Re: [Discuss-gnuradio] why can't use iwconfig when I run the gr-ieee802-11

[Marcus Müller]

Hi Siyu,


On 06/09/2017 11:04 PM, zhan siyu wrote:
> Hi Bastian,
>
> Thanks for reply. I know they are two different values. 5M is for
> the physical layer and 115K bits/s is for the application layer.  The
> application layer's upper bound should be the physical layer.
um, yeah, it's a strict upper bound. But there's a significant
difference between physical layer bitrate and payload rate. You don't
occupy the channel 100% of time, and you don't have 100% coding efficiency.
> So the value of application layer should always be less than the
> physical layer' s.
Indeed.
> However, I think my program is right which can take as much as the
> physical layer bandwidth can support.
No, that's not how IEEE802.11 works.
> You can see my program's result. The 802.11 average speed is 22Mb/s
> and I can get 15Mb/s. So I think, as the bandwidth is 5M, then the
> coding rate should be 5/3 = 1.67M bits/s. So I expect a throughput
> around 1M b/s.  But the result is just 1/10 of my expection. Too far
> away. I'm new to the communication. Not sure if I'm right. Please
> correct me if I'm wrong.
You're wrong, because, as Basti said, gr-ieee-802.11 doesn't support
backpressure, and thus, your method of measuring doesn't work.
>
> My professor wants me to use it in his prototype, please help me. And
> could you tell me if there is any software which can correctly measure
> the physical layer bit rate?
An approach here would be simply to calculate the maximum packet rate
you can transmit and multiply that with the packet payload size. It's
easy as that!

Best regards,
Marcus