openlte-announcements — General announcements

[Openlte-announcements] V00.20.05

[Ben W. <bwo...@gm...>]

All,

Version 00.20.05 is available.  This version added two codeword support,
refactored PUCCH channel decoding for 1, 1A, and 1B, added SR configuration
index mapping, added liblte_common unit tests, and properly handling FI
fields for multiple AMD PDUs in the LTE library, added helper functions and
interprocess message queue support to the tools library, added SR support,
moved GW, MME, and PDCP away from the singleton pattern, added direct IPC
to a UE for MAC and PHY, removed QOS and fixed UL scheduling in favor of SR
and BSR based UL scheduling, and refactored the AMD reception buffer to
handle more than one PDU per SN in LTE_fdd_enodeb, and fixed a typo in
lte_pucch_demod.m.

I plan on focusing on picking through the emails and forum posts in the
coming weeks.  Please keep providing any and all feedback.

Thanks,
Ben

[Openlte-announcements] V00.20.04

[Ben W. <bwo...@gm...>]

All,

Version 00.20.04 is available.  This version added band 26, 27, and 28
support (thanks to Jeremy Quirke), fixed the length skipping in
liblte_mme_unpack_pdn_address_ie (thanks to Przemek Bereski), and fixed the
turbo encoder tail bits (thanks to Laurent Louf and Jeremy Quirke) in the
LTE library and fixed the handling of multiple RLC AMD PDUs throughout.

Please keep providing any and all feedback.

Thanks,
Ben

Re: [Openlte-announcements] srsGUI issues

[Ben W. <bwo...@gm...>]

Kia,

Please reach out to Ismael Gomez and Paul Sutton at srsLTE
<https://github.com/orgs/srsLTE/people>.

Thanks,
Ben

On Mon, Oct 31, 2016 at 12:43 PM, Kia Rodriguez <kro...@fi...> wrote:

> Hi,
>
> We're trying to run srsGUI while running srsUE.  The GUI does not run.  Is
> there a change in the configuration that needs to be updated in order to
> run the analytics of the GUI while running srsUE?
>
> Thank you,
>
> --
> *Kia M. Rodriguez*
> *FIU Undergraduate Research Assistant, MPACT Lab*
>
>
> ------------------------------------------------------------
> ------------------
> Developer Access Program for Intel Xeon Phi Processors
> Access to Intel Xeon Phi processor-based developer platforms.
> With one year of Intel Parallel Studio XE.
> Training and support from Colfax.
> Order your platform today. http://sdm.link/xeonphi
> _______________________________________________
> Openlte-announcements mailing list
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/openlte-announcements
>
>

[Openlte-announcements] srsGUI issues

[Kia R. <kro...@fi...>]

Hi,

We're trying to run srsGUI while running srsUE.  The GUI does not run.  Is
there a change in the configuration that needs to be updated in order to
run the analytics of the GUI while running srsUE?

Thank you,

-- 
*Kia M. Rodriguez*
*FIU Undergraduate Research Assistant, MPACT Lab*

[Openlte-announcements] V00.20.03

[Ben W. <bwo...@gm...>]

All,

Version 00.20.03 is available.  This version properly discards sequence
type extensions in liblte_rrc, properly typecast for
bladerf_get_timestamp() in LTE_fdd_enodeb_radio, and updated
enodeb_nat_script.sh to be more generic with ethernet interfaces.

Please keep providing any and all feedback.

