Merge remote-tracking branch 'origin/aerial_25-3' into integration_2026_w05 (!3868)
CI: Upgrade Aerial pipeline to 25-3 Upgrade Aerial CI pipeline to 25-3 after aerial-cuda-accelerated-ran became open source.
This commit is contained in:
Jaroslava Fiedlerova
@@ -418,13 +418,13 @@ class Containerize():
|
|||||||
elif image != 'ran-build':
|
elif image != 'ran-build':
|
||||||
cmd.run(f'sed -i -e "s#ran-build:latest#ran-build:{imageTag}#" docker/Dockerfile.{pattern}{self.dockerfileprefix}')
|
cmd.run(f'sed -i -e "s#ran-build:latest#ran-build:{imageTag}#" docker/Dockerfile.{pattern}{self.dockerfileprefix}')
|
||||||
if image == 'oai-gnb-aerial':
|
if image == 'oai-gnb-aerial':
|
||||||
cmd.run('cp -f /opt/nvidia-ipc/nvipc_src.2025.10.09.tar.gz .')
|
cmd.run('cp -f /opt/nvidia-ipc/nvipc_src.2026.01.07.tar.gz .')
|
||||||
logfile = f'{lSourcePath}/cmake_targets/log/{name}.docker.log'
|
logfile = f'{lSourcePath}/cmake_targets/log/{name}.docker.log'
|
||||||
ret = cmd.run(f'{self.cli} build {self.cliBuildOptions} --target {image} --tag {name}:{imageTag} --file docker/Dockerfile.{pattern}{self.dockerfileprefix} {option} . > {logfile} 2>&1', timeout=1200)
|
ret = cmd.run(f'{self.cli} build {self.cliBuildOptions} --target {image} --tag {name}:{imageTag} --file docker/Dockerfile.{pattern}{self.dockerfileprefix} {option} . > {logfile} 2>&1', timeout=1200)
|
||||||
t = (name, archiveArtifact(cmd, ctx, logfile))
|
t = (name, archiveArtifact(cmd, ctx, logfile))
|
||||||
log_files.append(t)
|
log_files.append(t)
|
||||||
if image == 'oai-gnb-aerial':
|
if image == 'oai-gnb-aerial':
|
||||||
cmd.run('rm -f nvipc_src.2025.10.09.tar.gz')
|
cmd.run('rm -f nvipc_src.2026.01.07.tar.gz')
|
||||||
# check the status of the build
|
# check the status of the build
|
||||||
ret = cmd.run(f"{self.cli} image inspect --format=\'Size = {{{{.Size}}}} bytes\' {name}:{imageTag}")
|
ret = cmd.run(f"{self.cli} image inspect --format=\'Size = {{{{.Size}}}} bytes\' {name}:{imageTag}")
|
||||||
if ret.returncode != 0:
|
if ret.returncode != 0:
|
||||||
|
|||||||
@@ -13,16 +13,7 @@ gNBs =
|
|||||||
|
|
||||||
// Tracking area code, 0x0000 and 0xfffe are reserved values
|
// Tracking area code, 0x0000 and 0xfffe are reserved values
|
||||||
tracking_area_code = 1;
|
tracking_area_code = 1;
|
||||||
plmn_list = ({
|
plmn_list = ({ mcc = 208; mnc = 97; mnc_length = 2; snssaiList = ({ sst = 1; }) });
|
||||||
mcc = 208;
|
|
||||||
mnc = 97;
|
|
||||||
mnc_length = 2;
|
|
||||||
snssaiList = (
|
|
||||||
{
|
|
||||||
sst = 1;
|
|
||||||
}
|
|
||||||
);
|
|
||||||
});
|
|
||||||
|
|
||||||
nr_cellid = 12345678L;
|
nr_cellid = 12345678L;
|
||||||
|
|
||||||
@@ -186,19 +177,19 @@ gNBs =
|
|||||||
|
|
||||||
MACRLCs = (
|
MACRLCs = (
|
||||||
{
|
{
|
||||||
num_cc = 1;
|
num_cc = 1;
|
||||||
remote_s_address = "127.0.0.1"; // pnf addr [!]
|
remote_s_address = "127.0.0.1"; // pnf addr [!]
|
||||||
local_s_address = "127.0.0.2"; // vnf addr
|
local_s_address = "127.0.0.2"; // vnf addr
|
||||||
local_s_portc = 50001; // vnf p5 port
|
local_s_portc = 50001; // vnf p5 port
|
||||||
remote_s_portc = 50000; // pnf p5 port [!]
|
remote_s_portc = 50000; // pnf p5 port [!]
|
||||||
local_s_portd = 50011; // vnf p7 port [!]
|
local_s_portd = 50011; // vnf p7 port [!]
|
||||||
remote_s_portd = 50010; // pnf p7 port [!]
|
remote_s_portd = 50010; // pnf p7 port [!]
|
||||||
tr_s_preference = "aerial";
|
tr_s_preference = "aerial";
|
||||||
tr_n_preference = "local_RRC";
|
tr_n_preference = "local_RRC";
|
||||||
pusch_TargetSNRx10 = 280; # 150;
|
pusch_TargetSNRx10 = 280; # 150;
|
||||||
pucch_TargetSNRx10 = 200; #200;
|
pucch_TargetSNRx10 = 200; #200;
|
||||||
dl_max_mcs = 28;
|
dl_max_mcs = 28;
|
||||||
ul_max_mcs = 28;
|
ul_max_mcs = 28;
|
||||||
}
|
}
|
||||||
);
|
);
|
||||||
|
|
||||||
@@ -222,13 +213,13 @@ security = {
|
|||||||
|
|
||||||
log_config :
|
log_config :
|
||||||
{
|
{
|
||||||
global_log_level ="info";
|
global_log_level ="info";
|
||||||
hw_log_level ="info";
|
hw_log_level ="info";
|
||||||
phy_log_level ="info";
|
phy_log_level ="info";
|
||||||
mac_log_level ="info";
|
mac_log_level ="info";
|
||||||
rlc_log_level ="info";
|
rlc_log_level ="info";
|
||||||
pdcp_log_level ="info";
|
pdcp_log_level ="info";
|
||||||
rrc_log_level ="info";
|
rrc_log_level ="info";
|
||||||
ngap_log_level ="debug";
|
ngap_log_level ="debug";
|
||||||
f1ap_log_level ="debug";
|
f1ap_log_level ="debug";
|
||||||
};
|
};
|
||||||
|
|||||||
@@ -80,7 +80,7 @@
|
|||||||
<testCase>
|
<testCase>
|
||||||
<class>Iperf</class>
|
<class>Iperf</class>
|
||||||
<desc>iperf (DL/UDP/800M)(30 sec)(single-ue profile)</desc>
|
<desc>iperf (DL/UDP/800M)(30 sec)(single-ue profile)</desc>
|
||||||
<iperf_args>-u -b 800M -t 30 -R</iperf_args>
|
<iperf_args>-u -b 800M -t 30 -R -l 1428</iperf_args>
|
||||||
<svr_id>oc-cn5g-20897-aerial</svr_id>
|
<svr_id>oc-cn5g-20897-aerial</svr_id>
|
||||||
<id>up2-aerial</id>
|
<id>up2-aerial</id>
|
||||||
<iperf_packetloss_threshold>25</iperf_packetloss_threshold>
|
<iperf_packetloss_threshold>25</iperf_packetloss_threshold>
|
||||||
@@ -101,7 +101,7 @@
|
|||||||
<class>Iperf</class>
|
<class>Iperf</class>
|
||||||
<desc>Optional iperf (DL/UDP/1000M)(30 sec)(single-ue profile)</desc>
|
<desc>Optional iperf (DL/UDP/1000M)(30 sec)(single-ue profile)</desc>
|
||||||
<may_fail>true</may_fail>
|
<may_fail>true</may_fail>
|
||||||
<iperf_args>-u -b 1000M -t 30 -R</iperf_args>
|
<iperf_args>-u -b 1000M -t 30 -R -l 1428</iperf_args>
|
||||||
<svr_id>oc-cn5g-20897-aerial</svr_id>
|
<svr_id>oc-cn5g-20897-aerial</svr_id>
|
||||||
<id>up2-aerial</id>
|
<id>up2-aerial</id>
|
||||||
<iperf_packetloss_threshold>25</iperf_packetloss_threshold>
|
<iperf_packetloss_threshold>25</iperf_packetloss_threshold>
|
||||||
|
|||||||
@@ -101,7 +101,7 @@
|
|||||||
<testCase>
|
<testCase>
|
||||||
<class>Iperf</class>
|
<class>Iperf</class>
|
||||||
<desc>iperf (DL/UDP/200M)(30 sec)(single-ue profile)</desc>
|
<desc>iperf (DL/UDP/200M)(30 sec)(single-ue profile)</desc>
|
||||||
<iperf_args>-u -b 200M -t 30 -R</iperf_args>
|
<iperf_args>-u -b 200M -t 30 -R -l 1428</iperf_args>
|
||||||
<svr_id>oc-cn5g-20897-aerial</svr_id>
|
<svr_id>oc-cn5g-20897-aerial</svr_id>
|
||||||
<id>up2-aerial</id>
|
<id>up2-aerial</id>
|
||||||
<iperf_packetloss_threshold>25</iperf_packetloss_threshold>
|
<iperf_packetloss_threshold>25</iperf_packetloss_threshold>
|
||||||
|
|||||||
@@ -4,8 +4,6 @@
|
|||||||
cuBB_Path="${cuBB_SDK:-/opt/nvidia/cuBB}"
|
cuBB_Path="${cuBB_SDK:-/opt/nvidia/cuBB}"
|
||||||
|
|
||||||
cd "$cuBB_Path" || exit 1
|
cd "$cuBB_Path" || exit 1
|
||||||
# Add gdrcopy to LD_LIBRARY_PATH
|
|
||||||
export LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/opt/mellanox/dpdk/lib/x86_64-linux-gnu:/opt/mellanox/doca/lib/x86_64-linux-gnu:/opt/nvidia/cuBB/cuPHY-CP/external/gdrcopy/build/x86_64/
|
|
||||||
|
|
||||||
# Restart MPS
|
# Restart MPS
|
||||||
# Export variables
|
# Export variables
|
||||||
@@ -14,7 +12,7 @@ export CUDA_MPS_PIPE_DIRECTORY=/var
|
|||||||
export CUDA_MPS_LOG_DIRECTORY=/var
|
export CUDA_MPS_LOG_DIRECTORY=/var
|
||||||
|
|
||||||
# Stop existing MPS
|
# Stop existing MPS
|
||||||
sudo -E echo quit | sudo -E nvidia-cuda-mps-control
|
echo quit | sudo -E nvidia-cuda-mps-control
|
||||||
|
|
||||||
# Start MPS
|
# Start MPS
|
||||||
sudo -E nvidia-cuda-mps-control -d
|
sudo -E nvidia-cuda-mps-control -d
|
||||||
@@ -24,11 +22,11 @@ sudo -E echo start_server -uid 0 | sudo -E nvidia-cuda-mps-control
|
|||||||
# Check if an argument is provided
|
# Check if an argument is provided
|
||||||
if [ $# -eq 0 ]; then
|
if [ $# -eq 0 ]; then
|
||||||
# No argument provided, use default value
|
# No argument provided, use default value
|
||||||
argument="P5G_FXN_GH"
|
argument="P5G_WNC_GH"
|
||||||
else
|
else
|
||||||
# Argument provided, use it
|
# Argument provided, use it
|
||||||
argument="$1"
|
argument="$1"
|
||||||
fi
|
fi
|
||||||
|
|
||||||
sudo -E "$cuBB_Path"/build/cuPHY-CP/cuphycontroller/examples/cuphycontroller_scf "$argument"
|
sudo -E "$cuBB_Path"/build.$(uname -m)/cuPHY-CP/cuphycontroller/examples/cuphycontroller_scf "$argument"
|
||||||
sudo -E ./build/cuPHY-CP/gt_common_libs/nvIPC/tests/pcap/pcap_collect
|
sudo -E ./build.$(uname -m)/cuPHY-CP/gt_common_libs/nvIPC/tests/pcap/pcap_collect
