Integration `2024.w18`
See merge request oai/openairinterface5g!2713
* !2705 Fix CRC check in UE DCI detection
* !2708 Adding few more NR bands to the list of supported ones
* !2701 gNB archi fixes
* !2203 DL-MMSE
* !2684 remove asserts on min RX/TX, set the RX/TX shift as 2 in the UE
* !2594 F1AP refactoring to align with O-RAN.WG5.C.1-v11
Taking the example of the AMF Set ID, the previous version of the macro
used another macro to encode the number, then marked the last 6 bits in
the bit string (of length 16/2 bytes) as unused. This is wrong, because,
assuming AMF Set ID=3, it basically truncates the number by encoding all
16 bits and cutting off the last 6, like this:
------------------- encode 16 bits
0000 0000 0000 0011
------------ take these 10 bits
So what remains is these bit positions of the number:
(16,15,14,13,12,11,10,9,8,7) -- the last 6 bits are cut (and the first 6
are either 0 or garbage, because they don't exist).
Instead, manually encode bits (10,9,8,7,6,5,4,3) in the first byte, and
(2,1) in the second byte of the bit string, like so:
------------ encode 10 bits
0000 0000 1100 0000
------------ take these 10 bits
Do the same with the pointer.
The AMF Region ID above is correct, as it is exactly 8 bits.
In the case of PDU sessions in the NG UE context setup request, we first
have to set up security. Hence, in this commit, implement the "delayed"
set up of theses PDU sessions, and acknowledge the setup request after
reconfiguration, as requested by O-RAN.WG5.C.1-v11.
Trigger UE ctxt setup response for the case of "registration request"
(no PDU sessions in UE Context setup response), as mandated by
O-RAN.WG5.C.1-v11.
As of this commit, since we now send the UE security command as a DL NAS
message, there is no F1 UE Context setup request. Hence, instead of
sending a UE context modif request, send the UE context setup request
after the E1 bearer setup response (which logically also makes more
sense, as the E1 setup procedure comes before the F1 setup procedure).
This reverts commit 0f100a6e94, which
introduced a hack to possibly wait with a PDU session setup at the RRC
in case a RRC UE capability transaction (requesting UE capabilities from
a UE) was ongoing. This happened, as we delayed the UE capabilities to
after the first RRC reconfiguration; in that case, certain UEs were
requesting the next PDU session, and if RRC did not delay the new PDU
session (as requested from the core), this procedure might occur while
UE capability enquiry was ongoing, leading to failures in these
transactions.
The default RRC reconfiguration was previously sent after the security
mode command as a "first" RRC reconfiguration. However, it is simply not
needed, as it will be triggered through a subsequent reconfiguration
that also sets up DRBs.
Move the Measurement Config to the "other/dedicated" RRC
reconfiguration.
This reconfiguration would have forwarded a NAS PDU (typically a
registration accept). This is now done by a dedicated forwarding of NAS.
This reverts commit 4a7d7975ce.
Trigger the UE capability right after security mode complete, as
specified in O-RAN WG5.C.1-v11. Also, there is no "need" for the
"default" RRC Reconfiguration (it will come once PDU sessions are
requested), so remove this as well.
Previously, the CU sent the Security Mode Command as part of a UE
context setup request. This was done "because it was possible", not
because there was an inherent need to do this. However the LiteOn DU
does not like this, as it expects to also have a DRB in the UE context
setup request procedure, which is not always the case.
Hence, send the Security Mode Command in a normal DL RRC msg transfer
over F1. As of this commit, there is not UE context Setup Request (so it
might not work with all DUs), but the OAI DU is cool and does not care,
so RFsim still works.
This also aligns the CU's behavior with O-RAN.WG5.C.1-v11.
Finally, as of this commit, we do not trigger a UE context setup
request, so we cannot handle PDU sessions inside the initial UE context
setup request at the same time as the security mode command (which was
done previously before reaching this point). This will be fixed in a
later commit.
In the DU, read slice information from the DU config file. Handle
multiple slices within the F1 Setup Request structure representation,
and forward it via F1 to CU. The DU slice information is stored in the
setup request structure.
This reverts commit eaf28414ca.
This assertion was introduced as we (wrongfully) thought we needed to
calculate the SSB ARFCN at the CU. Since PBCH might carry one bit, which
is not available at the CU, only certain SSB ARFCNs would have been
possible.
With the availability of the SSB ARFCN through the
MeasurementTimingConfiguration, this is not necessary anymore.
MeasTimingConfig may not come with F1 Setup Request: some DUs send it
with the DU configuration update. Handle this gracefully by checking for
MTC before using it.
Use the correct MIB data structure, MIB_t. This is required by 38.473,
and the LiteOn DU sends only that. So our CU needs to handle it, so
align the OAI DU as well.
Integration `2024.w17`
See merge request oai/openairinterface5g!2702
* !2659 NR bandwidth index fix
* !2693 Harmonize frequency range structures
* !2699 Add support for USRP X410 to run with 200 MHz bandwidth in FR2 at 120 kHz SCS
* !2692 Fix for the overflow issue while processing GPS based timestamp from RU
* !2665 NR UE improvements in handling RRC Release
* !2682 dci11type0
* !2686 CI: Maintenance, fixes, improvements
* !2677 NR UE trigger RA for SR failure
* !2683 Ue small fixes
* !2668 Refactor PDCP Reestablishment and introduce PDCP Suspend at gNB
* Add rhel9.4 to the list of OAI supported distributions
* !2673 speedup integrity computation
* !2706 doc: Update of CI testbenches
- according to 5.3.7.4 of TS 38.331: 1> re-establish RLC for SRB1;
- gNB:
1) re-establish RLC for SRB1 in nr_rlc_update_id
2) re-establish RLC for remaining RBs concurrently with RRCReconfiguration
issue: gNB) in nr_rlc_update_id the gNB is re-establishing RLC for all
RBs at the same time when receiving RRCReestablishmentRequest
from UE
UE) according to the specs, the UE is re-establishing RLC for
SRB1 with RRCReestablishmentRequest the other RBs are
re-established during RRCReconfiguration (in our case in
nr_rrc_ue_process_RadioBearerConfig) when receiving
reestablishRLC IE in rlc_BearerToAddModList
RLC TX) this leads to RLC counters mismatch after re-establishment:
control PDUs discarded until max nb of retx is reached
and then UL failure occurs
By using a struct we may have warnings at compilation time in some
cases if the type is not used properly.
The type definition becomes a bit weird, so we need to properly
comment.
If too strange, we can revert to the classical "void" definition.
- rename "lib_ctx" to "mac_implementation" which is what is stored because
of previous commit
- rename "mac" to "mac_context" to be clearer with what is stored in this
variable too
The main work of this commit is to rewrite cipher_aes_128_cbc_cmac()
which was too slow and created issues when running the 4G eNB on some
machines (LLL when using a b210).
Instead of creating the security context, use it, and delete it, all in
cipher_aes_128_cbc_cmac(), we introduce the function init_aes_128_cbc_cmac()
to create it, and we change cipher_aes_128_cbc_cmac() to just use it,
and we also adapt free_aes_128_cbc_cmac(). The function aes_128_cbc_cmac()
is removed, it does not follow the new init/use/free API.
Doing so, on the primary test machine used to do this work (a powerful
and fast machine where it is possible to set the CPUs' frequencies)
it was possible to limit the CPUs' frequencies to 500 MHz (it was not
possible to go lower) without any LLL (using a b210, quectel UE in
idle, no data traffic, only periodic measurement reports). Before this
commit, it was necessary to set a frequency of at least 1.8 GHz
otherwise some LLL would appear.
Then we need to adapt the rest of the code to take into account those
changes.
A notion of security container (stream_security_container_t) containing
ciphering and integrity contexts is introduced.
And the code is harmonized for, hopefully, better readability.
The (opaque) type stream_security_context_t is introduced, used by
both integrity and ciphering contexts. The idea is to use a
init()/use()/free() API for all the security procedures, identical for
all the ciphering and integrity procedures.
These functions enables (un)deployment of the OAI CN v2.0.1, increase
reliability and speedup overall CN (un)deployment compared to current
implementation.
Use "--wait" helm option to wait for deployment/undeployment of the CN.
Wait for pods to be in a ready state before marking the release as
successful. If not deployed/undeployed before timeout, marked as
unsuccessful.
Hardcode path for OC CN5G deployment/undeployment
- Remove unused AmariUE config file, replace by new one - for PLMN 00102
- Change frequency of AW2S pipeline - Move center frequency by 40 MHz to avoid interferency with other setups
in the lab. Set center frequency to ~ 3450 MHz.
The absslot_tx overflow issue observed during integration testing of AW2S
RU with GPS as the source for the clock synchronization. Causes the gNB
not to start the TX thread and stops the processing.