Thanks,
Ben

[Openlte-announcements] (no subject)

[Idress M. <ken...@ya...>]

hi, i am facing problems in installing openLTE. after running cmake and i run sudo make i get these errors:[ 70%] Built target LTE_file_recorder
[ 72%] Building CXX object LTE_fdd_enodeb/CMakeFiles/LTE_fdd_enodeb.dir/src/LTE_fdd_enb_radio.cc.o
LTE_fdd_enb_radio.cc: In member function ‘void LTE_fdd_enb_radio_bladerf::receive(LTE_FDD_ENB_RADIO_PARAMS_STRUCT*)’:
LTE_fdd_enb_radio.cc:975:65: error: invalid conversion from ‘uint64* {aka long long unsigned int*}’ to ‘uint64_t* {aka long unsigned int*}’ [-fpermissive]
In file included from /usr/local/open/LTE_fdd_enodeb/hdr/LTE_fdd_enb_radio.h:54:0,
                 from LTE_fdd_enb_radio.cc:61:
/usr/local/include/libbladeRF.h:3002:15: error:   initializing argument 3 of ‘int bladerf_get_timestamp(bladerf*, bladerf_module, uint64_t*)’ [-fpermissive]
 int CALL_CONV bladerf_get_timestamp(struct bladerf *dev, bladerf_module mod,
               ^
make[2]: *** [LTE_fdd_enodeb/CMakeFiles/LTE_fdd_enodeb.dir/src/LTE_fdd_enb_radio.cc.o] Error 1
make[1]: *** [LTE_fdd_enodeb/CMakeFiles/LTE_fdd_enodeb.dir/all] Error 2
make: *** [all] Error 2
please help! i can't understand why i am getting these errors.thnx!

[Openlte-announcements] V00.20.02

[Ben W. <bwo...@gm...>]

All,

Version 00.20.02 is available. This version changed the QoS DL bytes per
subframe for default data to ensure QPSK only for 5MHz other DL modulation
schemes seem to not be functioning correctly.

Same standard warning, I have only been testing LTE_fdd_enodeb at 5MHz.

Still on the roadmap for future releases:
- incorporation of outstanding patches that have been submitted

Please keep providing any and all feedback.

Thanks,

Ben

Re: [Openlte-announcements] [Openlte-discuss] V00.20.01

[Sushree S. M. <Sus...@co...>]

Hi all,

I have difficulty in running openLTE. All the other functionalities run
fine except LTE_fdd_enodeb. I am running it on Fedora 23 and probably that
is the reason for failure. The enodeb script doesnt run successfully on
fedora as it doesn not have iptables. However, I believe that IP packets
tunneling is a later part of the LTE only when my device is connected. As
soon as I run LTE_fdd_enodeb ,it says 'fail cant start '. Can somebody help
me out in figuring out things for fedora. Does it run on Fedora?

Any help would be appreciated
Thanks,
Sushree

On Aug 1, 2016 7:47 AM, "Ben Wojtowicz" <bwo...@gm...> wrote:

> All,
>
> Version 00.20.01 is available. This version reworked the MAC RTS handling
> logic and limited the number of HARQ retransmissions in LTE_fdd_enodeb.
>
> Same standard warning, I have only been testing LTE_fdd_enodeb at 5MHz.
>
> On the roadmap for the next release:
> - incorporation of outstanding patches that have been submitted
>
> Please keep providing any and all feedback.
>
> Thanks,
>
> Ben
>
>
> ------------------------------------------------------------------------------
>
> _______________________________________________
> Openlte-discuss mailing list
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/openlte-discuss
>
>

[Openlte-announcements] V00.20.01

[Ben W. <bwo...@gm...>]

All,

Version 00.20.01 is available. This version reworked the MAC RTS handling
logic and limited the number of HARQ retransmissions in LTE_fdd_enodeb.

Same standard warning, I have only been testing LTE_fdd_enodeb at 5MHz.

On the roadmap for the next release:
- incorporation of outstanding patches that have been submitted

Please keep providing any and all feedback.

Thanks,

Ben

[Openlte-announcements] How to use wireshark to capture the message exchage on control plane (e.g., RRC, NAS )

[《資策會》張漢魁 <hk...@ii...>]

Dear OpenLTE Providers,

 

Because I had the problem on ping address via Samsung Note 3 (N9002) whie
"user authentication successful", "user full attached", "default bearer
setup", and "getting a IP from OpenLTE eNB+EPC" are fulfilled, 

I would like to dissect the message exchange on the control plane among UE,
eNB, and EPC. However, I can’t do it by using wireshark. 

