Re: [Discuss-gnuradio] RTT time of ping based on OFDM tunnel

[Alex Zhang]

thanks for those constructive questions.

On Thu, May 17, 2012 at 7:22 PM, Brian Padalino <address@hidden> wrote:

On Thu, May 17, 2012 at 7:06 PM, Alex Zhang <address@hidden> wrote:
> It seems that the GMSK modulation is doing well in supporting higher
> datarate over GNURadio, than OFDM modulation.
>
> Today I did test using the GMSK between two USRP N210 nodes, the maximum
> data rate with acceptable packet loss rate of ping (2%) is 3Mbps. It can
> support video transmission.
> This result is much better than the existing OFDM implementation in
> GNURadio. I beleive that OFDM signal processing is much more complex than
> GMSK and cost much more resource than GMSK, but I did not expect that
> GNURadio OFDM link can only support less than 200kbps data rate.

What's the EVM of your OFDM signal coming out of the RF daughterboard?
 Are you backing off enough to allow for OFDM's high PAPR to remain in
the linear region?

I think backoff time is enough, as I actually did not see the conflict occurs.

 

GMSK has the nice advantage of being constant envelope, so compression
doesn't matter.  OFDM, on the other hand, doesn't get that luxury and,
in fact, requires a significant backoff.

Have you tried running the code without a USRP and just simulating
AWGN and possibly a slight frequency offset to see if the OFDM code is
adequately robust?  What are your subcarrier modulations?  Are you
running with any FEC?

No FEC. Actually, I used the code of tunnel.py without any fundamental modification. I just want to know if any other have ever tried this code and what the performance is. 

Sorry for all the questions.

Brian

> Can any guru give your testing result?
>
> On Wed, May 16, 2012 at 10:41 PM, Alex Zhang <address@hidden>
> wrote:
>>
>> Hi all,
>>
>> I did some experiments of Ping, between two USRP N210 nodes, over the OFDM
>> link. I need to know if any other has the same performance as I did.
>>
>> This OFDM link is using two frequencies,  900MHz and 1800MHz, as the SBX
>> daughter board is used. The antenna is VERT900 which is working in  824 to
>> 960 MHz, 1710 to 1990 MHz Quad-band .
>> Two computers of intel i7 (4 cores @3.4GHz) are used to run gnuradio.
>>
>> My performance is that as below:
>>
>> Static ARP entries are used for both the computer. Tunnel devices are
>> setup receptively to run ping on it.
>>
>> Bandwidth = 500K, occ=400/512, average RTT time is 14ms, and packet loss
>> of ping is 17%
>> Bandwidth = 1M, occ = 400/512, average RTT time is 8ms, and packet loss of
>> ping 33%
>> Bandwidth = 1.2M, occ = 400/512, average RTT time is 7ms, but packte loss
>> of ping is degraded to >90%, means the PC can not support such kind
>> bandwidth.
>>
>> If I use two intel core 2 PC, the supported bandwidth with acceptable
>> paccket loss rate (35%) is a little higher than 500K (RTT = 20ms). With more
>> bandwidth, the computer would be slow down significantly and the link
>> quality is near crash.
>>
>> So I need to know if any other friend have tested such experiment and what
>> the corresponding performance is.
>>
>> --
>>
>> Alex,
>> Dreams can come true – just believe.
>>
>
>
>
> --
>
> Alex,
> Dreams can come true – just believe.
>
>

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

-- 

Alex,

Dreams can come true – just believe.