Re: [Discuss-gnuradio] uhd_sink+uhd_source VS virtual_sink+virtual

2015-11-05 23:15 GMT+08:00 Marcus Müller <address@hidden>:

Hi Ekko,

please reply to the list, not to individual people :)

1.i can use source block generate digita signal,
Every block in GNU Radio is a digital block (your computer is a digital device)!

2.i am not sure whether "chunks to symbol"block make the job of digital-to-analog

no. Digital to analog conversion only happens in your SDR peripheral, in your case, the USRP.

3.mixing it up to carrier ferq is done by uhd_sink right?

No, not by the usrp_sink (which is but a piece of software), but by the hardware in the USRP.

Best regards,
Marcus
On 11/05/2015 04:11 PM, chai E wrote:
hello marcus
as A newer ,i have made some Stupid questions，

so very thank you , for your Patiently answers，

i connect the usrp“tx and rx pin whit a wire to imitate that no error,so i think the ber should be zero.

i got that this opion is wrong from your last email

also

my question:
1.i can use source block generate digita signal,

2.i am not sure whether "chunks to symbol"block make the job of digital-to-analog

3.mixing it up to carrier ferq is done by uhd_sink right?

thank you

---Ekko
2015-11-05 22:33 GMT+08:00 Marcus Müller <address@hidden>:
Hi Ekko,

i connect the usrp“tx and rx pin whit a wire to imitate that no error,so i think the ber should be zero.
If you still think that, you have understood neither mine nor Marcus Leech's email; I'm afraid I can only ask you to read them again.

my problem is that if i want to use uhd_sink and source to test this grc,is there some other work i need to do compare with the connect of virtual sink +source.

Again: Virtual sink/source are not doing *anything* to the signal, whereas the USRP does.
I feel like I shouldn't be simplifying as much, but consider this:
digital complex baseband is in theory equivalent to a real RF bandpass signal, but it's not the same.
If you directly connect digital things (which is what virtual sink/source does) then it would be like taking a cake recipe, and giving it to your neighbor, to read it. Your neighbor will have no problems reading it.

Transmitting something with a real piece of SDR hardware and receiving it is more like:

writing down that recipe (generating the digital signal)

baking that cake, (digital-to-analog conversion, yielding a physical baseband signal)

putting that in a box, (mixing it up to the carrier frequency)

bringing that box to a post office, (transporting it through a coax/microwave connection to your channel)

sending that box to your neighbor, (sending the RF signal through that channel, i.e. your direct cable)

your neighbor getting that box, (your receiver getting that signal)

taking the cake out of the box, (receiver mixes down to I/Q baseband),

and decomposing the cake to finally (analog-to-digital conversion)

get an estimate for the original recipe (SDR algorithm)

You will never get a zero BER (which means of infinitely many recipes you transmit with the bake-cake-box-post-box-cake-decompose method, every single one is perfectly estimated), but you can get one that is sufficiently small for a given application if you can get your SNR to be good enough.

Best regards,
Marcus
On 11/05/2015 02:37 PM, chai E wrote:
hello marcus
i connect the usrp“tx and rx pin whit a wire to imitate that no error,so i think the ber should be zero.

my problem is that if i want to use uhd_sink and source to test this grc,is there some other work i need to do compare with the connect of virtual sink +source.

this is my question : with the connect of virtual sink+source the ber is zero ,i think with the connect of uhd_sink+source the ber should be 0.01 or other number clear to zero,but when i use uhd_sink +source the ber is clear to 1.

thank for you reply

--Ekko
2015-11-03 22:49 GMT+08:00 Marcus Müller <address@hidden>:
Hi Ekko,

>difference between uhd_sink+uhd_source with virtual_sink+virtual_source

uhd_sink / _source talk to radio hardware, virtual sink and source are just "invisible" GRC connections, and simply talk to the block that is attached to them.

> in this grc when i use virtual_sink+virtual_source,the ber is zero

obviously, because it's simply a direct connection, no noise, hence, no errors.

> but when i use uhd_sink+uhd_source, and i connect the B200's tx and rx port with a RF wire. but the ber is not zero.

Real world radios have noise (and other imperfections), and hence there's bit errors. That's the core problem of digital communications, and has been for the last couple of decades.

> how can i use uhd_sink+uh_source just like virtual_sink+virtual_source

You can't. One reproduces perfectly, the other is the real RF world. Getting information across the wireless channel without error is what science has been working on, and since Shannon we know that you can not achieve a transmission with arbitrarily low BER over a channel with limited SNR at arbitrary speed. Every real channel has limited SNR.

What you do is apply your signal processing and communication theory skills to counteract the effects of the channel, check correctness and correct errors (if possible), possibly make the sender retransmit etc; but you will never break Shannon's limits.

Best regards,
Marcus

On 11/03/2015 01:58 PM, chai E wrote:
hello all
i have a question about the difference between uhd_sink+uhd_source with virtual_sink+virtual_source
the accessory is the grc file i am running.
in this grc when i use virtual_sink+virtual_source,the ber is zero ,in other words,the sorce file is same with the dest file
but when i use uhd_sink+uhd_source, and i connect the B200's tx and rx port with a RF wire. but the ber is not zero.
i want to konw what is the difference ,and how can i use uhd_sink+uh_source just like virtual_sink+virtual_source

--Ekko
_______________________________________________
Discuss-gnuradio mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
_______________________________________________
Discuss-gnuradio mailing list
address@hidden
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio

From:	chai E
Subject:	Re: [Discuss-gnuradio] uhd_sink+uhd_source VS virtual_sink+virtual_source
Date:	Thu, 5 Nov 2015 23:22:13 +0800

Re: [Discuss-gnuradio] uhd_sink+uhd_source VS virtual_sink+virtual_sourc