 

Can you please instruct me how to use wireshark to dissect the message
exchange on the control plane among UE, eNB, and EPC?

For example, RRC, NAS.

 

Sincerely

 

Han-Kui Chang

[Openlte-announcements] Problem on pinging address via Samsung Note 3 (N9002) after "user authentication successful", "user full attached", "default bearer setup", and "getting a IP from OpenLTE eNB+EPC" are fulfilled

[《資策會》張漢魁 <hk...@ii...>]

Dear OpenLTE Providers,

 

First of all, Thank you for offering this excellent project.

 

I have tried to ping external address (e.g., 8.8.8.8) via the UE device,
Samsung Note 3 (N9002), after "user authentication successful", "user full
attached", "default bearer setup", and "getting a IP from OpenLTE eNB+EPC"
are fulfilled as illustrated in the attached photo; however, it failed !!

 

Can you please provide me some suggestions about how to successfully fulfil
the task? 

 

As the attached files, I provide the config. parameters for setting up this
connection.

 

Thank you in advance.

 

Sincerely

 

Han-Kui Chang

Re: [Openlte-announcements] [Openlte-discuss] V00.20.00

[Tom T. <tom...@et...>]

On Sun, Jul 17, 2016 at 11:35 PM, Ralph A. Schmid, dk5ras
<ra...@sc...> wrote:
> Is there a chance that also 1.4 MHz bandwidth works, or is it for sure that
> I do not even need to try it? :)

SIB1 from the eNB is recoverable at 1.4 MHz. On the scanner, MIB
decodes at 1.4 MHz though that shouldn't be any surprise as it's the
same for all bandwidths.

PDCCH : DCI Length 21, CRC 65535, Agg. Level 4, Block 0
PDCCH : DCI Format 1A:
        Format 0/1A differentiation              1
        Local/distributed VRB assignment         0
        Resource block assignment                18
        Modulation and coding                    2
        HARQ process number                      0
        New data indicator                       0
        Redundancy version                       1
        TPC command for PUCCH                    1
FRAME : SFN 0966, Ns 05
PDSCH : RIV 18, start 0, N_step -1, L_crbs 4
PDSCH : Transport block size A=144, Physical bits G=816
PDSCH : RNTI 65535, Modulation 2, Redundancy version 1
PDSCH : CRC Segmentation block r=0
PDSCH :     Rate match length E=816
PDSCH :     Rate match length D=172
PDSCH :     Turbo decode length K=168
PDSCH : Decoded transport block:
        60 40 04 03 00 01 00 00 00 18 03 50 00 00 42 00 00 00

  -TT

[Openlte-announcements] V00.20.00

[Ben W. <bwo...@gm...>]

All,

V00.20.00 is available.  This version increased the maximum messages size,
fixed extended power headroom MAC CE, added classmark3 support, added
tracking area update request support, fixed the fill bits in packing mobile
ID IE for IMEISV and the length in unpacking UE security capabilities IE
(thanks to Przemek Bereski for finding this), allowing NULL to be passed
for key_256 in service reject message packing (thanks to Przemek Bereski
and Peter Nguyen for finding this), added PDCCH size defines, added an
error return to DCI 1C unpack, clearing punctered turbo decode bits before
decoding, added AMD PDU segment support, padded all RRC enums to powers of
2, added support for several RRC IEs and messages (with the help of Przemek
Bereski) in the LTE library, restructured the radio classes to ease the
addition of new radio types, set the processor affinity to allow the
PHY/Radio to run on a particular core, fixed memcpy lengths in MAC, fixed
bugs in print_all_users (thanks to Sultan Qasim Khan for finding this) and
when receiving a service request message for a non-existent user (thanks to
Peter Nguyen for finding this), fixed a long standing PHY bug that was
including UL PDCCH allocations in a check of DL PRB usage, added error
logging for received AMD PDU segments, increased the DL datarate for
default data QoS, using C-RNTI when no IMSI/IMEI are available for printing
an attached user, Przemek Bereski added UE capability enquiry and UE
capability information support and bladeRF support.

