of the eventstream source, but I would expect perhaps 2X the PDU size of latency if it has a PDU length number of zeros in it's buffer before inserting the PDU.

Thank you,
DL
_______________________________________________
Discuss-gnuradio mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio

reply via email to

[Prev in Thread] Current Thread [Next in Thread]