|
||||||
|
|||||||
@@ -0,0 +1,243 @@
|
|||||||
|
# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
|
||||||
|
# SPDX-License-Identifier: Apache-2.0
|
||||||
|
#
|
||||||
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||||
|
# you may not use this file except in compliance with the License.
|
||||||
|
# You may obtain a copy of the License at
|
||||||
|
#
|
||||||
|
# http://www.apache.org/licenses/LICENSE-2.0
|
||||||
|
#
|
||||||
|
# Unless required by applicable law or agreed to in writing, software
|
||||||
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||||
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||||
|
# See the License for the specific language governing permissions and
|
||||||
|
# limitations under the License.
|
||||||
|
|
||||||
|
aerial_sdk_version: 25-3-cubb
|
||||||
|
l2adapter_filename: l2_adapter_config_P5G_GH.yaml
|
||||||
|
aerial_metrics_backend_address: 127.0.0.1:8081
|
||||||
|
|
||||||
|
# CPU core shared by all low-priority threads
|
||||||
|
low_priority_core: 10
|
||||||
|
enable_ptp_svc_monitoring: 0
|
||||||
|
ptp_rms_threshold: 20 #ns
|
||||||
|
nic_tput_alert_threshold_mbps: 85000
|
||||||
|
|
||||||
|
cuphydriver_config:
|
||||||
|
standalone: 0
|
||||||
|
validation: 0
|
||||||
|
num_slots: 8
|
||||||
|
log_level: DBG
|
||||||
|
profiler_sec: 0
|
||||||
|
dpdk_thread: 10
|
||||||
|
dpdk_verbose_logs: 0
|
||||||
|
accu_tx_sched_res_ns: 500
|
||||||
|
accu_tx_sched_disable: 0 #Flag applicable only for CX6-DX : BF3/CX7 HW and beyond support accurate send scheduling directly on timestamp, for CX6-DX this is implemented via software emulation
|
||||||
|
fh_stats_dump_cpu_core: -1
|
||||||
|
pdump_client_thread: -1
|
||||||
|
use_green_contexts: 0
|
||||||
|
use_batched_memcpy: 0 # Experimental feature; requires >= CUDA 12.8
|
||||||
|
mps_sm_pusch: 100
|
||||||
|
mps_sm_pucch: 2
|
||||||
|
mps_sm_prach: 2
|
||||||
|
mps_sm_ul_order: 20
|
||||||
|
mps_sm_pdsch: 102
|
||||||
|
mps_sm_pdcch: 10
|
||||||
|
mps_sm_pbch: 2
|
||||||
|
mps_sm_gpu_comms: 16
|
||||||
|
mps_sm_srs: 16
|
||||||
|
pdsch_fallback: 0
|
||||||
|
dpdk_file_prefix: cuphycontroller
|
||||||
|
nics:
|
||||||
|
- nic: '0000:01:00.0'
|
||||||
|
mtu: 9216
|
||||||
|
cpu_mbufs: 196608
|
||||||
|
uplane_tx_handles: 64
|
||||||
|
txq_size: 8192
|
||||||
|
rxq_size: 16384
|
||||||
|
gpu: 0
|
||||||
|
cus_port_failover: 0
|
||||||
|
gpus:
|
||||||
|
- 0
|
||||||
|
# Set GPUID to the GPU sharing the PCIe switch as NIC
|
||||||
|
# run nvidia-smi topo -m to find out which GPU
|
||||||
|
workers_ul: [4,5]
|
||||||
|
workers_dl: [6,7,8]
|
||||||
|
debug_worker: -1
|
||||||
|
workers_sched_priority: 95
|
||||||
|
prometheus_thread: -1
|
||||||
|
start_section_id_srs: 3072
|
||||||
|
start_section_id_prach: 2048
|
||||||
|
enable_ul_cuphy_graphs: 1
|
||||||
|
enable_dl_cuphy_graphs: 1
|
||||||
|
ul_order_kernel_mode: 0 # 0: Ping-Pong, 1: dual CTA
|
||||||
|
ul_order_timeout_cpu_ns: 8000000
|
||||||
|
ul_order_timeout_log_interval_ns: 1000000000
|
||||||
|
ul_order_timeout_gpu_ns: 3000000
|
||||||
|
ul_order_timeout_gpu_srs_ns: 5200000
|
||||||
|
ul_srs_aggr3_task_launch_offset_ns: 500000
|
||||||
|
ul_order_timeout_gpu_log_enable: 0
|
||||||
|
ul_order_max_rx_pkts: 512
|
||||||
|
ul_order_rx_pkts_timeout_ns: 100000
|
||||||
|
cplane_disable: 0
|
||||||
|
gpu_init_comms_dl: 1
|
||||||
|
cell_group: 1
|
||||||
|
cell_group_num: 1
|
||||||
|
fix_beta_dl: 0
|
||||||
|
bfw_beta_prescaler: 16384
|
||||||
|
pusch_workCancelMode: 2 # set cuphyPuschStatPrms_t workCancelMode field to PUSCH_DEVICE_GRAPHS; this only sets the mode and does not cancel work;
|
||||||
|
# work cancellation (i.e., not launching some of the device graphs) can only occur after an explicit internal
|
||||||
|
# memory write is timely observed by a PUSCH graph.
|
||||||
|
pusch_sinr: 2
|
||||||
|
pusch_rssi: 1
|
||||||
|
pusch_tdi: 1
|
||||||
|
pusch_cfo: 1
|
||||||
|
pusch_dftsofdm: 0
|
||||||
|
pusch_to: 1
|
||||||
|
pusch_select_eqcoeffalgo: 3
|
||||||
|
pusch_select_chestalgo: 1
|
||||||
|
pusch_enable_perprgchest: 0
|
||||||
|
pusch_tbsizecheck: 1
|
||||||
|
pusch_deviceGraphLaunchEn: 1
|
||||||
|
pusch_waitTimeOutPreEarlyHarqUs: 3000
|
||||||
|
pusch_waitTimeOutPostEarlyHarqUs: 3000
|
||||||
|
puxch_polarDcdrListSz: 8
|
||||||
|
enable_cpu_task_tracing: 0
|
||||||
|
enable_prepare_tracing: 0
|
||||||
|
disable_empw: 0 #1=>Disables Multi packet WQE feature
|
||||||
|
cqe_tracer_config:
|
||||||
|
enable_dl_cqe_tracing: 0
|
||||||
|
cqe_trace_cell_mask: 1 #[64 bit mask: Bit0->Cell0, Bit1->Cell1...]