A few notes:
- The PHY bug fix in LTE_fdd_enodeb for not including UL PDCCH allocations
in a check of DL PRB usage was causing a lot of grief in the PHY.  It ended
up throwing many DL packets and UL grants away with very little printed to
the log file.  It has been around ever since UL was added (which is quite a
while).  After fixing I have seen much more stable IP traffic.
- A huge shout-out to Przemek Bereski for the work on adding bladeRF!

Same standard warning, I have only been testing LTE_fdd_enodeb at 5MHz.

On the roadmap for the next release:
- incorporation of outstanding patches that have been submitted
- fixing of the synchronization error between PHY and MAC in LTE_fdd_enodeb
that several people have reported recently

Please keep providing any and all feedback.

Thanks,
Ben

[Openlte-announcements] V00.19.04

[Ben W. <bwo...@gm...>]

All,

V00.19.04 is available.  This version fixed the calculation of n_prime_p in
generate_dmrs_pucch.

The next release will include pulling in outstanding patches.

Thanks,
Ben

[Openlte-announcements] V00.19.03

[Ben W. <bwo...@gm...>]

All,
V00.19.03 is available.  This version added support for PUCCH decode in
liblte, added support for H-ARQ and PUCCH in LTE_fdd_enodeb, and fixed a
bug related to improper cleanup of the inactivity timer in LTE_fdd_enodeb.

Same warning as the previous releases: I have only been testing
LTE_fdd_enodeb at 5MHz.  Other bandwidths (specifically 1.4 and 3MHz) are
not currently working.

The next release will include pulling in outstanding patches.

Please keep providing any and all feedback.

Thanks,
Ben

[Openlte-announcements] V00.19.02

[Ben W. <bwo...@gm...>]

All,

V00.19.02 is available.  This version pulled in unit tests for the turbo
decoder from Ziming He, optimized the PBCH detection algorithm (thanks to
Ziming He), and fixed the setting of id for IMEISV in Mobile Idenity IE
(thanks to Mikhail Gudkov) in the LTE library, added the ability to print
all registered users, removed extra boost includes, added an RRC connection
reestablishment state machine, added a UE inactivity timer, actually
properly constructing MNC (thanks to Mikhail Gudkov), memcpy-ing IP packet
headers instead of typecasting (thanks to Damian Jarek), initialized
present flags and changed packet filter evaluation precedence in activate
dedicated EPS bearer context (thanks to Pedro Batista), moved the message
queue empty error to a warning, and properly updating the user list
iterator after erasing a user (thanks to Damian Jarek) in LTE_fdd_enodeb.

Same warning as last time: I have only been testing LTE_fdd_enodeb at
5MHz.  Other bandwidths (specifically 1.4 and 3MHz) are not currently
working.

Several people have provided lots of valuable feedback that has gone into
this release (see above for details).  But, there are still a few
outstanding patches that I am working on pulling in.  The next release will
include more of these patches and also some PUCCH functionality.

Please keep providing any and all feedback.

Thanks,
Ben

Re: [Openlte-announcements] [Openlte-discuss] V00.19.01

[Ben W. <bwo...@gm...>]

Denis,

The current release supports USRP N2X0 for LTE_file_recorder and
LTE_fdd_dl_scan.  LTE_fdd_enodeb currently only works with USRP B2X0.

Hope this helps,
Ben

On Wed, Dec 30, 2015 at 5:51 AM, Denis Lapshin <de...@mi...> wrote:

