Use flat buffer for txdataF in gNB L1
This MR changes txdataF buffer format to hold freq domain data starting from
PRB0 instead of circular buffer that starts from PRB N/2. The motive is to
simplify RE mapping function in L1 and copying data to xran lib.
- Add num_ues and ue_id parameters for server and client configuration
- Implement ul combining and dl distribution to all ues
- Each ue uses dedicated antenna offsets based on ue_id
- Works with no-chanmod mode only and with uniform antenna dimensions across all ues
The original error was simply "failed to write to RF", which is not very
precise. Update with sample numbers for context, and add it in another
similar case for consistency.
Co-authored-by: Sagar Arora <sagar.arora@openairinterface.org>
At least writerProcessWaitingQueue() checks for the exact number of
samples written. In many other places, we check that the return is
non-zero.
The usrp-tx-write-thread functionality returned 0, which triggers
various problems. Return the number of samples, as in the "direct write"
case (if-block before else-block).
Co-authored-by: Sagar Arora <sagar.arora@openairinterface.org>
So far txdataF buffer has DC RE as the first element. This was prefered
most likely to avoid memcpy before iFFT but comes with a small
complexity in RE mapping functions.
Instead, we use a flat buffer that holds freq domain data starting from
first negative SC. After resource mapping is done on all channels, each
OFDM symbol is shifted before iFFT so that the buffer start with DC RE.
This reduces complexity in mapping functions and later on when passing
beam IDs to 7.2 split RU especially when a beam is associated to
interleaved RB and/or REs.
There is already a memcpy in place to copy txdataF from gNB stuct to RU
struct. So this change does not introduce new memcpy.
Remove L1 dependence from NR band number
L1 band information relied on get_band function that couldn't provide an
unambiguous result. Also there is no SCF compliant parameter for L1 and
generally we don't really need that information in L1.
The MR also remove a RU section that used nr_band but was never really
accessed in NR.
Prior to this commit, constants NB_ANTENNAS_RX/TX are provided on the
command line to each compiler invocation. This has the drawback (1) that
also files that don't need that get the constants, and (2) it is not
easy to find them using ctags or similar, as the only definition is in
CMakeLists.txt (where it is not found).
In this commit, use cmake configure_file() to write a header file with
these constants. The file will be generated during cmake invocation, and
written to <build-dir>/common/cmake_defs.h. The relevant files that
need this definition will pick it from the header.
This has been used previously also in a commit to write git information.
See-also: 8bffd1666d ("Avoid complete rebuild on cmake run with git info change")
All the use cases we tested so far, are supported with xran F release.
Therefore, we can proceed with removing the support with E release.
In addition, our plan is to bring the support for K release which is ongoing activity in
the MR https://gitlab.eurecom.fr/oai/openairinterface5g/-/merge_requests/3444.
Cleanup Cmake: remove some double compilation
Harmonize some cmake code to avoid double compilation, notably around
polar and smallblock. In develop, we just add their sources to other
targets, whereas now we define STATIC libraries (further changed to
shared library in !3856) to avoid double compilation.
With this command line:
cmake .. -GNinja -DENABLE_TESTS=ON && ninja
shows that before we compile 12631 files, where after it is 12583
there is additional supporting cleanup (move code, remove headers,
types, ...).
there is more harmonization to do, I limited myself to some easy wins.
For instance, we do the same with PDCP/SDAP, and this should be
refactored into STATIC libraries as well.
Remove VCD signal dumper from NR code
We don't test VCD in 5G NR code (and haven't tested it in 4G for years).
Remove the VCD traces in 5G code to avoid any ambiguity. 4G code remains
in place, as well as the tools to analyze existing traces.