Integration `2024.w16`
See merge request oai/openairinterface5g!2689
* !2680 CI: Modification of log collection in UndeployObject()
* !2681 remove a useless copy and specific buffer for all UE UL payload
* !2685 Clang: make executable run, fix clang warnings, fix memsan warnings
* !2690 Remove hardcoding of 5G-S-TMSI on nrUE
After the recent update of matix, collected logs contain color codes
and we did not find the way to switch them completely off.
Replacement of "docker compose logs" by "docker logs" resolves this
issue. This commit includes changes related to "docker logs" usage.
Reset all the circular buffer to zero; this ensures that we cannot read
uninitialized data. Since this is RFsimulator, performance is not a
problem here either (and zeroing out happens only once).
Memsan complains that uninitialized reads originate because of these
warnings. Most of the corresponding variables are actually read into
through by libconfig; nevertheless, initializing them should not have a
bad side-effect (because the initialized value is overwritten).
We access G_N at indexes that are not necessarily 32-byte aligned,
leading to segfaults. Use simde_mm256_loadu_si256() to load from memory
at an unaligned memory address to avoid the segfault.
Discovered using clang; it is unclear why gcc did not encounter the same
segfault.
-rdynamic is necessary for clang as well to properly link shared
executables. Add furthermore an option to ignore unused command line
arguments, to suppress warning
clang: warning: argument unused during compilation: '-rdynamic' [-Wunused-command-line-argument]
This is likely because -rdynamic is only relevant for the linker, but
the corresponding option add_link_options() is only available starting
cmake 3.13, while we are still at 3.12 (for Ubuntu 18, to my
understanding).
- remove PDCP reestablishment for DRBs from RRCReestablishment message generation
- do PDCP reestablishment and CUUP notification calls after reception or RRCReestablishmentComplete
- this is necessary to make it consistent with the RRCReconfiguration message that contains
RadioBearerConfig with the list of bearers to add/release
which is followed by PDCP re-establishment at UE
- SRB1 shall be re-established at the time of RRCReestablishmentRequest
according to 5.3.7.4 of 3GPP TS 38.331
- SRB2 shall be re-established when creating srb-ToAddModList
according to 5.3.5.6.3 of 3GPP TS 38.331
Integration `2024.w15`
Closes#727, #765, #776
See merge request oai/openairinterface5g!2676
* !2561 CI - add T2 tests to timing pipeline
* !2620 Fix cppcheck errors
* !2624 NR UE MAC SR improvements
* !2647 NR UE mixed fixes
* !2671 Reduce stack memory size for layer mapping
* !2669 CI: create SA 2x2 pipeline
* !2660 fix for 3072 fft size when using 80 MHz and -E
* !2539 CI: Add channel emulation in one test
* !2635 SCTP problems: allow to bind separate F1-C/F1-U interfaces; allow to use DNS
* !2438 PSBCH RX,TX and SLSS SEARCH procedures
* !2675 (doc): Update benetel firmware version and configuration files
* Update of O-RAN 7.2 gNB configuration file for Benetel 650
Unlike the previous commit, in noS1, we cannot use Attach(), as noS1
automatically brings up an interface, no matter whether the connection
succeeded or not. Therefore, consistently wait a bit longer to increase
the chance that the radio is really up.
For various reasons, the UE might take longer to connect to the eNB. A
simple sleep might not be enough; instead, use Attach() to actively wait
that the UE gets an IP address. The added bonus of this function is that
it will also stop&restart a UE if it did not attach, i.e., it will try
multiple times to attach the UE.
The code (author) confused MCE_PARAMS_DESC and MCE_NETPARAMS_DESC and
tried to read the MCE name from MCE_NETPARAMS_DESC when in reality it
should come from MCE_PARAMS_DESC. This commit fixes to use the right
array.
In a previous commit, the functionality of handling IPv6 (and DNS names)
as part of the IPv4 was introduced (in IPv4 field for backwards
compatibility). This commit now removes superfluous configuration
fields:
- IPv6: handled by IPv4 field
- active: users can just remove an entry
- preference: not needed
- Update the configuration for benetel 650 and 550 for new firmware version
- Remove the old firmware configuration file and description in the document
- Add a new RU configuration section
- Add How to get support from the OAI Community to clarify the support procedure.
Rewrite sctp_create_new_listener() to use getaddrinfo() to look up
IPv4/IPv6 addresses via a common interface. This allows to use e.g. DNS,
and enables the SCTP library to listen on IPv6 ports.
To support the above, the corresponding message is changed to not take
an IPv4 address on 4 bytes (and separate IPv6), but a single name that
will be resolved via getaddrinfo()
The previous init message used to allow multiple IPv4/IPv6 addresses;
since this is not used anywhere, remove the corresponding functionality.
Modify all uses of this ITTI message to copy te SCTP init bind addr into
ITTI message.
Use getaddrinfo() to look up IP (v4/v6) addresses. The old
interface of separately specifying IPv4 and IPv6 addresses is
preserved, but both fields allow to specify types of addresses, and we
will remove one field in the future.
After the last commit, it is not possible to bind on separate interfaces
for F1-C and F1-U. In other words, it is for instance not possible to
have the F1-C on one subnet on one physical interface, and F1-U on
another subnet on another physical interface. This commit introduces a
second option MACRLCs.[0].local_n_address_f1u to allow to bind F1-U on
an alternative address; if it is not specified, the interface of F1-C is
reused.
The above only works for the DU; no corresponding option exists for the
CU. For the CU, use separate CU-CP and CU-UP.
Use it in the CI.
This reverts commit 284116522b.
The problem is that (1) we don't really support multi-homing, but (2) on
some system, if the other IP addresses that the operating system
considers part of the multi-homed connection are HEARTBEAT'ed (e.g.,
docker), they answer with SCTP ABORT, which breaks the entire
connection.
In other words, force single-homing by binding. For F1-U uses, another
option will be introduced that allows to bind to a separate interface
for F1-U (separate from F1-C), if necessary.
Closes#727
- adding a new test case
- the test case simulate a DL radio channel with noise and runs Iperf test with such channel conditions
- use CustomCommand class to send commands to a Telnet server
- switched from netcat’s unsupported -N option (OpenBSD) to --send-only (Ncat 7.94) for compatibility with CI
re-sync CI test in F1 RFSIM
- to be used in the future, for CI of RRCReestablishment in RFSIM
- adding four new test cases (040021, 040022, 040023, 040023) to the RFSIM F1 configuration
- the test cases simulate a disruption of the radio channel, verify UE sync loss and resync
and restore the original channel conditions
- use CustomCommand class to send commands to a Telnet server
temp xml
- this change simplifies the enum definition and makes it easier
to add, remove, or modify states in the future
- the states are now defined in the NR_UE_L2_STATES macro,
and the enum members are generated automatically from this macro
Add ulsch decoding measurements, which are reported by nr_ulsim. Start ulsch_decoding_stats measurement before
nr_ulsch_decoding is called and stop it after collection of all decoder tasks from Tpool. In the context of nr_ulsim,
ULSCH_id is allways == 0, testing is performed per one TB, therefore measurement is started and stopped only once per trial.
Reported ulsch decoding statistics show the mean ulsch decoding time over x trials (specified by -n command line option of nr_ulsim).
This measurement does not provide valid information in case of nr-softmodem, when the loop is done over multiple ULSCH IDs -
start of the measurement is done multiple times and stops only once in this case.
For unitary physical simulators, ULSCH processing loop is executed only once
- the max number of PUSCH is set to 1. This ensures exactly one start and stop
of the ULSCH decoding time measurement per processed TB.
Note: ulsch decoding statistics reports average time of segmentation, rate matching, interleaving and decoding of the TB
- Enable both nr_ulsim and nr_dlsim test with Run_T2Test
- Introduce time threshold variable for T2/CPU LDPC tests
- Set threshold individually for each test - T2 encoding/decoding and CPU encoding/decoding
- Check if encoder/decoder processing time is below the threshold
- Remove NOK return option - keep only OK (processing time below threshold) or KO (proceesing time beyond threshold)
- Rename function to run and check results of CUDA ldpctest
- Use new SSH class in function for check T2 test results, use f-strings
Integration `2024.w14`
Closes#775 and #773
See merge request oai/openairinterface5g!2663
* !2411 Simple and generic trx_write support of out of time order requests
* !2532 moving decision on DCI format type to UE MAC
* !2565 improvements in security of nrUE
* !2656 remove stored G in UE DL, not needed and creating issues with DCI format 11/type0
* !2658 Fix sleeping ue at synchro
* !2649 NR SRS RRC assrtion removal
* !2662 Avoid concurrent write while filling RX/CRC.indication
* Remove obsolete "version" parameter from yaml files used on avra
* !2664 Remove duplicated NAS message send, already done above
* !2605 Improve performance polar encoder i.e., 3GPP 38.212 5.3.1.2 function
* !2629 Msg2 and Msg3 at gNB: check allocation before committing to Msg2
* !2657 replace transient malloc by stack allocation, add defense code on block size
* !2631 chore(ci): some docker improvements
* !2654 Pucch2 polar rx dynamic range fix
* !2666 [E2 Agent] Fix E2 Setup Request for ASan
* !2661 CI: iperf/iperf3 UDP analysis function fix → review in progress, review to be completed
Simplification of the code by replacing iperf_opt by target_bitrate (returned from Iperf_ComputeModifiedBW) - no need to repeat the steps to retrieve information about requested bitrate from iperf_opt. Compute iperf bitrate performance as receiver bitrate over requested (target) bitrate.