> Hi All,
>
> Can latest package fully support USRP N210 hardware?
>
> Thanks,
> Denis
>
>
> On 07.12.2015 5:30, Ben Wojtowicz wrote:
>
> All,
>
> V00.19.01 is available.  This version added support for all ID types for
> the mobile identity IE in MME, added support for more PDU types in PDCP,
> added defines for sizes of fixed arrays in LIBLTE_PHY_STRUCT, fixed the
> order of sizes for rate match and unmatch arrays in LIBLTE_PHY_STRUCT
> (thanks to Ziming He), added return values for liblte_phy_get_n_cce (thanks
> to Damian Jarek) and liblte_rlc_unpack_status_pdu (thanks to Mikhail
> Gudkov) in the LTE library, added a scoped lock RAII class for sem_t and
> pthread_mutex_t in the tools library, changed from using boost::mutex to
> pthread_mutex_t and_sem_t throughout, added helper functions in MAC,
> changed user deletion and C-RNTI release procedures, properly initializing
> SIB scheduling info, properly construct MNC (thanks to Mikhail Gudkov),
> fixed an issue with RTS response in MAC, properly initialized PHY
> allocation structure in MAC (thanks to Markus Grab), fixed the updating of
> VT(A) in the RLC retransmission buffer, and changed the QoS parameters for
> default data in LTE_fdd_enodeb.
>
> The highlights in this release are:
> - Lots of updates from folks using openLTE!!
> - The fix for VT(A) which allows for much more reliable IP traffic.
>
> WARNINGS:
> 1) I have only been testing LTE_fdd_enodeb at 5MHz.  Other bandwidths
> (specifically 1.4 and 3MHz) are not currently working.
>
> There were several people that provided feedback that went into this
> release.  In addition, there are still several outstanding patches from
> people that I'm working to pull in.  Sorry for the slower than normal
> response time.  This is my top priority for the next release.
>
> Please keep providing any and all feedback.
>
> Thanks,
> Ben
>
>
> ------------------------------------------------------------------------------
> Go from Idea to Many App Stores Faster with Intel(R) XDK
> Give your users amazing mobile app experiences with Intel(R) XDK.
> Use one codebase in this all-in-one HTML5 development environment.
> Design, debug & build mobile apps & 2D/3D high-impact games for multiple OSs.http://pubads.g.doubleclick.net/gampad/clk?id=254741911&iu=/4140
>
>
>
> _______________________________________________
> Openlte-discuss mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/openlte-discuss
>
>
> --
> Denis Lapshin
> mailto: de...@mi...
>
>

Re: [Openlte-announcements] [Openlte-discuss] V00.19.01

[Denis L. <de...@mi...>]

Hi All,

Can latest package fully support USRP N210 hardware?

Thanks,
Denis

On 07.12.2015 5:30, Ben Wojtowicz wrote:
> All,
>
> V00.19.01 is available.  This version added support for all ID types 
> for the mobile identity IE in MME, added support for more PDU types in 
> PDCP, added defines for sizes of fixed arrays in LIBLTE_PHY_STRUCT, 
> fixed the order of sizes for rate match and unmatch arrays in 
> LIBLTE_PHY_STRUCT (thanks to Ziming He), added return values for 
> liblte_phy_get_n_cce (thanks to Damian Jarek) and 
> liblte_rlc_unpack_status_pdu (thanks to Mikhail Gudkov) in the LTE 
> library, added a scoped lock RAII class for sem_t and pthread_mutex_t 
> in the tools library, changed from using boost::mutex to 
> pthread_mutex_t and_sem_t throughout, added helper functions in MAC, 
> changed user deletion and C-RNTI release procedures, properly 
> initializing SIB scheduling info, properly construct MNC (thanks to 
> Mikhail Gudkov), fixed an issue with RTS response in MAC, properly 
> initialized PHY allocation structure in MAC (thanks to Markus Grab), 
> fixed the updating of VT(A) in the RLC retransmission buffer, and 
> changed the QoS parameters for default data in LTE_fdd_enodeb.
>
> The highlights in this release are:
> - Lots of updates from folks using openLTE!!
> - The fix for VT(A) which allows for much more reliable IP traffic.
>
> WARNINGS:
> 1) I have only been testing LTE_fdd_enodeb at 5MHz.  Other bandwidths 
> (specifically 1.4 and 3MHz) are not currently working.
>
> There were several people that provided feedback that went into this 
> release.  In addition, there are still several outstanding patches 
> from people that I'm working to pull in. Sorry for the slower than 
> normal response time.  This is my top priority for the next release.
>
> Please keep providing any and all feedback.
>
> Thanks,
> Ben
>
>
> ------------------------------------------------------------------------------
> Go from Idea to Many App Stores Faster with Intel(R) XDK
> Give your users amazing mobile app experiences with Intel(R) XDK.
> Use one codebase in this all-in-one HTML5 development environment.
> Design, debug & build mobile apps & 2D/3D high-impact games for multiple OSs.
> http://pubads.g.doubleclick.net/gampad/clk?id=254741911&iu=/4140
>
>
> _______________________________________________
> Openlte-discuss mailing list
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/openlte-discuss