- Add more sampling rates to support, e.g., 40MHz in 5G
- Allocate memory only at end to avoid memory leaks in case of failure
- Group recurring configuration functionality in configure_channel() to
reduced code duplication and uniform output
- Read maximum TX gain to set maximum automatically, and analogous to
USRP radio lib
- Check if BladeRF config (buffer size, num transfers, ...) fulfills
requirements as specified in documentation [1]
- Set BladeRF logging early to see BladeRF output
[1] https://www.nuand.com/bladeRF-doc/libbladeRF/v2.6.0/group___f_n___s_t_r_e_a_m_i_n_g___a_s_y_n_c.html#ga72752f2a047b95544e7686596a409abd
Co-authored-by: Brice Robert <robert@eurecom.fr>
Closes: #845
Don't enable radios twice, and only do it on device startup. Simplify
that function by using a new BLADERF_CHECK macro that checks return
codes and that will be reused in the following commits.
Fix for split 7.2 DU configurations including DU & RU ethernet addresses
Fix parsing of DU configurations that include both RU and DU ethernet
addresses in the fhi_72 section. The extraneous DU addresses are used in
MPlane to configure the RU.
Fix parsing of DU configurations that include both RU and DU ethernet
addresses in the fhi_72 section. The extraneous DU addresses are used
in MPlane to configure the RU.
Add xran_queue_length atomic variable to count the number of messages in
oran_sync_fifo (as notifiedFIFO_t does not do this). Use this to
tolerate some double messages (e.g. excessive processing delay), but
start to skip if messages accumulate.
Prior to this change, "Detected double sync message" was printed each
time there was more than one message, which reflected the previous
behavior when using polling, but the gNB might catch up if it is only
once. The new logic ensure that we don't skip individual slots but do so
once messages start to accumulate.
In xRAN, it appears that T1a/Ta4 values are assumed to be within a TTI.
When T1a/Ta4 values with larger than a TTI are used, this causes two issues: (1) C-plane packets to be generated later than expected and (2) oai_xran_fh_rx_callback() receives the wrong timing information resulting in potentially empty/stale PUSCH/PRACH data to be read.
In this patch, we modify xRAN to make it handle large T1a/Ta4 values properly.
The changes include: (1) three additional variables in offset_num_slots_* (xran_dev.h), (2) modify "xran_timing_create_cbs" and "xran_timer_arm_for_deadline" (xran_cb_proc.c), and (3) revise "tx_cp_dl_cb" and "tx_cp_ul_cb" (xran_main.c).
Additionally, we add additional guard conditions to prevent xran_tx_proc.c from sending U-plane packets when it is not a downlink slot.
[FHI72] [xran F] Handle fragmented (un)compressed packets
In both E and F, the radio-transport fragmentation is not supported, and
all the "fragmented" (not seen as fragments in the xran, only as
sections) packets contain "E-bit" equal to 1. Per spec, the value 1 for
"E-bit" signifies the last fragment of one symbol. For more info, please
see "ecpriSeqid" section in the spec.
In E:
- nRBStart and nRBSize is updated according to the sent/received packet;
- max sections = 1;
- when a packet is fragmented, the xran uses memcpy(pos, iq_data_start,
size) function, and nothing is stored in start_prbu , num_prbu ,...
In F:
- nRBStart and nRBSize is only used for C-plane messages;
- nPrbElm represents the number of sections for DL only;
- based on the nPrbElm, each section contains UP_nRBStart and UP_nRBSize
parameters which represent the start and number of PRBs in one section
or one fragment;
- when a packet is fragmented, the xran uses nSecDesc and sec_desc to
extract the packet section content, and the data is stored in
start_prbu , num_prbu ,...
Previous work on the fragmentation in F:
- #884
- !3394 - only increases the XRAN_MAX_FRAGMENT from 4 to 6, since it was
only tested with BFP 8, but not end-to-end unfortunately. But this MR
fixes all the issues.
Tested with VVDN RU, 2x2, 16 bits, 100MHz, 30kHz, MTU <= 9216 B, and
Quectel UE RM520N-GL Revision: RM520NGLAAR03A03M4G.
The sufficient number of fragments needed for MTU 1500 and 8b compression is 6.