Fix an issue with misdetection of missing report line - "result" was defined 2x in the function. With this MR, result can be only None or stores output of re.search() for matching the report line pattern, which ensure proper detection of the report line as well as presence off all required components in the report line.
Sender bitrate is not always equal to the requested bitrate (set by "-b" iperf option) in iperf3 (v3.9) UDP tests, so we have to compture bitrate performance of UDP iperf3 test as receiver over requested bitrate. Bitrate performance then express percentage of requested bitrate, which is received on the receiver side.
For iperf3 (version 3.9) UDP, sender bitrate is not always equal to requested bitrate (defined by '-b' option). We need to retrieve information about the requsted bitrate from iperf command line options specified for each UDP testcase. This parameter is then used in UDP analysis functions for proper evaluation of the test results.
Prior to this commit, it might happen that the gNB used TDA 0 (Start 0
len 13) for mixed slot, in which case the UE could not send msg3
(because a mixed slot has less than 13 symbols).
fixes the overflow issue in the PUCCH2 RX with 4+ RX antennas and 12+
bits (polar code format). The fix scales the polar decoding LLR
computation according to the number of RX antennas. The issue was clear
for 100 MHz 4 RX antennas. Even
nr_pucchsim -s -5 -q 8 -P 2 -b 12 -R 273 -z4 -n1000
doesn't converge to 0 BLER with increasing SNR. The issue didn't appear
in CI because of the bandwidth and antenna count limitations.
- docker on CI machine avra was updated to version 26.0.0, which reports
parameter "version" as obsolete
- this additional log causes fail of the container undeployment to stage
We observe these asserts while running concurrent UL traffic on many
UEs:
Assertion (crc->rnti == rx->rnti) failed!
In handle_nr_ulsch() ../../../openair2/NR_PHY_INTERFACE/NR_IF_Module.c:193
mis-match between CRC RNTI 7221 and RX RNTI 35cf
CRC and RX indication need to be pairwise for one UE, when here we have
each for different UE. However, investigation seemed to indicate that
both arrays for CRC and RX indication contained only one PDU; it seemed
the RNTI changed while writing the CRC/RX.indication messages in
nr_fill_indication().
In this patch, we change the logic to mark a ULSCH free AFTER calling
nr_fill_indication(). It is difficult to say if this makes the assertion
disappear, but preliminary tests did not exhibit the same failure
states.
Closes#773
Move decision of Msg2 allocation time instant from instance we receive
RA indication (only containing preamble time information) to actual RA
scheduler (containing current DL slot time). This has the advantage:
- if Msg2 cannot be sent, we might try in a later slot
- no timing-based problems were a RACH.indication FAPI message might
come late, we calculate a Msg2 frame/slot, but this instant has
already passed in the DL scheduler
- no assumption about slot_ahead (an L1 constant)
The previous code calculated and verified an SSB index; I am not sure
this code is correct or used. I leave the function
ssb_index_from_prach() (hence the gcc attribute), but remove any checks.
Similarly, the beam association list might be wrong. I deleted the
corresponding code.
Add an API to allow out-of-order writes to a radio interface, unlike the
current interface.
this function is called in UE in this commit, even if the tx might
still be in order.
Later, i plan to improve UE pulti-threading, that will lead to out of
order tx. also, the gNB will benefit of this new function to replace
several specific pieces of code that reorder tx with over complex and
slow systems
Introduce the structure nr_pdcp_integrity_data_t and adapt code to use it.
Note: in nr_pdcp_sdu_t we keep the 'count' variable (which is also present
in nr_pdcp_integrity_data_t). They represent the same value, but 'count'
in nr_pdcp_integrity_data_t is to be specifically used for integrity while
the other one has other uses, so it's better to keep it.
The case securityModeFailure didn't seem to be implemented
properly, so I just removed it entirely.
The variable 'securityMode' did not make much sense, removed as well.
Plus for integrity, there was: securityMode |= 1 << 5 for nea1
and << 6 for nea2, which does not seem correct (I would expect << 4
and << 5 respectively), so it was properly incorrect.
'securityModeCommand->criticalExtensions.choice.securityModeCommand'
was accessed before checking that
'securityModeCommand->criticalExtensions.present'
is 'NR_SecurityModeCommand__criticalExtensions_PR_securityModeCommand',
which is wrong.
The tests 'securityMode >= NO_SECURITY_MODE' and 'securityMode != 0xff'
don't make sense/are unclear, so removed too.
So let's simplify this function, wrong in several places. And put
some AssertFatal() so that the code won't do weird things. The
AssertFatal() can be removed later and the function improved later
if needed. The case securityModeFailure can also be handled later,
but I think there is more work to do than just encoding the message
and send it to the gNB, so it's not bad to remove it for the moment.
The use of security_mode_completed in the PDCP entity was a hack.
Plus it was not working at all with the deregistration request
when nea2 is used, sent when doing ctrl+c in the nrUE.
So let's remove it.
And let's handle activation of integrity and ciphering more in
accordance with what 38.331 says. SecurityModeComplete has to
be sent integrity protected, but not ciphered. Only after should
the ciphering be activated. (See 38.331 5.3.4.3.)
Integration `2024.w13`
Closes#696 and #762
See merge request oai/openairinterface5g!2653
* !1911 Preparatory work for NR DL 4-layer MIMO at gNB
* !2610 fix ug on pilot 2 in nr_pbch_dmrs_correlation() that is lacking signal...
* !2639 NR gNB PUCCH2 CQI workaround
* !2642 NR UE demoting to LOG_D another log that cause flooding
* !2644 NR UE prevent segfault at detach
* !2645 add error code for pdu session reject in the ue
* !2600 Handle Msg3 with SRB1 RRCReestablishmentComplete and MAC CE C-RNTI
* trigger RF-sim E1+F1 test without asan
* !2636 Fix various F1 problems
* !2641 NR DL and UL channel estimation fix
* !2652 Provide additional troubleshooting documentation for QtScope
* !2646 NRUE fix asn_sequence_del to clear a list
* !2504 CI: use iperf3, refactor Iperf_Module
* !2617 use UE ID instead of RNTI internally in NR RLC
* !2619 Minor PDCP fixes
* !2648 Minor CI fixes, stack protection, fix E1+F1 Quectel test, L2sim5G test
From 3GPP TS 38.321 Section 5.4.3 Multiplexing and assembly, we need to process the highest priority subPDUs first:
- MAC CE for C-RNTI is the one with higher priority
38.473 section 8.2.3.1 says that upon F1 Setup Request, UE associated
signalling should be erased. In this commit, we trigger a release of all
UEs that were still connected to a DU when we received the SCTP
shutdown, i.e., when the DU disconnects.
- Four mounting of files, we need the absolute path. Even if the script
gets relative paths, realpath will give the absolute path
- Correctly recognize and handle gNB images with AW2S support
- no need to specify "single-ue" profile when running iperf test for 1 UE only
- default profile "balanced" can be used instead, computes BW for 1 UE correctly
Integration `2024.w12`
See merge request oai/openairinterface5g!2637
* !2640 Fix UPF crash at startup after latest Docker and Ubuntu updates
* !2579 Add telnetsrv_rrc for triggering RRCRelease
* !2613 NR UE fix for procedure after T300 expiry
* !2625 Changes to support 32 UEs in the gNB.
* !2632 NR fix for initialization of gold sequences for DMRS
* !2588 fix(doc): add ptp4l service configuration file in 7.2 documentation
* !2630 Move BladeRF config to right subfolder
* !2626 On RFsimulator end, close open connections
* !2628 NR UE fix double sync
* !2634 feat(phy-test): disable CQI and RSRP errors when running in phy-test mode → review to be addressed
* !2621 CI: set memory limit for docker build cache
* !2604 NR UE BSR improvements → waiting for final changes
* !2638 NR UE data scrambling fixes
* !2602 [E2 agent] KPM & RC SMs improvements
- RAN Function Definition implemented
- ran_func_kpm.c - use function pointers based on the matching condition,
and node type
- ran_func_kpm_subs.h/c - for measurement collection within
MAC/RLC/PDCP layers; for CU-UP node only PDCP included (otherwise,
we would get undefined reference to MAC/RLC)
- use mutual E2SM UE ID file
In phy-test, the scheduler has a hardcoded UE allocation; if CQI and RSRP
are not evaluated, there is no change in the scheduler function. Also,
this avoids misleading error messages if no UE is connected.