|
||||||
|
cqe_trace_slot_mask: 196800 #[20 bit mask: Bit0->Slot0, Bit1->Slot1...] 196800=>All *6,*7 DL Slots
|
||||||
|
ul_rx_pkt_tracing_level: 0
|
||||||
|
ul_rx_pkt_tracing_level_srs: 0
|
||||||
|
pmu_metrics: 0 #Enable and select PMU metrics, 0=disabled, 1=General Counters (platform agnostic), 2=Topdown Metrics (Grace only), 3=Cache Metrics (Grace only)
|
||||||
|
enable_h2d_copy_thread: 1
|
||||||
|
h2d_copy_thread_cpu_affinity : 11
|
||||||
|
h2d_copy_thread_sched_priority : 95 #0->SCHED_OTHER, >0->Actual
|
||||||
|
split_ul_cuda_streams: 0 # 1=Put UL slot 4 and slot 5 on different streams for DDDSUUDDDD pattern
|
||||||
|
serialize_pucch_pusch: 0 # 1=Force serialization of PUSCH/PUCCH
|
||||||
|
# Note: for Early Harq (EH) order is PUSCH EH -> PUCCH -> PUSCH non EH
|
||||||
|
# for non-Early Harq order is PUCCH -> all PUSCH processing
|
||||||
|
aggr_obj_non_avail_th: 5 # Threshold for consecutive non-availability of Aggregated objects(UL/DL) or DL/UL buffers
|
||||||
|
dl_wait_th_ns:
|
||||||
|
- 500000 #H2D copy wait threshold
|
||||||
|
- 4000000 #cuPHY DL channel wait threshold
|
||||||
|
sendCPlane_timing_error_th_ns : 0
|
||||||
|
sendCPlane_ulbfw_backoff_th_ns: 300000 # ULBFW must be completed by cplane send time - this is the backoff time
|
||||||
|
sendCPlane_dlbfw_backoff_th_ns: 300000 # DLBFW must be completed by cplane send time - this is the backoff time
|
||||||
|
pusch_forcedNumCsi2Bits: 0
|
||||||
|
mMIMO_enable: 0
|
||||||
|
enable_srs: 1
|
||||||
|
ue_mode: 0
|
||||||
|
mCh_segment_proc_enable: 0
|
||||||
|
pusch_aggr_per_ctx: 3
|
||||||
|
prach_aggr_per_ctx: 2
|
||||||
|
pucch_aggr_per_ctx: 4
|
||||||
|
srs_aggr_per_ctx: 3
|
||||||
|
ul_input_buffer_per_cell: 10
|
||||||
|
ul_input_buffer_per_cell_srs: 6
|
||||||
|
max_harq_pools: 384
|
||||||
|
max_harq_tx_count_bundled: 10
|
||||||
|
max_harq_tx_count_non_bundled: 4
|
||||||
|
ok_testbench_config:
|
||||||
|
enable_ok_tb: 0
|
||||||
|
num_ok_tb_slot: 0 #Presently not used
|
||||||
|
max_ru_unhealthy_ul_slots: 100
|
||||||
|
srs_chest_algo_type: 0
|
||||||
|
srs_chest_tol2_normalization_algo_type: 1
|
||||||
|
srs_chest_tol2_constant_scaler: 32768.0
|
||||||
|
bfw_power_normalization_alg_selector: 1
|
||||||
|
total_num_srs_chest_buffers: 6144
|
||||||
|
send_static_bfw_wt_all_cplane: 1
|
||||||
|
ul_pcap_capture_enable: 0
|
||||||
|
ul_pcap_capture_thread_cpu_affinity: 10
|
||||||
|
ul_pcap_capture_thread_sched_priority: 95
|
||||||
|
pcap_logger_ul_cplane_enable: 0
|
||||||
|
pcap_logger_dl_cplane_enable: 0
|
||||||
|
pcap_logger_thread_cpu_affinity: 10
|
||||||
|
pcap_logger_thread_sched_prio: 95
|
||||||
|
pcap_logger_file_save_dir: ./
|
||||||
|
dlc_bfw_enable_divide_per_cell: 0 # Enable/disable dividing DL BFW C-Plane transmission window among multiple cells in a slot
|
||||||
|
ulc_bfw_enable_divide_per_cell: 0 # Enable/disable dividing UL BFW C-Plane transmission window among multiple cells in a slot
|
||||||
|
dlc_alloc_cplane_bfw_txq: 0 # Allocates a second separate queue for BFW DL C-Plane transmission
|
||||||
|
ulc_alloc_cplane_bfw_txq: 1 # Allocates a second separate queue for BFW UL C-Plane transmission
|
||||||
|
#static beam id range
|
||||||
|
static_beam_id_start: 1
|
||||||
|
static_beam_id_end: 16527
|
||||||
|
#dynamic beam id range
|
||||||
|
dynamic_beam_id_start: 16528 #Dynamic beam ID start should be after static beam ID end
|
||||||
|
dynamic_beam_id_end: 32767 #15bit max value
|
||||||
|
bfw_c_plane_chaining_mode: 0 # 0: Default memcpy of BFW weights, 1: CPU mbuf chaining of BFW weights, 2: GPU mbuf chaining
|
||||||
|
enable_l1_param_sanity_check: 0
|
||||||
|
pusch_nMaxLdpcHetConfigs: 32
|
||||||
|
notify_ul_harq_buffer_release: 0
|
||||||
|
|
||||||
|
# data_config:
|
||||||
|
# datalake_core: 12 # This must be set for Data Lake and/or E3 Agent to be used
|
||||||
|
# datalake_db_write_enable: 1 # Enable/disable ClickHouse writes based on datalake_data_types
|
||||||
|
# datalake_samples: 1000000
|
||||||
|
# datalake_address: localhost
|
||||||
|
# datalake_engine: "MergeTree() PRIMARY KEY (TsTaiNs)"
|
||||||
|
# # datalake_drop_tables: 0 # Set to 1 to drop Clickhouse tables at startup
|
||||||
|
# datalake_data_types: [fh, pusch, hest] # Data types to write to ClickHouse DB (all types always collected in memory for E3 Agent):
|
||||||
|
# fh = Fronthaul IQ samples,pusch = PUSCH/FAPI data, hest = channel estimates
|
||||||
|
# datalake_store_failed_pdu: 0 # Set to 1 to store PDU data even when CRC fails
|
||||||
|
# e3_agent_enable: 0 # Enable/disable E3 Agent and dApp capabilities
|
||||||
|
# num_rows_fh: 120
|
||||||
|
# num_rows_pusch: 200
|
||||||
|
# num_rows_hest: 140
|
||||||
|
# e3_pub_port: 5555 # E3 publisher port for indication messages
|
||||||
|
# e3_rep_port: 5556 # E3 reply port for E3AP requests
|
||||||
|
# e3_sub_port: 5560 # E3 subscriber port for Manager messages
|
||||||
|
|
||||||
|
cells:
|
||||||
|
- name: O-RU 0
|
||||||
|
cell_id: 1
|
||||||
|
ru_type: 1
|
||||||
|
# set to 00:00:00:00:00:00 to use the MAC address of the NIC port to use
|
||||||
|
src_mac_addr: 00:00:00:00:00:00
|
||||||
|
dst_mac_addr: e8:c7:cf:ad:12:0b #WNC n78
|
||||||
|
nic: '0000:01:00.0'
|
||||||
|
vlan: 500
|
||||||
|
pcp: 7
|
||||||
|
txq_count_uplane: 1
|
||||||
|
eAxC_id_ssb_pbch: [0, 1, 2, 3]
|
||||||
|
eAxC_id_pdcch: [0, 1, 2, 3]
|
||||||
|
eAxC_id_pdsch: [0, 1, 2, 3]
|
||||||
|
eAxC_id_csirs: [0, 1, 2, 3]
|
||||||
|
eAxC_id_pusch: [0, 1, 2, 3]
|
||||||
|
eAxC_id_pucch: [0, 1, 2, 3]
|
||||||
|
eAxC_id_srs: [8, 9, 10, 11]
|
||||||
|
eAxC_id_prach: [4, 5, 6, 7]
|
||||||
|
dl_iq_data_fmt: {comp_meth: 1, bit_width: 9}
|
||||||
|
ul_iq_data_fmt: {comp_meth: 1, bit_width: 9}
|
||||||
|
section_3_time_offset: 484
|
||||||
|
fs_offset_dl: 7
|
||||||
|
exponent_dl: 4
|
||||||
|
ref_dl: 0
|
||||||
|
fs_offset_ul: -5
|
||||||
|
exponent_ul: 4
|
||||||
|
max_amp_ul: 65504
|
||||||
|
mu: 1
|
||||||
|
T1a_max_up_ns: 280000
|
||||||
|
T1a_max_cp_ul_ns: 405000
|
||||||
|
Ta4_min_ns: 50000
|
||||||
|
Ta4_max_ns: 331000
|
||||||
|
Tcp_adv_dl_ns: 125000
|
||||||
|
ul_u_plane_tx_offset_ns: 280000
|
||||||
|
fh_len_range: 0
|
||||||
|
pusch_prb_stride: 273
|
||||||
|
prach_prb_stride: 12
|
||||||
|
srs_prb_stride: 273
|
||||||
|
pusch_ldpc_max_num_itr_algo_type: 1
|
||||||
|
pusch_fixed_max_num_ldpc_itrs: 10
|
||||||
|
pusch_ldpc_early_termination: 0
|
||||||
|
pusch_ldpc_algo_index: 0
|
||||||
|
pusch_ldpc_flags: 2
|
||||||
|
pusch_ldpc_use_half: 1
|
||||||
|
pusch_nMaxPrb: 273
|
||||||
|
pusch_nMaxRx: 4
|
||||||
|
ul_gain_calibration: 78.68
|
||||||
|
lower_guard_bw: 845
|
||||||
|
tv_pusch: cuPhyChEstCoeffs.h5
|
||||||
@@ -18,12 +18,13 @@ services:
|
|||||||
- /lib/modules:/lib/modules
|
- /lib/modules:/lib/modules
|
||||||
- /dev/hugepages:/dev/hugepages
|
- /dev/hugepages:/dev/hugepages
|
||||||
- /usr/src:/usr/src
|
- /usr/src:/usr/src
|
||||||
|
- ./cuphycontroller_P5G_WNC_GH.yaml:/opt/nvidia/cuBB/cuPHY-CP/cuphycontroller/config/cuphycontroller_P5G_WNC_GH.yaml
|
||||||
- ./aerial_l1_entrypoint.sh:/opt/nvidia/cuBB/aerial_l1_entrypoint.sh
|
- ./aerial_l1_entrypoint.sh:/opt/nvidia/cuBB/aerial_l1_entrypoint.sh
|
||||||
- /var/log/aerial:/var/log/aerial
|
- /var/log/aerial:/var/log/aerial
|
||||||
- ../../../cmake_targets/share:/opt/cuBB/share
|
- ../../../cmake_targets/share:/opt/cuBB/share
|
||||||
userns_mode: host
|
userns_mode: host
|
||||||