However, XRAN_MAX_FRAGMENT must be incremented by one (6+1 = 7).
In the MR https://gitlab.eurecom.fr/oai/openairinterface5g/-/merge_requests/3394, XRAN_MAX_FRAGMENT was increased to 6,
but the Tx data was actually copied 5 times, and unfortunately I haven't tested end-to-end back then.
This commit was successfully tested with Liteon FR1.
Even though the jumbo frame has 9600 bytes by the spec, the max MTU is limited to the MTU of a switch, if any used.
However, the value can be < max MTU of a switch.
This is very important for fragmentation purposes. If not properly set, the PDSCH packets will be missing.
semi-manual remove of not used #define commands
As i was upset to see all these #define with no code under, i made a
global pass to remove the obvious cases semi-manual remove of not used
bodies
manual script, then a lot of manual review
grep -IRs "#define" ../openair* ../common/* ../executables/ ../nfapi/ ../radio/ | \
fgrep -v '(' | fgrep -v '//' |awk '{print $2}' | egrep -v "#define *$n"| \
while read n ; do
grep -IRs $n ../openair* ../common/* ../executables/ ../nfapi/ ../radio/ | grep -v "#define *$n" | grep -q $n || echo $n;
done
will give a decent list of #define never used. Then, I think we should
keep lists of # define when it comes from standards, even if the code
doesn't use it right now
NR UE: add support for multiple RF frontends
This MR adds basic support for multiple RF-frontends at the UE.
There are multiple reasons why we might want this:
1. There might be different antennas needed for different bands (e.g. NTN or
TN antennas). These antennas might be connected to different RF-frontends.
In case of Handover, the UE needs to dynamically switch between these
RF-frontends.
2. Each rf-simulator connection carries baseband I/Q samples. To be able to
perform a handover between two cells at different frequencies, there must
be two rf-simulator connections, as the baseband channels should not
interfere and thus cannot simply be added.
3. When we want to support multiple UE simulation within one instance of
nr-uesoftmodem, and allow multiple cells and probably even handover between
these cells, we need either:
1. dedicated RF-frontends for each simulated UE, or
2. dedicated RF-frontends for each connected cell and dynamically assign
simulated UEs accordingly.
This MR implements what is needed for 1., 2. and 3.1.
Fix FFT size for xran_fh_tx_send_slot_BySymbol() and xran_fh_rx_read_slot_BySymbol()
This MR handles
- Fix FFT size for xran_fh_tx_send_slot_BySymbol() and
xran_fh_rx_read_slot_BySymbol() (!3809)
- Fix E release build
Even after this commit, we still have duplicated UL C-plane packets due to xran F release internal architecture.
Note: in E, all PRBs were in one section of one C-plane DL and UL. It is working even though the U-plane has "fragmented" packets.
Co-authored-by: Mario Joa-Ng <mario.joa-ng@openairinterface.com>
Previously, in F, only the following (subframe, slot, symbol) combinations were sent every frame:
(0, 0, 0)
(2, 1, 0)
(5, 0, 0)
(7, 1, 0)
Now, all the DL symbols are sent.
Notes:
In both E and F, the radio-transport fragmentation is not supported, and
all the "fragmented" (not seen as fragments in the xran, only as sections)
packets contain "E-bit" equal to 1.
Per spec, the value 1 for "E-bit" signifies the last fragment of one symbol.
For more info, please see "ecpriSeqid" section in the spec.
In E:
- nRBStart and nRBSize is updated according to the sent/received packet;
- max sections = 1.
In F:
- nRBStart and nRBSize is only used for C-plane messages;
- nPrbElm represents the number of sections for DL only;
- based on the nPrbElm, each section contains UP_nRBStart and UP_nRBSize parameters
which represent the start and number of PRBs in one section or one fragment.
Co-authored-by: Mario Joa-Ng <mario.joa-ng@openairinterface.com>
Robert Schmidt