- The DL MCS adapted based on the BLER for the minimum number of
DL(numd_sched) grants for the UE in the pre_configured number of frames.
The current threshold value is set to be >9 for 16 UE capacity.
- With increased number of UE to 32 and when all the UEs are DL data active
the num_dl_sched is not meeting the threshold value. Due to whcih the MCS
of all the UEs are stuck with MCS 10-12 in good channel condition.
- The threshold value is adjusted to support 32 UE capacity increase.
- Added the limit as MAX_DCI_CORESET for max_sched_ues in DL/UL scheduler.
This is needed because gNB crashes when tested with 100MHz cell as the
scheduler tries to schedule 11 UEs in a slot and the DCI informaiton is
corrupted due to the limitation in NFAPI interface because of
MAX_DCI_CORESET size.
Use the correct function in order to close open RFsim connections,
instead of a simple close(). If this is not done correctly, it might
happen that the softmodem closes, but an L1 thread might still read
samples (bug in higher layers). Then, RFsim tries to read a socket
(because it has an open connection), but the actual connection had been
close()d. This commit fixes this behavior.
Integration: `2024.w11`
See merge request oai/openairinterface5g!2623
* !2435 Add a sequence container in OAI
* !2552 NR UE RRC re-establishment
* !2611 RRC: ignore random value during RRC Setup Request
* !2587 NR UE RRC handle MeasConfig
* !2616 NR UE remove unused phy_config_ind queue
* !2580 Improve function that map ro to ssb in the UE: make dynamic smaller memory...
* !2614 F1AP: move GTP from enc/dec to logic handlers
* !2622 gNB LOG with UE command line parameters
* !2595 remove 5G dependency on 4G mac.h include, separate RA state enums between 4G/5G and UE/gNB
* !2615 Enhancements to traffic flow in SW architecture documentation
The F1AP message encoders and decoders should do only that, encoding and
decoding messages. For the actual logic handling, including set up and
teardown of bearers, we should use the handlers.
In this commit, delete GTP tunnel handling from encoders/decoders. For
the CU, this happens already in the CU-UP/E1 handlers in
cucp_cuup_handler.c. For the DU, add it in the handlers in
mac_rrc_dl_handler.c.
From the spec, it does not seem to be mandated to check that the random
identity needs to be unique within a cell. Also, from experience we see
that this only happens when the same UE "comes back"; so since one UE
can only have one connection, the other RRC UE contexts (with the same
identity) should be removed at some point, anyway, leaving only one with
the same identity.
Integration: `2024.w10`
See merge request oai/openairinterface5g!2608
* !2427 Scan build: Result of operation is garbage or undefined
* !2477 cleanup of nfapi_nr_interface.h and more
* !2523 CI: AW2S pipeline - test with 16 UEs including RAN code fixes
* !2530 Rework of MAC UE handling of DAI
* !2583 Small-bug-from- cdb18d38bfa6aee288a68d9e8881615a48addd83-and-one-gcc-warning
* !2592 NR DMRS RX comments
* !2599 Log module small cleaning
* CI: mark as failed if UE log analysis fails
* !2465 NVIDIA Aerial Integration
* !2409 Ue rework keep sync in frame
* !2603 tx_timestamp computation doesnt count slots from rx timestamp, and a small...
* !2607 preventing negative RX absolute slot in NR softmodem
* !2606 Instructions on how to debug using container images and core dump of CI
* !2609 Upgrade Open Cells Project UICC/SIM programming tool and rename FR2 SA configuration file
* fixup after merge of Aerial
Some places (e.g., 4G eNB/UE nfapi mode) use numbers for the FAPI mode,
directly mapping to an enum. Aerial being in the middle messes up these
numbers. Put it at the end to limit possible problems.
nr_fill_indication() fills CRC and RX requests, to be sent from L1 to
L2. Before this commit, nr_fill_indication() is called when
- we finally decoded LDPC (or not)
- low signal on PUSCH
This might happen at the same time, i.e., nr_fill_indication() might try
to fill both CRC and RX, in the same list, from different threads. This
can lead to this assertion:
Assertion (crc->rnti == rx->rnti) failed!
In handle_nr_ulsch() ../../../openair2/NR_PHY_INTERFACE/NR_IF_Module.c:190
mis-match between CRC RNTI e071 and RX RNTI 5e3b
e.g., e071 is low energy, and while 5e3b related message are filled in
one thread, e071 is being put into the message structure as well.
At least that is my understanding. I could not actually reproduce this
assertion; to be seen if it still happens.
Do not assert if the list of phy statistics is full; instead return
NULL.
In most places, this is handled appropriately (we do not store anything
if we can't). In nr_decode_pucch0(), the return value of get_phy_stats()
is used a little bit all over the function, and a constant
if (uci_stats)
uci_stats->variable++;
would look ugly, so make uci_stats point to the stack if we cannot store
in the phy_stats.
the uci_stats variable (pointing to per-UE statistics for UCI) is used
in function nr_decode_pucch0() to store information which is important
beyond statistics, e.g., uci_stats->pucch0_thres as a threshold (which
comes from somewhere else)
This commit refactors to use original or intermediate variables to not
mix uci_stats usage with actual processing.
Also, group uci_stats a bit more.
They use 4G NUMBER_OF_UE_MAX, which is not supposed to be present in 5G.
Further, the mutex and variable are global state; for Paging, we should
use F1.
We can only support up to 30 UEs for SRS. In this commit, we change:
- if the total number of UEs is > 32, throw error during compilation to
raise awareness (32 so that a multiple of two still works, everything
else is confusing, see also next point)
- during runtime, handle gracefully if UID is 30 or 31 to not abort (we
can simply not allocate SRS resources)
Also, refactor config_srs() to always give back new memory; the code is
shorter and clearer. Rename to get_contig_srs() to make it explicit.
SCF 222.10.02, Table 3-35: the PDUSize parameter of a DL_TTI.request PDU
includes the 4 bytes required for PDUType and PDUSize parameters (each 2
bytes), not 2.
This is required for the Aerial interoperability.
- use new ssh class
- do proper logging from multiple services
- use yaml dir
- Use splitlines(), minor modifs of syntax
- check that services are running
- Check all copyin() operations, not only last
- switch to docker compose:
with the use of the new ssh class, we cannot (easily) use the old
docker-compose anymore (which was an alias to docker compose). Switch to
"docker compose".
Co-authored-by: Robert Schmidt <robert.schmidt@openairinterface.org>
Set ulimit of core to -1 enable core dumps for file-based core dumping.
Note that it is not necessary for e.g., systemd-coredumpd, but put it
for completeness.
- Return code says if copying succeeded
- Improve handling of the recursive code paths, which copy files or
directories recursively into another directory
- both for RemoteCmd/LocalCmd
Some CI users (OAI developers) expressed that the warning of the `docker
volume prune` before starting RFsim is confusing. To avoid the warning in
the HTML, force the removal; if the volume does not exist, no warning is
emitted.
To still catch possible problems, do not do this in the `terminate.xml`
so we might still see if there are other problems.
The original UL max MCS of 9 was chosen for performance reasons. In the
meantime, machines became faster, and the ULSCH decoding faster. Improve
UL max MCS to be 28 by default to use the complete range of possible
MCS.
Doing only 5 LDPC iterations by default leads to a poor reception
performance, and correspondingly many retransmissions. Increase to 8 as
the new default to improve user experience.
Ensure that the UE has an IP address before continuing the test.
Otherwise, it might happen we ping before the UE is connected.
Do not do this in do-ra and phytest, as they use --noS1, which implies
statically assigned IP address, so IP address check always succeeds(!).
Make RFsim start up faster by starting core and RAN start at the same
time. This speeds up because we do not wait for "healthy" containers
serially, but parallely.
For the RAN, it is safe to start gNB and UE at the same time in SA mode
(as the UE will only attach once the radio, and hence the full gNB, is
ready).
We do not start in parallel in do-ra and phytest, as in these modes the
gNB first has to write a UE-specific configuration file. If we start
both gNB and UE at the same time, the UE might try to read the file
before it is written, leading to a failure.