ipc: "shareable"
|
ipc: "shareable"
|
||||||
image: cubb-build:25-2.1
|
image: cubb-build:25-3
|
||||||
environment:
|
environment:
|
||||||
- cuBB_SDK=/opt/nvidia/cuBB
|
- cuBB_SDK=/opt/nvidia/cuBB
|
||||||
command: bash -c "sudo rm -rf /tmp/phy.log && sudo chmod +x /opt/nvidia/cuBB/aerial_l1_entrypoint.sh && /opt/nvidia/cuBB/aerial_l1_entrypoint.sh"
|
command: bash -c "sudo rm -rf /tmp/phy.log && sudo chmod +x /opt/nvidia/cuBB/aerial_l1_entrypoint.sh && /opt/nvidia/cuBB/aerial_l1_entrypoint.sh"
|
||||||
@@ -45,7 +46,7 @@ services:
|
|||||||
- IPC_LOCK
|
- IPC_LOCK
|
||||||
ipc: "container:nv-cubb"
|
ipc: "container:nv-cubb"
|
||||||
environment:
|
environment:
|
||||||
USE_ADDITIONAL_OPTIONS: --log_config.global_log_options level,nocolor,time
|
USE_ADDITIONAL_OPTIONS: --log_config.global_log_options level,nocolor,utc_time
|
||||||
deploy:
|
deploy:
|
||||||
resources:
|
resources:
|
||||||
reservations:
|
reservations:
|
||||||
|
|||||||
@@ -4,8 +4,6 @@
|
|||||||
cuBB_Path="${cuBB_SDK:-/opt/nvidia/cuBB}"
|
cuBB_Path="${cuBB_SDK:-/opt/nvidia/cuBB}"
|
||||||
|
|
||||||
cd "$cuBB_Path" || exit 1
|
cd "$cuBB_Path" || exit 1
|
||||||
# Add gdrcopy to LD_LIBRARY_PATH
|
|
||||||
export LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/opt/mellanox/dpdk/lib/x86_64-linux-gnu:/opt/mellanox/doca/lib/x86_64-linux-gnu:/opt/nvidia/cuBB/cuPHY-CP/external/gdrcopy/build/x86_64/
|
|
||||||
|
|
||||||
# Restart MPS
|
# Restart MPS
|
||||||
# Export variables
|
# Export variables
|
||||||
@@ -14,7 +12,7 @@ export CUDA_MPS_PIPE_DIRECTORY=/var
|
|||||||
export CUDA_MPS_LOG_DIRECTORY=/var
|
export CUDA_MPS_LOG_DIRECTORY=/var
|
||||||
|
|
||||||
# Stop existing MPS
|
# Stop existing MPS
|
||||||
sudo -E echo quit | sudo -E nvidia-cuda-mps-control
|
echo quit | sudo -E nvidia-cuda-mps-control
|
||||||
|
|
||||||
# Start MPS
|
# Start MPS
|
||||||
sudo -E nvidia-cuda-mps-control -d
|
sudo -E nvidia-cuda-mps-control -d
|
||||||
@@ -24,11 +22,11 @@ sudo -E echo start_server -uid 0 | sudo -E nvidia-cuda-mps-control
|
|||||||
# Check if an argument is provided
|
# Check if an argument is provided
|
||||||
if [ $# -eq 0 ]; then
|
if [ $# -eq 0 ]; then
|
||||||
# No argument provided, use default value
|
# No argument provided, use default value
|
||||||
argument="P5G_FXN_GH"
|
argument="P5G_WNC_GH"
|
||||||
else
|
else
|
||||||
# Argument provided, use it
|
# Argument provided, use it
|
||||||
argument="$1"
|
argument="$1"
|
||||||
fi
|
fi
|
||||||
|
|
||||||
sudo -E "$cuBB_Path"/build/cuPHY-CP/cuphycontroller/examples/cuphycontroller_scf "$argument"
|
sudo -E "$cuBB_Path"/build.$(uname -m)/cuPHY-CP/cuphycontroller/examples/cuphycontroller_scf "$argument"
|
||||||
sudo -E ./build/cuPHY-CP/gt_common_libs/nvIPC/tests/pcap/pcap_collect
|
sudo -E ./build.$(uname -m)/cuPHY-CP/gt_common_libs/nvIPC/tests/pcap/pcap_collect
|
||||||
|
|||||||
@@ -0,0 +1,243 @@
|
|||||||
|
# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
|
||||||
|
# SPDX-License-Identifier: Apache-2.0
|
||||||
|
#
|
||||||
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||||
|
# you may not use this file except in compliance with the License.
|
||||||
|
# You may obtain a copy of the License at
|
||||||
|
#
|
||||||
|
# http://www.apache.org/licenses/LICENSE-2.0
|
||||||
|
#
|
||||||
|
# Unless required by applicable law or agreed to in writing, software
|
||||||
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||||
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||||
|
# See the License for the specific language governing permissions and
|
||||||
|
# limitations under the License.
|
||||||
|
|
||||||
|
aerial_sdk_version: 25-3-cubb
|
||||||
|
l2adapter_filename: l2_adapter_config_P5G_GH.yaml
|
||||||
|
aerial_metrics_backend_address: 127.0.0.1:8081
|
||||||
|
|
||||||
|
# CPU core shared by all low-priority threads
|
||||||
|
low_priority_core: 10
|
||||||
|
enable_ptp_svc_monitoring: 0
|
||||||
|
ptp_rms_threshold: 20 #ns
|
||||||
|
nic_tput_alert_threshold_mbps: 85000
|
||||||
|
|
||||||
|
cuphydriver_config:
|
||||||
|
standalone: 0
|
||||||
|
validation: 0
|
||||||
|
num_slots: 8
|
||||||
|
log_level: DBG
|
||||||
|
profiler_sec: 0
|
||||||
|
dpdk_thread: 10
|
||||||
|
dpdk_verbose_logs: 0
|
||||||
|
accu_tx_sched_res_ns: 500
|
||||||
|
accu_tx_sched_disable: 0 #Flag applicable only for CX6-DX : BF3/CX7 HW and beyond support accurate send scheduling directly on timestamp, for CX6-DX this is implemented via software emulation
|
||||||
|
fh_stats_dump_cpu_core: -1
|
||||||
|
pdump_client_thread: -1
|
||||||
|
use_green_contexts: 0
|
||||||
|
use_batched_memcpy: 0 # Experimental feature; requires >= CUDA 12.8
|
||||||
|
mps_sm_pusch: 100
|
||||||
|
mps_sm_pucch: 2
|
||||||
|
mps_sm_prach: 2
|
||||||
|
mps_sm_ul_order: 20
|
||||||
|
mps_sm_pdsch: 102
|
||||||
|
mps_sm_pdcch: 10
|
||||||
|
mps_sm_pbch: 2
|
||||||
|
mps_sm_gpu_comms: 16
|
||||||
|
mps_sm_srs: 16
|
||||||
|
pdsch_fallback: 0
|
||||||
|
dpdk_file_prefix: cuphycontroller
|
||||||
|
nics:
|
||||||
|
- nic: '0000:01:00.0'
|
||||||
|
mtu: 9216
|
||||||
|
cpu_mbufs: 196608
|
||||||
|
uplane_tx_handles: 64
|
||||||
|
txq_size: 8192
|
||||||
|
rxq_size: 16384
|
||||||
|
gpu: 0
|
||||||
|
cus_port_failover: 0
|
||||||
|
gpus:
|
||||||
|
- 0
|
||||||
|
# Set GPUID to the GPU sharing the PCIe switch as NIC
|
||||||
|
# run nvidia-smi topo -m to find out which GPU
|
||||||
|
workers_ul: [4,5]
|
||||||
|
workers_dl: [6,7,8]
|
||||||
|
debug_worker: -1
|
||||||
|
workers_sched_priority: 95
|
||||||
|
prometheus_thread: -1
|
||||||
|
start_section_id_srs: 3072
|
||||||
|
start_section_id_prach: 2048
|
||||||
|
enable_ul_cuphy_graphs: 1
|
||||||
|
enable_dl_cuphy_graphs: 1
|
||||||
|
ul_order_kernel_mode: 0 # 0: Ping-Pong, 1: dual CTA
|
||||||
|
ul_order_timeout_cpu_ns: 8000000
|
||||||
|
ul_order_timeout_log_interval_ns: 1000000000
|
||||||
|
ul_order_timeout_gpu_ns: 3000000
|
||||||
|
ul_order_timeout_gpu_srs_ns: 5200000
|
||||||
|
ul_srs_aggr3_task_launch_offset_ns: 500000
|
||||||
|
ul_order_timeout_gpu_log_enable: 0
|
||||||
|
ul_order_max_rx_pkts: 512
|
||||||
|
ul_order_rx_pkts_timeout_ns: 100000
|
||||||
|
cplane_disable: 0
|
||||||
|
gpu_init_comms_dl: 1
|
||||||
|
cell_group: 1
|
||||||
|
cell_group_num: 1
|
||||||
|
fix_beta_dl: 0
|
||||||
|
bfw_beta_prescaler: 16384
|
||||||
|
pusch_workCancelMode: 2 # set cuphyPuschStatPrms_t workCancelMode field to PUSCH_DEVICE_GRAPHS; this only sets the mode and does not cancel work;
|
||||||
|
# work cancellation (i.e., not launching some of the device graphs) can only occur after an explicit internal
|
||||||
|
# memory write is timely observed by a PUSCH graph.
|
||||||
|
pusch_sinr: 2
|
||||||
|
pusch_rssi: 1
|
||||||
|
pusch_tdi: 1
|
||||||
|
pusch_cfo: 1
|
||||||
|
pusch_dftsofdm: 0
|
||||||
|
pusch_to: 1
|
||||||
|
pusch_select_eqcoeffalgo: 3
|
||||||
|
pusch_select_chestalgo: 1
|
||||||
|
pusch_enable_perprgchest: 0
|
||||||
|
pusch_tbsizecheck: 1
|
||||||
|
pusch_deviceGraphLaunchEn: 1
|
||||||
|
pusch_waitTimeOutPreEarlyHarqUs: 3000
|
||||||
|
pusch_waitTimeOutPostEarlyHarqUs: 3000
|
||||||
|
puxch_polarDcdrListSz: 8
|
||||||
|
enable_cpu_task_tracing: 0
|
||||||
|
enable_prepare_tracing: 0
|
||||||
|
disable_empw: 0 #1=>Disables Multi packet WQE feature
|
||||||
|
cqe_tracer_config:
|
||||||
|
enable_dl_cqe_tracing: 0
|
||||||
|
cqe_trace_cell_mask: 1 #[64 bit mask: Bit0->Cell0, Bit1->Cell1...]