-- 
Denis Lapshin
mailto: de...@mi...

[Openlte-announcements] V00.19.01

[Ben W. <bwo...@gm...>]

All,

V00.19.01 is available.  This version added support for all ID types for
the mobile identity IE in MME, added support for more PDU types in PDCP,
added defines for sizes of fixed arrays in LIBLTE_PHY_STRUCT, fixed the
order of sizes for rate match and unmatch arrays in LIBLTE_PHY_STRUCT
(thanks to Ziming He), added return values for liblte_phy_get_n_cce (thanks
to Damian Jarek) and liblte_rlc_unpack_status_pdu (thanks to Mikhail
Gudkov) in the LTE library, added a scoped lock RAII class for sem_t and
pthread_mutex_t in the tools library, changed from using boost::mutex to
pthread_mutex_t and_sem_t throughout, added helper functions in MAC,
changed user deletion and C-RNTI release procedures, properly initializing
SIB scheduling info, properly construct MNC (thanks to Mikhail Gudkov),
fixed an issue with RTS response in MAC, properly initialized PHY
allocation structure in MAC (thanks to Markus Grab), fixed the updating of
VT(A) in the RLC retransmission buffer, and changed the QoS parameters for
default data in LTE_fdd_enodeb.

The highlights in this release are:
- Lots of updates from folks using openLTE!!
- The fix for VT(A) which allows for much more reliable IP traffic.

WARNINGS:
1) I have only been testing LTE_fdd_enodeb at 5MHz.  Other bandwidths
(specifically 1.4 and 3MHz) are not currently working.

There were several people that provided feedback that went into this
release.  In addition, there are still several outstanding patches from
people that I'm working to pull in.  Sorry for the slower than normal
response time.  This is my top priority for the next release.

Please keep providing any and all feedback.

Thanks,
Ben

[Openlte-announcements] V00.19.00

[Ben W. <bwo...@gm...>]

All,

V00.19.00 is available.  This version has several bug fixes and small
changes resulting in a more reliable system.  With this version, I was able
to successfully get end to end IP traffic through the system, including
loading a web page.  The biggest reason for this was switching all DRBs to
RLC AMD mode.  There is still an issue in the downlink, but this change
masks it.  Specifically, removed rb_id from all MME pack routines (thanks
to Przemek for finding this), fixed a bug with the default value of
filter_coeff in LIBLTE_RRC_UL_POWER_CONTROL_DEDICATED_STRUCT (thanks to
Paul Sutton for finding this), added the length indicator for eps_mobile_id
(thanks to Paul Sutton for finding this), added an error return for
dci_1a_unpack (thanks to Joel Carlson for the suggestion), properly
initializing vd_w_metric (thanks for ziminghe for finding this), and
limiting PUSCH allocations to 10 PRBs for performance reasons in the LTE
library, added TX and RX gains to the config file, combined UL and DL
schedule messages into a single PHY schedule message, moved QoS structure
from RB class to the user class, fixed bugs related to the RLC AM TX and RX
buffers, changed the default timer alignment timer to 10240 subframes,
removed boost::lexical_casts from HSS, moved the dedicated bearer QoS to 9,
and moved DRBs to RLC AMD mode in LTE_fdd_enodeb.