- PDCP reestablishment on CUCP (DRBs) triggering Bearer Context Modification procedures over E1
- Performing PDCP reestablishment for requested DRBs on CUUP
- Introduced function to notify re-establishment to CU-UP
- removed call to PDCP reestablishment for the DRBs on CUCP (it's done on CUUP)
cuup_notify_reestablishment:
- to fetch PDU session to setup in E1 following logic as in e1_send_bearer_updates
- fill bearer context modification request with stored F1-U tunnel information
during E1 reestablishment, otherwise it would initialized to 0 by default
- update GTP tunnel with the stored configuration after E1 reestablishment
The following ASSERT is tirggered while testing 16 UE setup.
Assertion (NPRB>0 && (NPRB + RBstart <= BWPsize)) failed!
In PRBalloc_to_locationandbandwidth0() /home/ran/common/utils/nr/nr_common.c:286
Illegal NPRB/RBstart Configuration (1,48) for BWPsize 48
Corrected the get_start_stop_allocation calculation to fix the issue.
The implementation was wrong. The rbstop shall be the last rb that could
be allocated. Also the following two conditions are taken care
rbStart needs to be strictly smaller than rbStop
rbStart+rbSize needs to be smaller or equal to rbStop
The problem was that we confounded both conditions, so rbStart could
be equal to rbStop, which then leads to one PRB being allocated at the end.
Setting rbStop to BWPsize - 1 and making sure the above two conditions are
true does not make it lead to wrong allocations anymore.
The previous code removed the "secondary UE" data (about DU association,
DU UE ID, associated CU UP), and only filled the F1-related data. In
other words, the associated CU-UP was lost. This commit reworks this
piece of code to retrieve old data, and update the relevant data without
destroying other.
- many structs and definitions are overlapping between the two different E1 procedures
- introduced naming specific to E1 Bearer Context Modification to improve readability
- grouped SDAP and PDCP configuration IEs for better reusability and readability
- introduced functions to set and get default PDCP config (DRBs and Bearer Contexts)
find_or_next_pdu_session:
- find_or_next_pdu_session refers to the E1 Bearer Context Modification Request
-- therefore it has to return pdu_session_to_mod_t* and takes in input e1ap_bearer_mod_req_t *
see !MR2545 for more context
- according to specs, in NR there are 4 SRBs
- in OAI the relevant define is used in the RRC procedures
- this commit harmonizes code by removing redundancy
- triggered by RRC reconfiguration events according to clause 5.3.5.6.5 of TS 38.331
- This procedure does QoS flow to DRB mapping according to clause 5.3.1 of TS 37.324
- QoS flows to DRBs to be released are also handled
CI: Integration Branch 2024 week 08
See merge request oai/openairinterface5g!2585
* !2540 NR UE max MIMO layers from UEcap
* !2554 NR UE RRC minor improvements
* !2563 Add missing backtick for proper code style in README.md
* !2567 Avoid integer overflows, buffer overflows, in channel levelling
* !2574 NR UE fix for configuring coreset0 and SS0
* !2575 NR UE fix init PDCP phy-test
* !2551 NR UE improvements for contention resolution timer
* !2570 CI: Replace nrmodule2 by up2
UBsan fails with errors without this commit. The commit changes:
- avoid two possible integer overflows in nr_dlsch_channel_level() and nr_dlsch_channel_level_median()
- nr_dlsch_channel_level_median() is simplified
- use MAX_ANT in avg to handle more than 4x4 config
- deleted: ../../../executables/rfsim.c
- removed unused command line options to set SINR and SNR
- removed unused functions to get SINR and SNR
- removed unused RFSIM struct members
for FR2 offsetToPointA is expressed in terms of 60 kHz SCS and k_SSB expressed in terms of the subcarrier spacing provided by the higher-layer parameter subCarrierSpacingCommon
Fix for CI runtime error: left shift of 1 by 31 places cannot be represented in type 'int'
Fix for CI runtime error: left shift of 128 by 24 places cannot be represented in type 'int'
Integration: 2024.w06
Closes#733 and #732
See merge request oai/openairinterface5g!2564
* !2490 reworking configuration of LogicalChannelConfig at MAC UE
* !2548 NR UE MSG3 buffer
* !2556 Support RC SM aperiodic subscription for "UE RRC State Change"
* !2557 fix duplicate call of RCconfig_NR_L1
* !2550 use pointer to structure instead of module_id inside MAC
* address #733
* !2560 chore(ci): force rebuild of ran-base for integration branches
* !2495 Sidelink configuration passed from RRC->MAC
* !2220 L1 tx thread
* !2559 FH 7.2 improvements
* !2566 Provide right E1 instance when creating E2 agent in CU-UP
* !2513 Update NR SA tutorials
CUuniqInstance is also used for GTP module identification, which
uses the UDP server socket as its ID. This leads to arbitrary IDs (>0),
which is wrong.
- Phy configuration will be prepared by MAC
- Sidelink preconfiguration parameters passed from RRC->MAC
- Only 1 SSB TA allocation used
- psbch payload prepared by MAC after receiving the tx slss req
PUSCH decoding is started multiple times in nr_ulsch_procedures() for
each ULSCH segment, and stopped once inside
phy_procedures_gNB_uespec_RX(). Not only is it not clear if that works
with the API, it also might measure more than just PUSCH decoding.
Remove the PUSCH decoding, it is misleading. Check L1 RX processing
instead.
The CI would only detect a deviation from the normalized value based on
a configurable threshold, typically 1.25 (so we would tolerate a 25%
increase). However, it often happens that a timing gets shorter (e.g.,
forgot to put the timing measurements). This commit introduces a
"Deviation Threshold" (smaller than 1) to not only check that we exceed
1.0 + "Deviation Threshold" (as before), but also that are not below 1.0
- "Deviation Threshold".
In other words, assuming a maximum divergence of 25%, instead of just
checking if val > 1.25, we now also check if val < 0.75 (and flag
error).
This commit introduces a separate thread (l1_tx_thread) that processes
gNB DL/TX slots. It receives a message from the ru_thread when a new
slot started, and starts processing.
The DL part of the scheduler is run in the l1_tx_thread. Therefore,
only call UL indication (for scheduler) in UL slots. The UL indication
previously triggered the DL scheduler -- hence, it had to be called in
every slot. Now, since the DL scheduler is moved into the DL thread, we
don't have to call the scheduler in every slot anymore.
The reorder thread is removed, as reordering with this scheme is not
necessary anymore.
The main advantage of this version is that the TX can be scheduled
earlier (sl_ahead). Further, there can no longer be race conditions in
the TX path, since the scheduler/L2, TX L1 processing and the RU tx
function are all execute in the same thread (L1_tX_thread).
Also, the scheduler prepares PRACH/PUSCH/PUCCH FAPI PDUs. As of this
commit, the scheduler runs in the TX processing chain, whereas these
PDUs are destined for RX processing. Thus, too avoid data races,
instead of triggering the RX processing at the same time as TX
processing in the RU thread, this commit changes the code to trigger RX
processing after the scheduler has been run in TX processing to avoid
data races.
Finally, we synchronize RU and L1 threads. This is important for
rfsimulator, since unlike most radios, rfsimulator can run slower or
faster, depending on I/O and CPU (it does not have a stable "tick").
Co-authored-by: kiran <saikiran@iitj.ac.in>
The UL_INFO_mutex is used in only one place, and therefore useless.
Also, the scheduler uses a lock internally to prevent concurrent access.
Hence, the UL_INFO_mutex is not needed.
Remove kill_gNB_proc(), as it is now dead code.
The next commit introduces a separate gNB TX processing thread. To
properly separate DL scheduler (for scheduling decisions) from UL
indication handling (for UL packets), this commit introduces a slot
indication "tick" to run the DL scheduler which will be put into the gNB
TX processing chain instead of the UL indication.
For different reasons, the frame/slot numbers might jump (e.g., gNB and
RU machines not properly synced). The scheduler would assert for PUCCH
allocations. Handle this gracefully by printing an error message, and
resetting the PUCCH structure.
Disable DPDK telemetry to skip creation of one/two threads (depending on
DPDK version). To force users to do this, bump the version of the patch
checked by cmake.
More information on DPDK telemetry:
https://doc.dpdk.org/guides/howto/telemetry.htmlCloses: #732 (suggestion 2)
Suggested-by: Andrew Sergeev <andrew.sergeev@adtran.com>
Force the CPU affinity for ru_thread explicitly to ru_thread_core. It is
already done via threadCreate(), but force again as per change request.
Delete commented 7.2 code that did the equivalent. We cannot set it in
the driver, as we don't have access to variable ru->ru_thread_core.
Closes: #732 (suggestion 2)
Suggested-by: Andrew Sergeev <andrew.sergeev@adtran.com>
system_core is the core reserved for DPDK control threads. It was
hardcoded to 0, but it seems that one some systems, CPU 0 cannot be
reasonably isolated from kernel threads. Make it therefore configurable
(default: 0)
Closes: #732 (suggestion 1)
Suggested-by: Andrew Sergeev <andrew.sergeev@adtran.com>
The main goal of this tutorial is to get 5G end-to-end connectivity as simple and fast as possible. For advanced topic the user should search for specific documentation.