|
||||||
|
cqe_trace_slot_mask: 196800 #[20 bit mask: Bit0->Slot0, Bit1->Slot1...] 196800=>All *6,*7 DL Slots
|
||||||
|
ul_rx_pkt_tracing_level: 0
|
||||||
|
ul_rx_pkt_tracing_level_srs: 0
|
||||||
|
pmu_metrics: 0 #Enable and select PMU metrics, 0=disabled, 1=General Counters (platform agnostic), 2=Topdown Metrics (Grace only), 3=Cache Metrics (Grace only)
|
||||||
|
enable_h2d_copy_thread: 1
|
||||||
|
h2d_copy_thread_cpu_affinity : 11
|
||||||
|
h2d_copy_thread_sched_priority : 95 #0->SCHED_OTHER, >0->Actual
|
||||||
|
split_ul_cuda_streams: 0 # 1=Put UL slot 4 and slot 5 on different streams for DDDSUUDDDD pattern
|
||||||
|
serialize_pucch_pusch: 0 # 1=Force serialization of PUSCH/PUCCH
|
||||||
|
# Note: for Early Harq (EH) order is PUSCH EH -> PUCCH -> PUSCH non EH
|
||||||
|
# for non-Early Harq order is PUCCH -> all PUSCH processing
|
||||||
|
aggr_obj_non_avail_th: 5 # Threshold for consecutive non-availability of Aggregated objects(UL/DL) or DL/UL buffers
|
||||||
|
dl_wait_th_ns:
|
||||||
|
- 500000 #H2D copy wait threshold
|
||||||
|
- 4000000 #cuPHY DL channel wait threshold
|
||||||
|
sendCPlane_timing_error_th_ns : 0
|
||||||
|
sendCPlane_ulbfw_backoff_th_ns: 300000 # ULBFW must be completed by cplane send time - this is the backoff time
|
||||||
|
sendCPlane_dlbfw_backoff_th_ns: 300000 # DLBFW must be completed by cplane send time - this is the backoff time
|
||||||
|
pusch_forcedNumCsi2Bits: 0
|
||||||
|
mMIMO_enable: 0
|
||||||
|
enable_srs: 1
|
||||||
|
ue_mode: 0
|
||||||
|
mCh_segment_proc_enable: 0
|
||||||
|
pusch_aggr_per_ctx: 9 #3
|
||||||
|
prach_aggr_per_ctx: 2
|
||||||
|
pucch_aggr_per_ctx: 4
|
||||||
|
srs_aggr_per_ctx: 3
|
||||||
|
ul_input_buffer_per_cell: 10
|
||||||
|
ul_input_buffer_per_cell_srs: 6
|
||||||
|
max_harq_pools: 384
|
||||||
|
max_harq_tx_count_bundled: 10
|
||||||
|
max_harq_tx_count_non_bundled: 4
|
||||||
|
ok_testbench_config:
|
||||||
|
enable_ok_tb: 0
|
||||||
|
num_ok_tb_slot: 0 #Presently not used
|
||||||
|
max_ru_unhealthy_ul_slots: 100
|
||||||
|
srs_chest_algo_type: 0
|
||||||
|
srs_chest_tol2_normalization_algo_type: 1
|
||||||
|
srs_chest_tol2_constant_scaler: 32768.0
|
||||||
|
bfw_power_normalization_alg_selector: 1
|
||||||
|
total_num_srs_chest_buffers: 6144
|
||||||
|
send_static_bfw_wt_all_cplane: 1
|
||||||
|
ul_pcap_capture_enable: 0
|
||||||
|
ul_pcap_capture_thread_cpu_affinity: 10
|
||||||
|
ul_pcap_capture_thread_sched_priority: 95
|
||||||
|
pcap_logger_ul_cplane_enable: 0
|
||||||
|
pcap_logger_dl_cplane_enable: 0
|
||||||
|
pcap_logger_thread_cpu_affinity: 10
|
||||||
|
pcap_logger_thread_sched_prio: 95
|
||||||
|
pcap_logger_file_save_dir: ./
|
||||||
|
dlc_bfw_enable_divide_per_cell: 0 # Enable/disable dividing DL BFW C-Plane transmission window among multiple cells in a slot
|
||||||
|
ulc_bfw_enable_divide_per_cell: 0 # Enable/disable dividing UL BFW C-Plane transmission window among multiple cells in a slot
|
||||||
|
dlc_alloc_cplane_bfw_txq: 0 # Allocates a second separate queue for BFW DL C-Plane transmission
|
||||||
|
ulc_alloc_cplane_bfw_txq: 1 # Allocates a second separate queue for BFW UL C-Plane transmission
|
||||||
|
#static beam id range
|
||||||
|
static_beam_id_start: 1
|
||||||
|
static_beam_id_end: 16527
|
||||||
|
#dynamic beam id range
|
||||||
|
dynamic_beam_id_start: 16528 #Dynamic beam ID start should be after static beam ID end
|
||||||
|
dynamic_beam_id_end: 32767 #15bit max value
|
||||||
|
bfw_c_plane_chaining_mode: 0 # 0: Default memcpy of BFW weights, 1: CPU mbuf chaining of BFW weights, 2: GPU mbuf chaining
|
||||||
|
enable_l1_param_sanity_check: 0
|
||||||
|
pusch_nMaxLdpcHetConfigs: 32
|
||||||
|
notify_ul_harq_buffer_release: 0
|
||||||
|
|
||||||
|
# data_config:
|
||||||
|
# datalake_core: 12 # This must be set for Data Lake and/or E3 Agent to be used
|
||||||
|
# datalake_db_write_enable: 1 # Enable/disable ClickHouse writes based on datalake_data_types
|
||||||
|
# datalake_samples: 1000000
|
||||||
|
# datalake_address: localhost
|
||||||
|
# datalake_engine: "MergeTree() PRIMARY KEY (TsTaiNs)"
|
||||||
|
# # datalake_drop_tables: 0 # Set to 1 to drop Clickhouse tables at startup
|
||||||
|
# datalake_data_types: [fh, pusch, hest] # Data types to write to ClickHouse DB (all types always collected in memory for E3 Agent):
|
||||||
|
# fh = Fronthaul IQ samples,pusch = PUSCH/FAPI data, hest = channel estimates
|
||||||
|
# datalake_store_failed_pdu: 0 # Set to 1 to store PDU data even when CRC fails
|
||||||
|
# e3_agent_enable: 0 # Enable/disable E3 Agent and dApp capabilities
|
||||||
|
# num_rows_fh: 120
|
||||||
|
# num_rows_pusch: 200
|
||||||
|
# num_rows_hest: 140
|
||||||
|
# e3_pub_port: 5555 # E3 publisher port for indication messages
|
||||||
|
# e3_rep_port: 5556 # E3 reply port for E3AP requests
|
||||||
|
# e3_sub_port: 5560 # E3 subscriber port for Manager messages
|
||||||
|
|
||||||
|
cells:
|
||||||
|
- name: O-RU 0
|
||||||
|
cell_id: 1
|
||||||
|
ru_type: 1
|
||||||
|
# set to 00:00:00:00:00:00 to use the MAC address of the NIC port to use
|
||||||
|
src_mac_addr: 00:00:00:00:00:00
|
||||||
|
dst_mac_addr: e8:c7:cf:ad:12:0b #WNC n78
|
||||||
|
nic: '0000:01:00.0'
|
||||||
|
vlan: 500
|
||||||
|
pcp: 7
|
||||||
|
txq_count_uplane: 1
|
||||||
|
eAxC_id_ssb_pbch: [0, 1, 2, 3]
|
||||||
|
eAxC_id_pdcch: [0, 1, 2, 3]
|
||||||
|
eAxC_id_pdsch: [0, 1, 2, 3]
|
||||||
|
eAxC_id_csirs: [0, 1, 2, 3]
|
||||||
|
eAxC_id_pusch: [0, 1, 2, 3]
|
||||||
|
eAxC_id_pucch: [0, 1, 2, 3]
|
||||||
|
eAxC_id_srs: [8, 9, 10, 11]
|
||||||
|
eAxC_id_prach: [4, 5, 6, 7]
|
||||||
|
dl_iq_data_fmt: {comp_meth: 1, bit_width: 9}
|
||||||
|
ul_iq_data_fmt: {comp_meth: 1, bit_width: 9}
|
||||||
|
section_3_time_offset: 484
|
||||||
|
fs_offset_dl: 7
|
||||||
|
exponent_dl: 4
|
||||||
|
ref_dl: 0
|
||||||
|
fs_offset_ul: -5
|
||||||
|
exponent_ul: 4
|
||||||
|
max_amp_ul: 65504
|
||||||
|
mu: 1
|
||||||
|
T1a_max_up_ns: 280000
|
||||||
|
T1a_max_cp_ul_ns: 405000
|
||||||
|
Ta4_min_ns: 50000
|
||||||
|
Ta4_max_ns: 331000
|
||||||
|
Tcp_adv_dl_ns: 125000
|
||||||
|
ul_u_plane_tx_offset_ns: 280000
|
||||||
|
fh_len_range: 0
|
||||||
|
pusch_prb_stride: 273
|
||||||
|
prach_prb_stride: 12
|
||||||
|
srs_prb_stride: 273
|
||||||
|
pusch_ldpc_max_num_itr_algo_type: 1
|
||||||
|
pusch_fixed_max_num_ldpc_itrs: 10
|
||||||
|
pusch_ldpc_early_termination: 0
|
||||||
|
pusch_ldpc_algo_index: 0
|
||||||
|
pusch_ldpc_flags: 2
|
||||||
|
pusch_ldpc_use_half: 1
|
||||||
|
pusch_nMaxPrb: 273
|
||||||
|
pusch_nMaxRx: 4
|
||||||
|
ul_gain_calibration: 78.68
|
||||||
|
lower_guard_bw: 845
|
||||||
|
tv_pusch: cuPhyChEstCoeffs.h5
|
||||||
@@ -18,12 +18,13 @@ services:
|
|||||||
- /lib/modules:/lib/modules
|
- /lib/modules:/lib/modules
|
||||||
- /dev/hugepages:/dev/hugepages
|
- /dev/hugepages:/dev/hugepages
|
||||||
- /usr/src:/usr/src
|
- /usr/src:/usr/src
|
||||||
|
- ./cuphycontroller_P5G_WNC_GH_ul.yaml:/opt/nvidia/cuBB/cuPHY-CP/cuphycontroller/config/cuphycontroller_P5G_WNC_GH.yaml
|
||||||
- ./aerial_l1_entrypoint.sh:/opt/nvidia/cuBB/aerial_l1_entrypoint.sh
|
- ./aerial_l1_entrypoint.sh:/opt/nvidia/cuBB/aerial_l1_entrypoint.sh
|
||||||
- /var/log/aerial:/var/log/aerial
|
- /var/log/aerial:/var/log/aerial
|
||||||
- ../../../cmake_targets/share:/opt/cuBB/share
|
- ../../../cmake_targets/share:/opt/cuBB/share
|
||||||
userns_mode: host
|
userns_mode: host
|
||||||
ipc: "shareable"
|
ipc: "shareable"
|
||||||
image: cubb-build:25-2.1_srs-ul-heavy
|
image: cubb-build:25-3
|
||||||
environment:
|
environment:
|
||||||
- cuBB_SDK=/opt/nvidia/cuBB
|
- cuBB_SDK=/opt/nvidia/cuBB
|
||||||