WARNINGS:
1) I have only been testing LTE_fdd_enodeb at 5MHz.  Other bandwidths
(specifically 1.4 and 3MHz) are not currently working.

There were several people who contributed to this release (see the notes
above).  Please keep the feedback coming.

Sorry this release took so long to get out, I hope to keep a more regular
release schedule moving forward.

Thanks,
Ben

[Openlte-announcements] V00.18.04

[Ben W. <bwo...@gm...>]

All,

V00.18.04 is available.  This version fixed the start/stop/start bug and
RTS timing bugs in LTE_fdd_enodeb and added UmTRX support to
LTE_fdd_dl_scan.  Specifically, this version fixed long BSR CE, added
extended power headroom CE support, added PDCP data PDU with short SN
support, and added header extension handling to RLC AMD to LTE library,
added UmTRX support to LTE_fdd_dl_scan, fixed the start/stop/start issues
by closing the TUN device on stop, fixed the RTS issues, added a mutex to
the message queue circular buffer, fixed uninitialized variables in the
configuration database, added header extension/multiple data support for
RLC AMD, added detach handling, and added a common routine for formatting
time

WARNINGS for LTE_fdd_enodeb:
1) I have been testing exclusively at 5MHz.  Other bandwidths (specifically
1.4MHz) are not currently working properly.

Please keep the feedback coming.  This release fixed some large
instabilities with LTE_fdd_enodeb, but there still may be more cases that
are not currently working.  When posting issues, please provide the output
of the debug log port and the LTE PCAP file.

Thanks,
Ben

[Openlte-announcements] V00.18.03

[Ben W. <bwo...@gm...>]

All,

V00.18.03 is available.  This version fixed several bugs in
LTE_fdd_enodeb.  Specifically, added more MME encoding and decoding, RLC
UMD header extension support, and higher order modulation scheme support to
PHY in the LTE library, split the error enum into common.h, moved to a new
message queue using semaphores and circular buffers, added IP PCAP support,
split UL/DL QoS TTI frequency, added user reset to RB class, added RLC UMD
multiple data support, added UTC timestamps to the debug log port, added
higher order modulation scheme support in MAC, fixed C-RNTI assign/release
management, changed RTS timing, fixed user switch in MME/RRC, added clear
RB support and fixed copy RB in the user class for LTE_fdd_enodeb.

With the fixes in this release, I am able to keep a UE on LTE_fdd_enodeb
and transferring IP packets in both downlink and uplink for long periods of
time.  I still do not see consistent results with data making through end
to end, but it appears to be related to timing issues with my B200.  For
those of you with GPSDO setups, I am very interested in the results of your
testing.

WARNINGS for LTE_fdd_enodeb:
1) I have been testing exclusively at 5MHz.  Other bandwidths (specifically
1.4MHz) are not currently working properly.

Please keep the feedback coming.  This release fixed several of the
instabilities with LTE_fdd_enodeb, but there still may be more cases that
are not currently working.  When posting issues, please provide the output
of the debug log port and the LTE PCAP file.

Thanks,
Ben

[Openlte-announcements] V00.18.02

[Ben W. <bwo...@gm...>]

All,

V00.18.02 is available.  This version fixed several bugs in
LTE_fdd_enodeb.  Specifically, cleaned up the Time Zone and Time IE and
added C-RNTI support to get_tbs_mcs_and_N_prb_for_dl in liblte, added
asymmetric QoS support, dynamically determining MCS for downlink data,
sending EMM information message, added more time spec information in radio
debug, copying SRB1 PDCP config to SRB2, fixed a UMD reassembly bug, and
did a hack to get around a crash when releasing a C-RNTI in LTE_fdd_enodeb.

With the fixes in this release, I am able to keep a UE on LTE_fdd_enodeb
for a significant amount of time (minutes) and was able to at least get
some data packets through end-to-end (ICMP pings to 8.8.8.8 from a UE get
responses at the UE).