No need to pull every single image independently, we can pull all the images with the corresponding tag that is in the 'docker-compose.yaml' with the 'docker compose pull' command
This tutorial aims to achieve end-to-end connectivity as simple as possible. Let's remove this complex topic from here.
Also, if a user copy-paste the section as it was, nrUE gets a segfault because there is no imeisv here, tutorial is out-dated.
- create hash table to save ric_req_id (key) and array of ran_param_id(s) (values), per each subscription
- create RB tree to store list of ric_req_id(s) for each ran_param_id
=> when the async event occurs, it is easier and faster to search per ran_param_id and send the indication message to all xApps (ric_req_id(s)) subscribed to the same ran_param_id
- it is important to mention that both data structures need to be maintained, especially when unsubscription occurs (free_aperiodic_subscription)
Integration: 2024.w05
See merge request oai/openairinterface5g!2549
* !2537 Fix for UE PUSCH scheduler DURATION_RX_TO_TX assertion
* !2538 Fix for UE MAC PDU LCID handling
* !2542 Fix L1_SSB_CSI_RSRP table mapping and check validity
* !2544 NR UE RRC timers improvement
* !2546 Correcting the location of packages
* !2493 Fix for computing SSB subcarrier offset
* !2496 MCS for retransmission fix in case of dci failed detection at NR UE
* !2503 Handle PDU Sessions in NG Initial UE context Setup and forward 5G-S-TMSI to core
* enable caching in the docker build of the CI build steps
* !2541 Fix RA bug, improve overall logging
We used to trigger the UECapabilityEnquiry right after
SecurityModeComplete, and before the RRCReconfiguration (which we call
"default" reconfiguration). However, 38.401 tells us that we should send
the RRCReconfiguration right after the SecurityModeComplete. In fact,
even though we get the UE capabilities, we cannot use them during this
reconfiguration, as we would first need to update the DU with a UE
Context Modification Request, which we cannot, as we just sent the UE
Context Setup Request, and the DU relies on first getting the
RRCReconfiguration.
Since we rely on a subsequent reconfiguration anyway, we can safely
trigger the UECapabilityEnquiry after RRCReconfigurationComplete.
(38.331 also says we can send UECapabilityEnquiry at any point.)
To cater for the possibility that there might not be any reconfiguration
coming afterwards, we check if a DRB has been set up. If not, we assume
a reconfiguration will come, and do not trigger one only for the UE
capabilities (this is what we do before this commit). If we already have
DRBs set up, they might have been set up during the "default" RRC
Reconfiguration, and another reconfiguration might not follow soon; in
this case, we trigger the reconfiguration by sending the UE capabilities
to the DU right away.
The next commit moves the UE Capability Enquiry after the first
reconfiguration. This has the effect that for some UEs (e.g., iPhone),
the Setup Requests come too close to each other, triggering RRC
Reconfigurations while previous transactions are ongoing.
I think the "true" solution would be to implement some tracking of
transactions across RRC, F1AP, E1AP, but this might require many
changes. For the moment, limit to delaying PDU session resource setups
to prevent above problem. Delaying is done using ITTI timers (to be able
to serve other UEs), waiting 10ms each time, up to 20 times (to not
deadlock the transaction -- after all, if the UE is unhappy, it will
drop the connection).
This commit allows the gNB to handle PDU sessions that the core requests
to setup during the NGAP Initial UE Context Setup. Previously, we only
managed them as part of PDU Session Resource Setup Request.
The RRC will, depending on whether a PDU session is in the NGAP Initial
UE Context Setup, either directly trigger the Security Command, or first
do a bearer setup at the CU-UP. Some asserts have been lifted, as now
the PDU sessions might be present before the RRC Connection is fully
established.
Implement the correct forwarding of the bearers in an F1 UE Context
Setup Request message.
This solves bug #672.
Prior to this commit, the handling of DRBs is complex: first the RRC
"guessed" a DRB ID when setting up DRBs via E1AP (in
rrc_gNB_process_NGAP_PDUSESSION_SETUP_REQ()), and later chose one for
real in fill_DRB_Configlist() (called in
rrc_gNB_generate_dedicatedRRCReconfiguration()).
To simplify, remove fill_DRB_Configlist(), and instead allocate the DRB
using generateDRB() before sending the message via E1AP, in
rrc_gNB_generate_dedicatedRRCReconfiguration(). The rest of the logic is
the same.
For PDU sessions, always mark PDU sessions as "done" to match pdu
session state logic.
It might happen that a UE has no CU-UP (e.g., never requested a PDU
session). When triggering a release, we previously and implicitly
associated a CU-UP in that case. That is not good, and confusing.
This commit adds a function to look up if the UE has an associated
CU-UP. We only send a release if it is the case.
The function get_existing_cuup_for_ue() now instead verifies that a
CU-UP exist, and does not implicitly create an association (which might
be unwanted, see above).
The function rrc_gNB_process_RRCReconfigurationComplete() does almost
nothing, we can delete it.
The variable name ue_reconfiguration_after_reestablishment_counter is
misleading, as it counts all reconfigurations; rename to make it clear.
The drb_active array keeps track of active DRBs. However, it only
replicates some of the information in established_drbs, and could lead
to a reuse of DRB IDs when two bearers are to be set up. Consider the
following:
1. trigger first DRB creation at RRC
2. DRB ID chosen from free drb_active entry
3. trigger second DRB creation at RRC
-> The first reconfiguration has not been acknowledged
-> drb_active is not marked as DRB_ACTIVE
4. The second DRB ID is chosen from a free drb_active entry, which is
the same as in 2.
By reusing established_drbs everywhere, this cannot happen, as we
1. select the DRB to be used using next_available_drb() and then
2. use generateDRB(), which marks the DRB used
all from within fill_DRB_configList, which gives a new DRB.
The logic is still overly complex, though.
If there are no active bearers, the function previously returned an
empty list. Return NULL if there are no bearers, so it is safe to call
this function even when no bearers are present. We can also pass it
every time to do_RRCReconfiguration(), which will not add an empty list
for DRBs.
Create a single function to activate an SRB. This function derives the
corresponding keys, triggers the setup in PDCP, and marks the SRB as set
up, instead of doing all of this common functionality in individual
steps. Use it at the corresponding places.
- show MAC bytes in one line
- add MCS table idx
- show ulsch errors before ulsch DTX (as in DL)
- show if UE in-sync/out-of-sync
- show CU UE ID next to RNTI to stats
Since we modify the format of the statistics, fix the format in the retx
checkers of the CI.
This commit implements retrieval of 5G-S-TMSI from Parts 1&2, or the
entire 5G-S-TMSI, depending on what the UE sends. Further, if it has
been retrieved, print a log of all constituent parts of it.
The UE-specific variable ng_5G_S_TMSI_Part2 is not needed, as we compute
the full 5G-S-TMSI as soon as we have Part2 (and 1).
The correct forwarding to the core is implemented in the following
commit.
We did not give a cause value, and therefore NGAP's "PDU sessions
failed" encoding failed. Give a cause value in all cases, so that this
cannot happen anymore.
To prevent against this, put an AssertFatal() into the encoder that
stops if an internally sent message does not have a correct cause value
(we can at least select a default).
Print an info message for all failed PDU sessions.
In rrc_gNB_send_NGAP_INITIAL_CONTEXT_SETUP_RESP(): if a PDU session is
already established, don't count it as failed.
If we are in one of these procedures, we basically must be in SA mode,
everything else is illogical. It also does not seem plausible to check
for this, as it is not apparent why only these couple of calls should
not be executed if not in SA. Hence, remove.
This implementation might work, but it simply triggers RRC setup on the new
DU when we might send a RRC reestablishment speeding up the whole
connection setup. For this:
- we should send the F1 UE context setup to the new DU
- we should still do the reestablishment, containing the reestablishment message
The latter should be sent once we received the F1 UE context setup
response. We don't need a new context. send a release to the old DU (and
don't release our own context!!).
We might also send the reestablishment to the UE encapsulated with the
F1 UE context setup response. Need to investigate if this is feasible
(because if the DU rejects [can that happen here?], the UE will still
send reestablishment complete -- what then?).
Also, we would need to send the E1 bearer context modification to the
CU-UP.
Keeping track if AS security is activated. This is important for a later
commit, where we use this to decide if we send a F1 UE Context Setup Req
with security command after receiving the E1 bearer context setup
response. The corresponding AssertFatal() removed in the same commit
implementing PDU session handling in NGAP Initial UE Context Setup.