command: bash -c "sudo rm -rf /tmp/phy.log && sudo chmod +x /opt/nvidia/cuBB/aerial_l1_entrypoint.sh && /opt/nvidia/cuBB/aerial_l1_entrypoint.sh"
|
command: bash -c "sudo rm -rf /tmp/phy.log && sudo chmod +x /opt/nvidia/cuBB/aerial_l1_entrypoint.sh && /opt/nvidia/cuBB/aerial_l1_entrypoint.sh"
|
||||||
@@ -45,7 +46,7 @@ services:
|
|||||||
- IPC_LOCK
|
- IPC_LOCK
|
||||||
ipc: "container:nv-cubb"
|
ipc: "container:nv-cubb"
|
||||||
environment:
|
environment:
|
||||||
USE_ADDITIONAL_OPTIONS: --log_config.global_log_options level,nocolor,time
|
USE_ADDITIONAL_OPTIONS: --log_config.global_log_options level,nocolor,utc_time
|
||||||
deploy:
|
deploy:
|
||||||
resources:
|
resources:
|
||||||
reservations:
|
reservations:
|
||||||
|
|||||||
+190
-104
@@ -1,4 +1,4 @@
|
|||||||
# OAI 7.2 Fronthaul Interface 5G SA Tutorial
|
# OAI - Aerial FAPI Split Tutorial
|
||||||
|
|
||||||
**Table of Contents**
|
**Table of Contents**
|
||||||
|
|
||||||
@@ -14,9 +14,12 @@ The hardware on which we have tried this tutorial:
|
|||||||
| Dell PowerEdge R750 (Dual Intel Xeon Gold 6336Y CPU @ 2.4G, 24C/48T (185W), 512GB RDIMM, 3200MT/s) |Ubuntu 22.04.3 LTS (5.15.0-72-lowlatency)| NVIDIA Converged Accelerator A100X (24.39.2048) |
|
| Dell PowerEdge R750 (Dual Intel Xeon Gold 6336Y CPU @ 2.4G, 24C/48T (185W), 512GB RDIMM, 3200MT/s) |Ubuntu 22.04.3 LTS (5.15.0-72-lowlatency)| NVIDIA Converged Accelerator A100X (24.39.2048) |
|
||||||
| Supermicro Grace Hopper MGX ARS-111GL-NHR (Neoverse-V2, 3.4GHz, 72C/72T, 576GB LPDDR5) | Ubuntu 22.04.5 LTS (6.5.0-1019-nvidia-64k) |NVIDIA BlueField3 (32.41.1000)|
|
| Supermicro Grace Hopper MGX ARS-111GL-NHR (Neoverse-V2, 3.4GHz, 72C/72T, 576GB LPDDR5) | Ubuntu 22.04.5 LTS (6.5.0-1019-nvidia-64k) |NVIDIA BlueField3 (32.41.1000)|
|
||||||
|
|
||||||
**NOTE**:
|
**Note**:
|
||||||
- These are not minimum hardware requirements. This is the configuration of our servers. The NIC card should support hardware PTP time stamping.
|
- These are not minimum hardware requirements. This is the configuration of our
|
||||||
- Starting from tag [2025.w13](https://gitlab.eurecom.fr/oai/openairinterface5g/-/tree/2025.w13?ref_type=tags) of OAI, we are only testing with the Grace Hopper server.
|
servers. The NIC card should support hardware PTP time stamping.
|
||||||
|
- Starting from tag
|
||||||
|
[2025.w13](https://gitlab.eurecom.fr/oai/openairinterface5g/-/tree/2025.w13?ref_type=tags)
|
||||||
|
of OAI, we are only testing with the Grace Hopper server.
|
||||||
|
|
||||||
PTP enabled switches and grandmaster clock we have tested with:
|
PTP enabled switches and grandmaster clock we have tested with:
|
||||||
|
|
||||||
@@ -54,10 +57,10 @@ The UEs that have been tested and confirmed working with Aerial are the followin
|
|||||||
To set up the L1 and install the components manually refer to this [instructions page](https://docs.nvidia.com/aerial/cuda-accelerated-ran/index.html).
|
To set up the L1 and install the components manually refer to this [instructions page](https://docs.nvidia.com/aerial/cuda-accelerated-ran/index.html).
|
||||||
|
|
||||||
**Note**:
|
**Note**:
|
||||||
- As of wk36, the L1 must be compiled with the following CMake flag: `-DSCF_FAPI_10_04_SRS=ON` , this is due to the usage
|
- To configure the Gigabyte server please refer to these
|
||||||
of the FAPI 10.04 version of the SRS PDU, and RX_Beamforming PDU.
|
[instructions](https://gitlab.eurecom.fr/oai/openairinterface5g/-/blob/2025.w13/doc/Aerial_FAPI_Split_Tutorial.md)
|
||||||
- To configure the Gigabyte server please refer to these [instructions](https://gitlab.eurecom.fr/oai/openairinterface5g/-/blob/2025.w13/doc/Aerial_FAPI_Split_Tutorial.md)
|
- The last release to support the Gigabyte server is **Aerial CUDA-Accelerated
|
||||||
- The last release to support the Gigabyte server is **Aerial CUDA-Accelerated RAN 24-1**.
|
RAN 24-1**.
|
||||||
|
|
||||||
#### CPU allocation
|
#### CPU allocation
|
||||||
|
|
||||||
@@ -67,21 +70,43 @@ of the FAPI 10.04 version of the SRS PDU, and RX_Beamforming PDU.
|
|||||||
|
|
||||||
**Grace Hopper MGX ARS-111GL-NHR**
|
**Grace Hopper MGX ARS-111GL-NHR**
|
||||||
|
|
||||||
| Applicative Threads | Allocated CPUs |
|
| Applicative Threads | Allocated CPUs |
|
||||||
|------------------------|----------------|
|
|--------------------------------------|--------------------------------|
|
||||||
| PTP & PHC2SYS Services | 41 |
|
| `workers_ul` | 4, 5 |
|
||||||
| OAI `nr-softmodem` | 13-16 |
|
| `workers_dl` | 6, 7, 8 |
|
||||||
|
| `timer_thread_config` | 9 |
|
||||||
|
| `message_thread_config` | 9 |
|
||||||
|
| `pcap_shm_caching_cpu_core` | 10 |
|
||||||
|
| `pcap_file_saving_cpu_core` | 10 |
|
||||||
|
| `dpdk_thread` | 10 |
|
||||||
|
| `ul_pcap_capture_thread_cpu_affinity`| 10 |
|
||||||
|
| `pcap_logger_thread_cpu_affinity` | 10 |
|
||||||
|
| `h2d_copy_thread_cpu_affinity` | 11 |
|
||||||
|
| `fh_stats_dump_cpu_core` | -1 |
|
||||||
|
| `pdump_client_thread` | -1 |
|
||||||
|
| `debug_worker` | -1 |
|
||||||
|
| `prometheus_thread` | -1 |
|
||||||
|
| | |
|
||||||
|
| OAI `nr-softmodem` | 13,14,15,16 |
|
||||||
|
|
||||||
### PTP configuration
|
**Note**:
|
||||||
|
- `-1` indicates that no explicit CPU pinning is applied and the thread
|
||||||
|
is scheduled by the Linux kernel.
|
||||||
|
- core 10 is a `low_priority_core` and it is shared by all low-priority threads.
|
||||||
|
|
||||||
1. You need to install the `linuxptp` debian package. It will install both ptp4l and phc2sys.
|
#### PTP configuration
|
||||||
|
|
||||||
|
1. Install the `linuxptp` debian package. It will install both
|
||||||
|
ptp4l and phc2sys.
|
||||||
|
|
||||||
```bash
|
```bash
|
||||||
#Ubuntu
|
#Ubuntu
|
||||||
sudo apt install linuxptp -y
|
sudo apt install linuxptp -y
|
||||||
```
|
```
|
||||||
|
|
||||||
Once installed you can use this configuration file for ptp4l (`/etc/ptp4l.conf`). Here the clock domain is 24 so you can adjust it according to your PTP GM clock domain
|
Once installed you can use this configuration file for ptp4l
|
||||||
|
(`/etc/ptp4l.conf`). Here the clock domain is 24 so you can adjust it according
|
||||||
|
to your PTP GM clock domain
|
||||||
|
|
||||||
```
|
```
|
||||||
[global]
|
[global]
|
||||||
@@ -94,25 +119,27 @@ tx_timestamp_timeout 30
|
|||||||
network_transport L2
|
network_transport L2
|
||||||
|
|
||||||
```
|
```
|
||||||
|
The service of ptp4l (`/lib/systemd/system/ptp4l.service`) should be configured
|
||||||
|
as below:
|
||||||
|
|
||||||
The service of ptp4l (`/lib/systemd/system/ptp4l.service`) should be configured as below:
|
|
||||||
```
|
```
|
||||||
[Unit]
|
[Unit] Description=Precision Time Protocol (PTP) service
|
||||||
Description=Precision Time Protocol (PTP) service
|
|
||||||
Documentation=man:ptp4l
|
Documentation=man:ptp4l
|
||||||
|
|
||||||
[Service]
|
[Service]
|
||||||
Restart=always
|
Restart=always
|
||||||
RestartSec=5s
|
RestartSec=5s
|
||||||
Type=simple
|
Type=simple
|
||||||
ExecStart=taskset -c 41 /usr/sbin/ptp4l -f /etc/ptp.conf
|
ExecStart=/usr/sbin/ptp4l -f /etc/ptp.conf
|
||||||
|
|
||||||
[Install]
|
[Install]
|
||||||
WantedBy=multi-user.target
|
WantedBy=multi-user.target
|
||||||
|
|
||||||
```
|
```
|
||||||
|
|
||||||
and service of phc2sys (`/lib/systemd/system/phc2sys.service`) should be configured as below:
|
and service of phc2sys (`/lib/systemd/system/phc2sys.service`) should be
|
||||||
|
configured as below:
|
||||||
|
|
||||||
```
|
```
|
||||||
[Unit]
|
[Unit]
|
||||||
Description=Synchronize system clock or PTP hardware clock (PHC)
|
Description=Synchronize system clock or PTP hardware clock (PHC)
