openLTE annoucement

GNU Radio
1

I am pleased to announce version 00.05 of openLTE, an open source LTE
project (openLTE download | SourceForge.net). This version includes a
gnuradio application that reads recorded I/Q downlink LTE data from a
file, decodes MIB and SIB1, and displays the decoded information.
This information includes:

  • Measured frequency offset
  • System Frame Number
  • Physical Cell ID
  • Number of transmit antennas
  • Bandwidth
  • PHICH format
  • PLMN Identity
  • Tracking Area Code
  • Cell ID
  • Cell Barred information
  • Reselection information
  • Frequency band
  • System Info scheduling
  • Duplexing mode

This version requires input data to be interleaved 8-bit I and Q (I
first) recorded at 30.72MSPS centered on an LTE channel. Please see
the README file for installation and usage.

There are limitations with this version that will be addressed in the
future. Some of the more pertinent ones are:

  • Only handles FDD
  • Only handles normal cyclic prefix
  • Only handles DCI 1A with localized VRBs on the PDCCH
  • Only handles normal PHICH duration

Enjoy,
Ben

2

Hi Ben,

Thanks for letting us know about your work. This is very exciting.

Have you heard about the Wireless Innovation Forum’s call for
proposals on open LTE frameworks?
http://www.businesswire.com/news/home/20120327006941/en/

This might be a good candidate to pursue it if you are interested.

Tom

3

Tom,

I had not caught wind of this. Thanks for sending the link along, I
will check it out and see where I fit in.

Ben

4

On Sat, Apr 28, 2012 at 9:26 PM, Ben W. [email protected]
wrote:

I am pleased to announce version 00.05 of openLTE, an open source LTE
project (openLTE download | SourceForge.net). This version includes a
gnuradio application that reads recorded I/Q downlink LTE data from a
file, decodes MIB and SIB1, and displays the decoded information.

This is awesome.

I pulled this from the local Verizon tower with a USRP2.

DL LTE Channel found:
Frequency Offset = -19.63
System Frame Number = 234
Physical Cell ID = 74
Number of TX Antennas = 2
Bandwidth (MHz) = 10
PHICH Duration = Normal
PHICH Resource = 1

I wasn’t able to decode the SIB. I have what appears to be a decent
signal with a discone antenna near the window, but the CRC check on
the DCI never clears. I’ll play with it some more when I have some
free time (unfortunately not in the near future). Really cool stuff
though.

Thomas

5

Thomas,

Thanks for giving it a try!! If you get me your VZW recording, I can
take a closer look to see what is going wrong with the SIB decoding.

Thanks,
Ben

6

Hi Ben,

Do not hesitate to ask questions about the call for proposals. We’re
looking for people with a strong open-source background in addition to
code contributions.

Inside of the Forum the work has started on a proper model of the PHY
and the code in Mathematica code for it will be released soon. I think
you could request access to it even before publishing. Code for
various parts of the PHY is more then welcome as well.

</WInnF CommBB vice chair hat off>

On Tue, May 1, 2012 at 01:39, Ben W. [email protected] wrote:

Thanks for letting us know about your work. This is very exciting.
On Sat, Apr 28, 2012 at 9:26 PM, Ben W. [email protected] wrote:

  • PHICH format
    first) recorded at 30.72MSPS centered on an LTE channel. Please see
    Ben

Discuss-gnuradio mailing list
[email protected]
Discuss-gnuradio Info Page

Discuss-gnuradio mailing list
[email protected]
Discuss-gnuradio Info Page

–
Regards,
Alexander C…
CEO, Fairwaves LLC / ООО УмРадио

7

Ben W. <bwojtowi gmail.com> writes:

  • Bandwidth
    This version requires input data to be interleaved 8-bit I and Q (I
    Enjoy,
    Ben

Good!

One more information I can offer is that actually we don’t need
30.72Msps as
it in 3GPP spec. I have developed an LTE Downlink Receiver which works
on
19.2Msps for 100RB LTE system and it decodes actual 4G/LTE (both TDD and
FDD) signal in field Beijing successfully (SIB information is parsed
finally).

Why 19.2Msps is important? I think some cheap SDR front-ends will
benefit
from that, such as hackRF which only support ~20Msps and 20MHz
bandwidth.

The method is simple: 19.2*8/5 = 30.72, which is a general fractional
sampling rate conversion method in Digital Signal Processing.

See the project here: JiaoXianjun (Jiao Xianjun) · GitHub LTE-Cell-Scanner
project.

8

On Wed, May 2, 2012 at 8:18 AM, Ben W. [email protected]
wrote:

Thomas,

Thanks for giving it a try!! If you get me your VZW recording, I can
take a closer look to see what is going wrong with the SIB decoding.

I’ll try to post some captures by the end of the week. I should
mention that I used some hacks to create the 30.72 Msps signal. I
captured the 10 MHz signal at 12.5 and 25 Msps and resampled (in real
time) with some test code linked to the OpenBTS transceiver. There
were some limitations to keep the samples moving, and it’s very
possible that I mangled the signal in the process.

I would like to pull a 40 Msps capture from test equipment (30.72 Msps
is not an option), resample off-line, and try again.

Thomas