38.401 Sec 8.1 asks for sending the Initial Context Setup Response after
a reconfiguration happened. Therefore, we can never send this message
directly from the handler in the Initial Context Setup Request
38.133 table 10.1.6.1.1 specifies dB for RSRP measurements. On the
"edges", the choice of INT_MIN/INT_MAX is poor, as it can quickly lead
to integer overflow/underflow for the wrong values. Instead, set proper
min/max RSRP values.
If we cannot send Msg.2, then basically the RA process is dead (because
the UE expects it at a fixed time). Hence, cancel the RA process, so
that we do not pick it up again after 1000 frames.
Integration: 2024.w04
See merge request oai/openairinterface5g!2535
* !2522 Integration of O-RAN 7.2 FH using OSC xRAN library
* !2416 fix-ue-scheduler-multi-thread-storage-list-mutex-protection
* !2488 LDPC T2 card encoding: rate matching and interleaving on the T2 card
* !2497 Read NoS1 bearer configuration from gNB-generated file
* !2518 Extension of E1AP procedures
* !2533 This limits PUCCH payload to 11 bits, 7 CSI + 3 ACKNACK + 1 SR maximum.
* !2526 prepare MSG3 in advance at RRC UE
* !2451 MAC procedures to handle DLSCH allocation type 0
* !2525 NR UE fix for function to find PUCCH resource set
* !2536 Improve Precoding Matrix computation taking in consideration the DCI format
* some documentation+config file updates for 7.2
* !2501 Improvements for NR implementation of timers
* !2531 Fix unit test build and build unit tests in CI
- Rate matching performed by card
- Due to rate matching on card, small TBs are encoded correctly, so the
limitation of encoding small TBs in CPU can be lifted (all TBs encoded
on T2), see a later commits
- CU-CP replying to the CU-UP with a Setup Failure message in case of unsuccessfull E1 Setup Request
- message is decoded by the CU-UP, and stops the process
Subsequent pipelines can/might use the ran-base image. Further, before
building a new ran-base image, we delete it. Hence, it suffices to only
delete the image when necessary (before building a new one), and
otherwise leave the current (develop, ci-temp) ran-base image.
Using AssertFatal() has the drawback that it requires the definition of
an exit_function() (which is typically declared in OAI main
executables). We don't want to pull in such dependency for every
consumer. Hence, use c stdlib's assert() instead.
This commit removes nr_rrc_mac_config_req_ue_logicalChannelBearer and handles configuration of logical channel bearers at MAC in nr_rrc_mac_config_req_cg.
This also avoids to configure LCs before resetting them again nr_rrc_mac_config_req_cg().
- update library location
- correct information for ioWorker
- disable HT on servers
- add local/lib trick for Fedora-based OS. Also clean-up on line terminations
- add another pkg-config trick.
- Proper version(s) of DPDK
- Support for both Bronze and E releases
- Minor fixes on E-patches and build system
- fixed a few typos and better explanation of the CPU allocation scheme
Co-authored-by: Raphael Defosseux <raphael.defosseux@eurecom.fr>
Co-authored-by: Jaroslava Fiedlerova <jaroslava.fiedlerova@openairinterface.org>
Co-authored-by: Robert Schmidt <jaroslava.fiedlerova@openairinterface.org>
- Add the O-RAN 7.2 FHI library, interfacing with OSC xRAN library (E
release)
- Add Findxran.cmake to detect the library and headers, including
version information
- Test for numa and DPDK presence
- Add patch for OSC xRAN to make interworking possible
- Add sample config files for LiteOn, Benetel, and VVDN units.
Documentation is provided in the next commit.
Co-authored-by: Raymond Knopp <raymond.knopp@eurecom.fr>
Co-authored-by: Cedric Roux <cedric.roux@eurecom.fr>
Co-authored-by: Manish Kumar Singh <manish1.kumar@amd.com>
Co-authored-by: thamizhselvan.k <thamizhselvan.k@vvdntech.in>
Co-authored-by: rajeshwari.p <rajeshwari.p@vvdntech.in>
Co-authored-by: Hongzhi Wang <hongzhi.wang@openairinterface.org>
Co-authored-by: Sofia Pison <Sofia.Pison@eurecom.fr>
For the initialization of the xran library (split 7.2), we need some
parameters such as PRACH or frame config. However, to date,
initialization of split at level 7 use the split 8 openair0_config
structure, which does not have the necessary information.
Hence, add a single structure that contains that information.
* make a single call to various function
* make fill_rf_config() static, it is only used in this file
* do not initialize the number of antennas in openair0_cfg in
init_NR_RU(), as it is already done in fill_rf_config()
Harmonize the call to openair0_transport_load(), by putting it in a
single place instead of multiple calls
Further, delay the call to ru_thread, for two reasons:
1) it is somewhat "symmetric" with local RF, which is done there as well
2) for the xran library, we need to read some parameters such as PRACH,
frame format, etc. At the RU, the corresponding FAPI config request
parameter structure is copied from the gNB to RU at the beginning of
the ru_thread, so we have to wait for that to happen.
Reformat code to properly copy memory if we skip precoding.
precoding_stats timing is not part of ru->feptx_ofdm, but
ru->feptx_prec, so always print it.
Previously, in order to speed up processing if precoding is disabled, we
just copy the pointer from txdataF to txdataF_BF, as opposed to do a
full memcpy() to duplicate the data. The same happens if we cannot have
precoding, i.e., in SISO. This has implications such as memory leaks,
which is why this is not being done. However, the code comment is
intentionally placed there.
For 7.2 FH, some RUs do phase compensation, whereas for others, the DU
does it. In fact, this can be "negotiated" using the M plane. Since we
don't have the M plane yet, add a configuration option to set where
phase compensation is performed.
For certain processing, e.g., in the case of FH 7.2, we need to ensure
that critical L1 processing happens fast. Try to force speedy processing
with this high priority.
Integration: 2024.w03
See merge request oai/openairinterface5g!2524
* !2519 fix(build): migrating rocky dockerfiles to Rocky-linux-9 to ease maintenance
* !2508 NR UE PDCP RLC indexing
* !2511 chore(ci): updating 5G RF / L2 simulator scenarios to release v2.0
* !2486 E2 agent: Use FlexRIC release v2 and add to CU-UP
* !2453 Liteon DU integration
* !2516 remove ASAN_OPTIONS: detect_leaks=0 from SA rfsim CI tests
* !2199 NR precoding matrix generation at MAC
RRC Version is a required IE. Hence, implement the encoder and decoder
for this message. At the DU and CU, read the RRC Version we use and fill
the corresponding field.
Co-authored-by: Robert Schmidt <robert.schmidt@openairinterface.org>
In NSA, we don't have SIB1. However, the DU system information requires
both MIB and SIB1, but is itself optional. So, in NSA, don't fill the DU
system information.
For LiteOn DU, we do not receive the MIB/SIB1 in the F1 Setup Request.
This is not a problem in itself, but it might happen that certain
procedures are executed without MIB/SIB1. Check for it and handle
appropriately:
- During normal connection establishment, we might not be able to
calculate measurement config
- During reestablishment, we might not be able to calculate ARFCN SSB
and cannot do reestablishment; do RRC setup instead.
read_version() takes a version string in format "x.y.z" with x being
major, y minor, and z patch number. It transforms reads them into
separate numbers, and outputs in the correspondingly named variables.
- e2_ran_func library divided into e2_ran_func_cuup (CU-UP only) and e2_ran_func_du_cucp_cuup (DU, CU, CU-CP and gNB-mono)
- GTP SM is not yet implemented in CU-UP -> CU-UP doesn't store PDU session information
- PDCP SM support for gNB-mono, CU and CU-UP
- KPM SM support for all node types (DU, CU, CU-UP, CU-CP and gNB-mono)
Integration: 2024.w02
See merge request oai/openairinterface5g!2514
* !2506 NR UE MAC config fixes
* !2478 Moving computation of base graph for LDPC at MAC in the UE
* !2481 fix UE RX/TX thread several bugs in synchronisation of events, add LOG_E and...
* !2494 Fix for building rach occasions in case of multiple BWPs
* !2498 Fix security header type for PDU session establishment, and update sqn, counters
* !2509 Fixed typos in init_ru
* !2512 Null pointer in nr generate msg3 retransmission
* !2505 USRP doc: Include b205-mini, improve markup
While testing the OAI 5G RAN in SA mode, it appears that the PDU session
establishment request is using the wrong security header type.
The current used header is: "Integrity protected and ciphered with new
5G NAS security context".
However, according to TS 24.501, Table 9.3.1, this header can only be used
by the "Security mode complete" procedure.
Changes:
- Hence, the new security header has been changed to "Integrity protected
and ciphered", which will use the security context generated during the
"Security mode complete".
- Sequences numbers and NAS counters have been updated to reflect that
change.