|
||||||
@@ -131,10 +158,25 @@ ExecStart=/usr/sbin/phc2sys -a -r -n 24
|
|||||||
WantedBy=multi-user.target
|
WantedBy=multi-user.target
|
||||||
|
|
||||||
```
|
```
|
||||||
|
**Note**: As of release 25-3 there's no need to allocate a core for PTP.
|
||||||
|
|
||||||
|
## Prepare the L1
|
||||||
|
|
||||||
|
- Follow these
|
||||||
|
[instructions](https://github.com/NVIDIA/aerial-cuda-accelerated-ran) to
|
||||||
|
clone the `aerial-cuda-accelerated-ran` repository, pull and build the cuBB
|
||||||
|
image.
|
||||||
|
- The CMake flags we use to compile the L1 are:
|
||||||
|
- `-DSCF_FAPI_10_04_SRS=ON` this flag must be used as of OAI tag `2025.w36`
|
||||||
|
and this is due to the usage of the FAPI 10.04 version of the SRS PDU, and
|
||||||
|
RX_Beamforming PDU.
|
||||||
|
- `-DENABLE_CONFORMANCE_TM_PDSCH_PDCCH=OFF` this option should only be
|
||||||
|
enabled when doing conformance testing with testmac.
|
||||||
|
|
||||||
## Build OAI gNB
|
## Build OAI gNB
|
||||||
|
|
||||||
If it's not already cloned, the first step is to clone OAI repository
|
If it's not already cloned, the first step is to clone OAI repository
|
||||||
|
|
||||||
```bash
|
```bash
|
||||||
git clone https://gitlab.eurecom.fr/oai/openairinterface5g.git ~/openairinterface5g
|
git clone https://gitlab.eurecom.fr/oai/openairinterface5g.git ~/openairinterface5g
|
||||||
cd ~/openairinterface5g/
|
cd ~/openairinterface5g/
|
||||||
@@ -142,65 +184,42 @@ cd ~/openairinterface5g/
|
|||||||
|
|
||||||
### Get nvIPC sources from the L1 container
|
### Get nvIPC sources from the L1 container
|
||||||
|
|
||||||
The library used for communication between L1 and L2 components is called nvIPC, and is developed by NVIDIA. It is not open-source and can't be freely distributed.
|
The library used for communication between L1 and L2 components is called
|
||||||
In order to achieve this communication, we need to obtain the nvIPC source files from the L1 container (cuBB) and place it in the gNB project directory `~/openairinterface5g`.
|
nvIPC, and is developed by NVIDIA. In order to achieve this communication, we
|
||||||
This allows us to build and install this library when building the L2 docker container.
|
need to obtain the nvIPC source files from the L1 container (cuBB) and place it
|
||||||
|
in the gNB project directory `~/openairinterface5g`. This allows us to build
|
||||||
Check whether your L1 container is running:
|
and install this library when building the L2 docker image.
|
||||||
```bash
|
|
||||||
~$ docker container ls -a
|
|
||||||
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
|
|
||||||
a9681a0c4a10 14dca2002237 "/opt/nvidia/nvidia_…" 3 months ago Exited (137) 10 days ago cuBB
|
|
||||||
|
|
||||||
```
|
|
||||||
If it is not running, you may start it and logging into the container by running the following:
|
|
||||||
```bash
|
|
||||||
~$ docker start cuBB
|
|
||||||
cuBB
|
|
||||||
~$ docker exec -it cuBB bash
|
|
||||||
aerial@c_aerial_aerial:/opt/nvidia/cuBB#
|
|
||||||
```
|
|
||||||
|
|
||||||
After logging into the container, you need to pack the nvIPC sources and copy them to the host ( the command creates a `tar.gz` file with the following name format: `nvipc_src.YYYY.MM.DD.tar.gz`)
|
|
||||||
```bash
|
|
||||||
~$ docker exec -it cuBB bash
|
|
||||||
aerial@c_aerial_aerial:/opt/nvidia/cuBB# cd cuPHY-CP/gt_common_libs
|
|
||||||
aerial@c_aerial_aerial:/opt/nvidia/cuBB/cuPHY-CP/gt_common_libs#./pack_nvipc.sh
|
|
||||||
nvipc_src.YYYY.MM.DD/
|
|
||||||
...
|
|
||||||
---------------------------------------------
|
|
||||||
Pack nvipc source code finished:
|
|
||||||
/opt/nvidia/cuBB/cuPHY-CP/gt_common_libs/nvipc_src.YYYY.MM.DD.tar.gz
|
|
||||||
aerial@c_aerial_aerial:/opt/nvidia/cuBB/cuPHY-CP/gt_common_libs# sudo cp nvipc_src.yyyy.mm.dd.tar.gz /opt/cuBB/share/
|
|
||||||
aerial@c_aerial_aerial:/opt/nvidia/cuBB/cuPHY-CP/gt_common_libs# exit
|
|
||||||
```
|
|
||||||
|
|
||||||
The file should now be present in the `~/openairinterface5g/ci-scripts/yaml_files/sa_gnb_aerial/` directory, from where it is moved into `~/openairinterface5g`
|
|
||||||
```bash
|
|
||||||
~$ mv ~/openairinterface5g/ci-scripts/yaml_files/sa_gnb_aerial/nvipc_src.YYYY.MM.DD.tar.gz ~/openairinterface5g/
|
|
||||||
```
|
|
||||||
|
|
||||||
Alternatively, after running `./pack_nvipc.sh` , you can exit the container and copy `nvipc_src.YYYY.MM.DD.tar.gz` using `docker cp`
|
|
||||||
|
|
||||||
```bash
|
```bash
|
||||||
~$ docker exec -it cuBB bash
|
# Start interactive development container
|
||||||
aerial@c_aerial_aerial:/opt/nvidia/cuBB# cd cuPHY-CP/gt_common_libs
|
./cuPHY-CP/container/run_aerial.sh
|
||||||
aerial@c_aerial_aerial:/opt/nvidia/cuBB/cuPHY-CP/gt_common_libs#./pack_nvipc.sh
|
|
||||||
nvipc_src.YYYY.MM.DD/
|
# Pack the nvIPC sources and copy them to the host ( the command creates a `tar.gz` file with the following name format: `nvipc_src.YYYY.MM.DD.tar.gz`)
|
||||||
...
|
aerial@c_aerial_oaicicd:/opt/nvidia/cuBB# cd cuPHY-CP/gt_common_libs
|
||||||
---------------------------------------------
|
aerial@c_aerial_oaicicd:/opt/nvidia/cuBB/cuPHY-CP/gt_common_libs#./pack_nvipc.sh
|
||||||
Pack nvipc source code finished:
|
nvipc_src.YYYY.MM.DD/ ... --------------------------------------------- Pack
|
||||||
/opt/nvidia/cuBB/cuPHY-CP/gt_common_libs/nvipc_src.YYYY.MM.DD.tar.gz
|
nvipc source code finished:/opt/nvidia/cuBB/cuPHY-CP/gt_common_libs/nvipc_src.YYYY.MM.DD.tar.gz
|
||||||
aerial@c_aerial_aerial:/opt/nvidia/cuBB/cuPHY-CP/gt_common_libs# exit
|
|
||||||
~$ docker cp nv-cubb:/opt/nvidia/cuBB/cuPHY-CP/gt_common_libs/nvipc_src.YYYY.MM.DD.tar.gz ~/openairinterface5g
|
|
||||||
```
|
```
|
||||||
*Important note:* For using docker cp, make sure to copy the entire name of the created nvipc_src tar.gz file.
|
|
||||||
|
Keep the interactive development container open and in another terminal use
|
||||||
|
`docker cp` to copy the library
|
||||||
|
|
||||||
|
```bash
|
||||||
|
docker cp c_aerial_oaicicd:/opt/nvidia/cuBB/cuPHY-CP/gt_common_libs/nvipc_src.YYYY.MM.DD.tar.gz ~/openairinterface/
|
||||||
|
```
|
||||||
|
*Important note:* For using docker cp, make sure to
|
||||||
|
copy the entire name of the created nvipc_src tar.gz file.
|
||||||
|
|
||||||
|
You can now exit the interactive development container.
|
||||||
|
|
||||||
|
With the nvIPC sources in the project directory, the L2 docker image can be built.
|
||||||
|
|
||||||
With the nvIPC sources in the project directory, the docker image can be built.
|
|
||||||
|
|
||||||
### Building OAI gNB docker image
|
### Building OAI gNB docker image
|
||||||
|
|
||||||
In order to build the target image (`oai-gnb-aerial`), first you should build a common shared image (`ran-base`)
|
In order to build the target image (`oai-gnb-aerial`), first you should build a
|
||||||
|
common shared image (`ran-base`)
|
||||||
|
|
||||||
```bash
|
```bash
|
||||||
~$ cd ~/openairinterface5g/
|
~$ cd ~/openairinterface5g/
|
||||||
~/openairinterface5g$ docker build . -f docker/Dockerfile.base.ubuntu --tag ran-base:latest
|
~/openairinterface5g$ docker build . -f docker/Dockerfile.base.ubuntu --tag ran-base:latest
|
||||||
@@ -210,52 +229,113 @@ In order to build the target image (`oai-gnb-aerial`), first you should build a
|
|||||||
|
|
||||||
## Running the setup
|
## Running the setup
|
||||||
|
|
||||||
In order to use Docker Compose to automatically start and stop the setup, we first need to create a pre-built image of L1 after compiling the L1 software and making the necessary adjustments to its configuration files.