WARNINGS for LTE_fdd_enodeb:
1) I have been testing exclusively at 5MHz.  Others have reported that
other bandwidths (specifically 1.4MHz) do not currently work.
2) Even with the improvements in this release, there are still
instabilities and crashes should be expected.  If you find a reproducible
crash, please compile with -g and -pg, run with gdb, and provide a
backtrace and debug port log.

This will almost certainly be the last release of 2014.  There has been a
lot of progress on openLTE this year, which has resulted in LTE_fdd_enodeb
beginning to be of use.  Thanks to everyone who is using openLTE and
providing valuable feedback.  Please keep this up in the new year!

Thanks especially to Csaba, who is now maintaining the wiki.  Lots of good
content has been added over the last day and it looks like it will be very
useful to newcomers.

As always, any and all feedback is welcome.

Thanks,
Ben

[Openlte-announcements] V00.18.01

[Ben W. <bwo...@gm...>]

All,

V00.18.01 is available.  This version added more MME encoding and decoding
to the LTE library, pulled in a patch from Ruben Merz to add USRP X300
support for LTE_fdd_dl_scan and LTE_fdd_enodeb, added to
enodeb_nat_script.sh, added delayed user deletion, ol extension to all
mesage queues, ability to send EMM information message, and QoS for default
data services to LTE_fdd_enodeb.

WARNINGS for LTE_fdd_enodeb:
1) I have been testing exclusively at 5MHz.  Other bandwidths should work,
but may have issues.
2) Not many improvements have been made to user plane data support in this
release and it is still quite unstable.  Crashes will be common.
3) In order to support user plane data, a script (enodeb_nat_script.sh) is
run at install time to setup NAT using iptables and LTE_fdd_enodeb creates
a TUN device (tun_openlte) at run time to send/receive IP packets.  This
configuration worked well for me, but I can't guarantee that it will work
on every installation.  Please provide feedback if you are having issues or
if you have better suggestions on how to configure this functionality.  In
order to view the IP traffic, it is best to open wireshark and capture
packets from tun_openlte.
4) I had some major USB issues as I was in the middle of testing this
version.  I don't think these issues are due to openLTE, but please provide
feedback if you experience USB issues while using LTE_fdd_dl_scan,
LTE_file_recorder, or LTE_fdd_enodeb.

Future releases will focus on stability of user plane data with
LTE_fdd_enodeb.

Any and all feedback is welcome.

Thanks,
Ben

[Openlte-announcements] V00.18.00

[Ben W. <bwo...@gm...>]

All,

V00.18.00 is available.  This is the first version that supports user plane
data (IP packets) in LTE_fdd_enodeb.  Here are the details of what changed:
Added the liblte prefix to value_2_bits and bits_2_value, converted
everything except PHY and RRC to byte message struct, added more MME
decoding and encoding, added PDCP user plane support, added RLC UMD
support, and fixed a bug in RRC connection reesablishment pack and unpack
(thanks to Paul Sutton for tracking this down) in the LTE library. Added
regeneration of eNodeB security data in HSS, service request, service
reject, and activate dedicated EPS bearer context request support in MME,
timer reset, and IP gateway support, using the latest LTE library, and
refactored C-RNTI assignement and release in LTE_fdd_enodeb.

WARNINGS for LTE_fdd_enodeb:
1) I have been testing exclusively at 5MHz.  Other bandwidths should work,
but may have issues.
2) This is the first release with user plane data and it is not the most
stable.
3) In order to support user plane data, a script (enodeb_nat_script.sh) is
run at install time to setup NAT using iptables and LTE_fdd_enodeb creates
a TUN device (tun_openlte) at run time to send/receive IP packets.  This
configuration worked well for me, but I can't guarantee that it will work
on every installation.  Please provide feedback if you are having issues or
if you have better suggestions on how to configure this functionality.  In
order to view the IP traffic, it is best to open wireshark and capture
packets from tun_openlte.

The next release will merge the patch from Ruben Merz for USRP x300
support.  Future releases will focus on stability of user plane data with
LTE_fdd_enodeb.

Any and all feedback is welcome.

Thanks,
Ben