Integration: 2024.w01
See merge request oai/openairinterface5g!2502
* !2482 Trigger UL Failure if Ack of Msg4 not received and implement RA Contention Resolution timer in gNB
* !2484 function to harmonize handling of NAS registraction accept
* !2487 Rework PUCCH codec with c16_t type and functions to make readable code
* !2491 improvements for NR UE detaching
* !2499 Update L1sim and RFsim documentation
* !2500 Fix FAPI CoreSetType enumeration options after the fix on the description made by SCF
* !2489 NR UE improvements RRCSetup
- update explanation for noS1
- better help for nr-uesoftmodem options
- consistent writing of name "RFsimulator"
- remove LDPC section at end, is already referenced in doc/README.md
- Add TOC
- Better describe what RFsim does
- Remove misleading, wrong, unnecessary information
* It is futile to mention --phy-test or other modes, RFsim works with
all
* Explaining how to ping is not the task of the RFsim documentation
(that is what tutorials are for)
* Remove useless/outdated build log information
- Add some more information on which flags to run
- Improve formatting, and minor clarifications
SCF FAPI 222.10.02 had a bug on the description of CoreSetType:
0: CORESET is configured by the PBCH or SIB1
(subcarrier 0 of CRB0 for DMRS mapping)
1: otherwise (subcarrier 0 of CORESET)
That was fixed in SCF 222.10.04:
0: CORESET is configured by the PBCH or SIB1
(subcarrier 0 of the CORESET)
1: otherwise (subcarrier 0 of CRB0 for DMRS mapping)
That wrong description created a misunderstanding, to overcome it we "wrongly" added a third value. This commit fixes that and aligns the CoreSetType with FAPI 222.10.04
- Do not list L1sim as a way of running OAI: it is untested, and might
not work
- Reword the tutorial where necessary, remove outdated/wrong information
(e.g., we don't test this mode AFAIK)
The UE will trigger new RA either with MAC CE C-RNTI or with new RRCSetupRequest:
- MAC CE C-RNTI: The ul_failure_timer will be reset
- RRCSetupRequest: New UE will be created and the previous one will be deleted
Integration: 2023.w51
See merge request oai/openairinterface5g!2483
* !2221 NR UE going to IDLE state
* !2464 move common .h file in common directory, remove mem_block_t
* !2479 CI: test with 2 UL layers in AW2S pipeline
* !2468 LDPC T2 card: fix and doc update
* !2480 Enabling multi-ue support in the open-source L2 proxy
* !2476 Fixes for multi-DU/CU-UP and CI test
* !2482 Trigger UL Failure if Ack of Msg4 not received
* !2484 function to harmonize handling of NAS registraction accept
The 5G-RFsim F1 test tested multiple DUs at one CU. This
functionality is retained in the new F1+E1 test.
The 5G-RFsim E1 test tested a single CU-UP at a single CU-CP. Here, we
extend to use multiple CU-UPs, one for each DU. The associated is
ensured by using three different F1-U transport networks, one for each
DU/CU-UP pair.
- Test only one DU at the CU (3 DUs is tested in F1+E1)
- Harmonize IP configuration to make it simpler to reuse in the F1+E1
test
- Speed up the test by reducing the number of steps to deploy core/RAN
- Add traffic test
binding the local address for SCTP is pointless, because we don't really
care through which interface SCTP traffic goes. Actually, we even might
want to set the DU remote_n_addr and the the local_n_addr to addresses
in different networks, because we might want to have GTP traffic go
through a different interface than SCTP, as done in the CI test for E1.
MR-2401 changed the argument to the init_nrUE_standalone_thread
to be zero instead of the ue_id_g. This causes every port to be
opened for multiple UEs to have the same value, therefore only
allowing one UE to be connected to the proxy. If the ue_id_g
global is an issue, we should open a new ticket for a proper fix.
We set the RFsim TCP socket to non-blocking. Thus, it may happen that we
cannot write() to the socket when the kernel cannot handle the request,
in which case we wait some time before trying again. This is the
correct/expected behavior.
Since this is *normal*, an error message is not justified; we did not
even print a message in this case before commit 21790f9c4f introduced it.
When setting up bearer(s) through F1, the corresponding F1 message has
QoS fields for the DRB(s) to be set up, as well as for all flows
associated to this DRB. In this commit, set the DRB QoS field to be
equal to the first flow (it is unclear/unspecified how it should be set
up, based on the flows), which was not set before and therefore
contained unitialized data.
The F1 module uses this data, and in split mode, F1 message generation
failed due to uninitialized data.
We reuse RRC UE IDs at the RRC. Previous to this commit, we did not
release a UE at the CU-CP's PDCP (for SRBs). Thus, when reconnecting a
UE, the PDCP might not be in a good state, since we did not release the
previous PDCP UE context.
Integration: 2023.w50
See merge request oai/openairinterface5g!2474
* !2460 bugfix: initialize TX and RX antennas separately
* !2471 fix a useless computation done in case we activate LOG_D
* !2417 PDSCH precoding optimization
* !2436 Fix the DoS vulnerability in RA procedure
* !2472 (ci): adding new packages in cuup dockerfiles
* !2463 NR UE RRC MAC configuration improvements
* !2475 Fix gNB RA procedures when Msg4 Ack not received
remove several unused files
remove mem_block_t that is never used in a different way than a simple heap buffer
move up IP address type and remove duplications of the same struct declaration
When the gNB does not receive PUCCH with Ack for Msg4 we have to consider Msg4 received by the UE, so we need to apply the CellGroupConfig sent to the UE.
There are two scenarios:
1 - The UE never decoded PDSCH for Msg4, the RA will fail because RA Contention Resolution timer, the UE will come with a new RA with new RRCSetupRequest
2 - The UE decoded PDSCH for Msg4, and if SR does not work and no resources for PUSCH, it will trigger RA with MAC CE C-RNTI in Msg3 (so, gNB must be ready for it)
Integration: 2023.w49
See merge request oai/openairinterface5g!2470
* !2264 Improvements in f1ap for qos
* !2377 Store allowed NSSAIs from NAS message
* !2462 Hotfix: swap UIDs after RRCReestablishment
* !2449 Improve performance of polar initialization
* !2469 handling configuration of MAC-CellGroupConfig at UE
* !2395 `nr_generate_pdsch()` clean up and little performance improves
* !2461 CI: add FDD test to SA B200 pipeline
* !2424 Improvements in SDAP RRC management
to use faster code
- replaced hundreds mallocs by either local arrays or by the OAI multi-dimensionalarray (G_N_tab array)
- made in SIMD a one bit per byte packing function to packed (ordinary) bits in bytes
The CellGroup is created considering UID, for example:
'pucchres0->startingPRB = 8 + uid;' and since we keep the CellGroup from
the previous RNTI/UID we should keep also the previous UID.
Considering the scenario where a UE did successful RRCReestablishment,
it was getting uid = 1, and uid = 0 was removed.
When another UE connected, it was getting UID = 0 (because 0 was already
free), and so the PUCCH 0 resource was conflicting with the first UE.
On that occasion, also correct the comment that describes the code, as
it was wrong/misleading.
- new cu and du config files for testing in band 1
- modification of yml file for CU DU deployment - with new config files
- modification of xml file for SA F1 testcase
- Implement: Extend the F1 encoding and decoding with Qos
- Implement: E1 decoding for QoS
- Fix: Modify the UE_MODIFICATION_REQUEST_MSG towards DU to contain QoS info based on E1AP context response
- Fix: Modify the E1AP and F1AP message structures
- Fill the Qos configuration to send to MAC
├── radio : Drivers for various radios such as USRP, AW2S, RFsim, ...
└── targets : Some configuration files; only historical relevance, and might be deleted in the future
```
# How to get support from the OAI Community #
You can ask your question on the [mailing lists](https://gitlab.eurecom.fr/oai/openairinterface5g/-/wikis/MailingList).
Your email should contain below information:
- A clear subject in your email.
- For all the queries there should be [Query\] in the subject of the email and for problems there should be [Problem\].
- In case of a problem, add a small description.
- Do not share any photos unless you want to share a diagram.
- OAI gNB/DU/CU/CU-CP/CU-UP configuration file in `.conf` format only.
- Logs of OAI gNB/DU/CU/CU-CP/CU-UP in `.log` or `.txt` format only.
- In case your question is related to performance, include a small description of the machine (Operating System, Kernel version, CPU, RAM and networking card) and diagram of your testing environment.
- Known/open issues are present on [GitLab](https://gitlab.eurecom.fr/oai/openairinterface5g/-/issues), so keep checking.
Always remember a structured email will help us understand your issues quickly.
result=re.search('got sync|Starting E1AP at CU UP|Starting F1AP at CU|Got sync|Waiting for RUs to be configured|cuPHYController initialized|Received CONFIG.response, gNB is ready',mySSH.getBefore())
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
