Add a CU-UP load tester and improve GTP performance
This adds a simple tester program for the CU-UP. It sets up a number of
UEs/bearers in the CU-UP via the E1 interface, then streams data in
downlink/uplink (as seen by the CU-UP). More information is available in
tests/nr-cuup/README.md.
Includes some fixes to properly verify that SCTP is installed, and use
"the right cmake way" to use it in cmake code (see commit for details).
Remove these F1-C port numbers. The configuration is not actually
applied, and the default F1-C port number 38472 (see F1AP_PORT_NUMBER)
is hardcoded in the F1 module.
For the moment, I leave the configuration option, in case we want to
allow to configure the port numbers later).
Now that we reliably receive SIB19, we can set the ntn-UlSyncValidityDuration-r17 to a sane value
and thus make sure that the UE always has up-to-date SIB19 information.
Fix Dedicated BWPs
- Improve SearchSpace selection
- Improve CORESET selection
- Fix fill 'initialDownlinkBWP->pdsch_Config' when there are
DedicatedBWPs
* This fixes the reported 'RSRP = 0' issue when the UE is in the
InitialBWP and we have DedicatedBWPs configured
- Add a 'common' type SearchSpace in the NR_BWP_DownlinkDedicated IE for
DedicatedBWPs
- Fix hardcoded 'controlResourceSetId' by making it dynamic
- Remove 'searchspaceid' from 'verify_agg_levels' function because it
was used only for LOG
- CI: Add DedicatedBWP to '40 MHz TDD F1+E1 SA' in
'RAN-SA-B200-Module-SABOX-Container'
Following frequency change in the pipeline, we observed degraded UL channel
conditions resulting in low UL throughput. Increasing attenuation on the gNB
side was found to mitigate this issue by improving UL channel quality
and reducing UL noise power.
Increase tested UL and DL bitrates in OAIUE pipeline.
bugfix: read msg_type only after deciphering
There was a problem with accessing msg_type from the input buffer before
deciphering, giving obviously wrong value.
CI team: we should add a test in the CI with NAS ciphering + integrity
(connection to the core) set to not null (I suggest nea2/nia2). With
openair's core, this is done by setting this in the amf config file
(nia1 and nea2 at the top of the list):
supported_integrity_algorithms:
- "NIA2"
- "NIA1"
- "NIA0"
supported_encryption_algorithms:
- "NEA2"
- "NEA1"
- "NEA0"
And also put full security in the AS. This is done with this in the gnb
config file:
security = {
ciphering_algorithms = ( "nea2", "nea0" );
integrity_algorithms = ( "nia2", "nia1", "nia0" );
drb_ciphering = "yes";
drb_integrity = "yes"
};
Ideally we should test this with both openair UE and cots UE. But as a
strict minimum do for openair UE (this is where the bug was).
No need to do much traffic, a few ping is enough. But if you want to
check throughput with full security, why not.
Enabling usage of ulsyncvalidityDuration timer in SIB19
ntn-UlSyncValidityDuration in SIB19 from Spec 38.331 v17.12 A validity
duration configured by the network for assistance information which
indicates the maximum time duration (from epochTime) during which the UE
can apply assistance information without having acquired new assistance
information. The unit of ntn-UlSyncValidityDuration is second. Value s5
corresponds to 5 s, value s10 indicate 10 s and so on.
ntn-UlSyncValidityDuration-r17 ENUMERATED{ s5, s10, s15, s20, s25, s30, s35,s40, s45, s50, s55, s60, s120, s180, s240, s900}
Implementation of epochtime is under discussion.
It can be configured in configuration file according to the above
enumeration. For GEO - 240s, MEO - 20s, LEO-5s can be used.
Following a power cycle of the CI machines in the NSA-B200 setup, the gNB
machine encountered a networking issue. Although packets were observed leaving
and returning to the host of gNB, they failed to reach the container.
Debugging steps taken:
- upgraded the gNB machine
- reinstalled Docker
- reviewed and verified IP rules and routing settings
Setting the container to use host network mode resolved the issue.
Improve radio for Metanoia RU (and likely others)
We introduced a regression that lowered the achievable DL throughput
with at least the Metanoia RU (i.e., in the configuration and the UE
that we use). git bisect showed up one commit, which does not seem
necessary. We just revert this commit, as it improves the throughput and
does not seem to have a negative impact on performance.
Further, increase RLC buffers a bit. Limit 7.2 pipeline number of layers
to 2.
With PMI reporting working (see HEAD^^), it seems that four layers put
two much processing strain on the machine (cacofonix), which asserts
immediately. In tests, it was found that the same test throughput
(600Mbps) can be achieved with two layers, while not crashing the
machine.
CI: Replacing Devkit server with a Grace Hopper in Aerial pipeline
Switch servers on the Aerial pipeline from Gigabyte (AMD) to Grace
Hopper (ARM), and make the necessary changes:
- Update XML, YAML, and gnb.conf
- Change the argument passed to the L1 in aerial_l1_entrypoint.sh to be
compatible with the GH
- Update the Aerial setup in TESTBench.md and Aerial_FAPI_Split_Tutorial.md docs
- Decrease the tested throughput due to an issue with the RU. Will open
another MR to increase it once the problem is fixed.
- Update the N2 and N3 IP addresses to match the Grace Hopper's in the gnb.conf.
- Remove the test XML file for pulling images in the ARM image builder.
- Update the Aerial XML file to pull ARM images.
- Decrease tested throughput (temporarily) due to problems with Foxconn RU
- use Ubuntu image, instead of RHEL image - gNB fails with "Illegal
instruction" during RA, when running with RHEL image
- init DPDK in VA mode
- increase number of cores Tpool
fhi_72: allow to provide dpdk_iova_mode to xran
- Introduce a gNB configuration option that enables users to select the
desired IOVA mode for DPDK ("PA" or "VA") with dpdk_iova_mode
parameters in fhi_72 configuration section. If not set, the default
remains PA mode to ensure backward compatibility with existing
configurations. Detailed information about IOVA modes, including their
advantages and disadvantages, can be found in the DPDK documentation:
Memory in DPDK Part 2: Deep Dive into IOVA
- Using IOVA VA mode allows DPDK to run without requiring SYS_ADMIN
capability for the docker container.
- Switch to IOVA VA mode in the FHI7.2 pipeline and reduce the required
capabilities for the docker container.
Jaroslava Fiedlerova