|
|
||||||
The process of preparing L1 is covered in NVIDIA's documentation and falls outside the scope of this document.
|
|
||||||
|
|
||||||
### Prepare the L1 image
|
|
||||||
After preparing the L1 software, the container needs to be committed to create an image where L1 is ready for execution, which can later be referenced by a `docker-compose.yaml` file.
|
|
||||||
|
|
||||||
*Note:* The default L1 configuration file is `cuphycontroller_P5G_FXN_GH.yaml`, located in `/opt/nvidia/cuBB/cuPHY-CP/cuphycontroller/config/`.
|
|
||||||
In this file the RU MAC address needs to be specified before commiting the image.
|
|
||||||
|
|
||||||
```bash
|
|
||||||
~$ docker commit nv-cubb cubb-build:25-2.1
|
|
||||||
~$ docker image ls
|
|
||||||
..
|
|
||||||
cubb-build 25-2.1 824156e0334c 2 weeks ago 23.9GB
|
|
||||||
-..
|
|
||||||
```
|
|
||||||
|
|
||||||
### Adapt the OAI-gNB configuration file to your system/workspace
|
### Adapt the OAI-gNB configuration file to your system/workspace
|
||||||
|
|
||||||
Edit the sample OAI gNB configuration file and check following parameters:
|
Edit the [OAI gNB configuration file](https://gitlab.eurecom.fr/oai/openairinterface5g/-/blob/develop/ci-scripts/conf_files/gnb-vnf.sa.band78.273prb.aerial.conf?ref_type=heads)
|
||||||
|
and check the following parameters:
|
||||||
|
|
||||||
* `gNBs` section
|
* `gNBs` section
|
||||||
* The PLMN section shall match the one defined in the AMF
|
* The PLMN section shall match the one defined in the AMF
|
||||||
|
* To calculate the `absoluteFrequencySSB` and `dl_absoluteFrequencyPointA`,
|
||||||
|
please follow these
|
||||||
|
[instructions](https://gitlab.eurecom.fr/oai/openairinterface5g/-/blob/develop/doc/gNB_frequency_setup.md)
|
||||||
* `amf_ip_address` shall be the correct AMF IP address in your system
|
* `amf_ip_address` shall be the correct AMF IP address in your system
|
||||||
* `GNB_IPV4_ADDRESS_FOR_NG_AMF` shall match your DU N2 interface IP address
|
* `GNB_IPV4_ADDRESS_FOR_NG_AMF` shall match your DU N2 interface IP address
|
||||||
* `GNB_IPV4_ADDRESS_FOR_NGU` shall match your DU N3 interface IP address
|
* `GNB_IPV4_ADDRESS_FOR_NGU` shall match your DU N3 interface IP address
|
||||||
|
|
||||||
The default amf_ip_address:ipv4 value is 192.168.70.132, when installing the CN5G following [this tutorial](https://gitlab.eurecom.fr/oai/openairinterface5g/-/blob/develop/doc/NR_SA_Tutorial_OAI_CN5G.md)
|
The default amf_ip_address:ipv4 value is 192.168.70.132, when installing the
|
||||||
Both 'GNB_IPV4_ADDRESS_FOR_NG_AMF' and 'GNB_IPV4_ADDRESS_FOR_NGU' need to be set to the IP address of the NIC referenced previously.
|
CN5G following [this
|
||||||
|
tutorial](https://gitlab.eurecom.fr/oai/openairinterface5g/-/blob/develop/doc/NR_SA_Tutorial_OAI_CN5G.md)
|
||||||
|
Both `GNB_IPV4_ADDRESS_FOR_NG_AMF` and `GNB_IPV4_ADDRESS_FOR_NGU` need to be
|
||||||
|
set to the IP address of the NIC referenced previously.
|
||||||
|
|
||||||
|
**Note**: If the Core Network is running on the same server, 3 cores should be
|
||||||
|
allocated to it. 2 for the UPF and 1 for the all the remaining services as shown
|
||||||
|
below.
|
||||||
|
|
||||||
### Running docker compose
|
```patch
|
||||||
#### Aerial L1 entrypoint script
|
diff --git a/docker-compose/docker-compose-basic-nrf.yaml b/docker-compose/docker-compose-basic-nrf.yaml
|
||||||
The `aerial_l1_entrypoint` script is used by the L1 container to start the L1 software and is called in the Docker Compose file.
|
index 7fbefc9..d5ef964 100644
|
||||||
It begins by setting up environment variables, restarting NVIDIA MPS, and finally running `cuphycontroller_scf`.
|
--- a/docker-compose/docker-compose-basic-nrf.yaml
|
||||||
|
+++ b/docker-compose/docker-compose-basic-nrf.yaml
|
||||||
|
@@ -28,6 +28,7 @@ services:
|
||||||
|
- 8080/tcp
|
||||||
|
volumes:
|
||||||
|
- ./conf/basic_nrf_config.yaml:/openair-udr/etc/config.yaml
|
||||||
|
+ cpuset: "15"
|
||||||
|
environment:
|
||||||
|
- TZ=Europe/Paris
|
||||||
|
depends_on:
|
||||||
|
```
|
||||||
|
|
||||||
The L1 software is executed with an argument that specifies which configuration file to use. If not modified, the default argument is set to `P5G_FXN_GH`.
|
### Aerial L1 entrypoint script
|
||||||
|
|
||||||
After building the gNB image, and preparing the configuration file, the setup can be run with the following command:
|
The `aerial_l1_entrypoint` script is used by the L1 container to start the L1
|
||||||
|
software and is mounted in the Docker Compose file. It begins by setting up
|
||||||
|
environment variables, restarting NVIDIA MPS, and finally running
|
||||||
|
`cuphycontroller_scf`.
|
||||||
|
|
||||||
|
The L1 software is executed with an argument that specifies which configuration
|
||||||
|
file to use. If not modified, the default argument is set to `P5G_WNC_GH`.
|
||||||
|
|
||||||
|
### Aerial L1 configuration
|
||||||
|
|
||||||
|
```bash
|
||||||
|
cd ~/aerial-cuda-accelerated-ran/cuPHY-CP/cuphycontroller/config/
|
||||||
|
```
|
||||||
|
|
||||||
|
Choose the configuration file that matches your setup and make the necessary
|
||||||
|
changes. We currently use `cuphycontroller_P5G_WNC_GH.yaml`.
|
||||||
|
|
||||||
|
Make sure that the following parameters are modified:
|
||||||
|
|
||||||
|
```
|
||||||
|
dst_mac_addr:
|
||||||
|
nic:
|
||||||
|
vlan:
|
||||||
|
```
|
||||||
|
|
||||||
|
### Docker compose
|
||||||
|
|
||||||
|
We recommend the user to apply the patch below for `docker-compose.yaml`.
|
||||||
|
```patch
|
||||||
|
diff --git a/ci-scripts/yaml_files/sa_gnb_aerial/docker-compose.yaml b/ci-scripts/yaml_files/sa_gnb_aerial/docker-compose.yaml
|
||||||
|
index 985fe9a6a3..0774d826ac 100644
|
||||||
|
--- a/ci-scripts/yaml_files/sa_gnb_aerial/docker-compose.yaml
|
||||||
|
+++ b/ci-scripts/yaml_files/sa_gnb_aerial/docker-compose.yaml
|
||||||
|
@@ -15,10 +15,10 @@ services:
|
||||||
|
stdin_open: true
|
||||||
|
tty: true
|
||||||
|
volumes:
|
||||||
|
+ - ~/aerial-cuda-accelerated-ran:/opt/nvidia/cuBB/
|
||||||
|
- /lib/modules:/lib/modules
|
||||||
|
- /dev/hugepages:/dev/hugepages
|
||||||
|
- /usr/src:/usr/src
|
||||||
|
- - ./cuphycontroller_P5G_WNC_GH.yaml:/opt/nvidia/cuBB/cuPHY-CP/cuphycontroller/config/cuphycontroller_P5G_WNC_GH.yaml
|
||||||
|
- ./aerial_l1_entrypoint.sh:/opt/nvidia/cuBB/aerial_l1_entrypoint.sh
|
||||||
|
- /var/log/aerial:/var/log/aerial
|
||||||
|
- ../../../cmake_targets/share:/opt/cuBB/share
|
||||||
|
```
|
||||||
|
|
||||||
|
By applying this change, it would easier for the user to modify the L1 code or
|
||||||
|
configuration files directly from the `aerial-cuda-accelerated-ran` repository.
|
||||||
|
|
||||||
|
`ci-scripts/yaml_files/sa_gnb_aerial/docker-compose.yaml` is tailored for our
|
||||||
|
CI.
|
||||||
|
We copy the `aerial-cuda-accelerated-ran` repository inside the cuBB
|
||||||
|
docker image using `docker cp` and commit the changes using `docker commit`.
|
||||||
|
We mount the `cuphycontroller_P5G_WNC_GH.yaml` in the `nv-cubb` volumes so
|
||||||
|
that it would override some parameters depending on the setup being run on CI.
|
||||||
|
|
||||||
|
After building the gNB image, and preparing the configuration file, the setup
|
||||||
|
can be run with the following command:
|
||||||
|
|
||||||
```bash
|
```bash
|
||||||
cd ci-scripts/yaml_files/sa_gnb_aerial/
|
cd ci-scripts/yaml_files/sa_gnb_aerial/
|
||||||
docker compose up -d
|
docker compose up -d
|
||||||
|
|
||||||
```
|
```
|
||||||
This will start both containers, beginning with `nv-cubb`, and `oai-gnb-aerial` will start only after it is ready.
|
This will start both containers, beginning with `nv-cubb`, and
|
||||||
|
`oai-gnb-aerial` will start only after it is ready.
|
||||||
|
|
||||||
The logs can be followed using these commands:
|
The logs can be followed using these commands:
|
||||||
|
|
||||||
@@ -263,8 +343,13 @@ The logs can be followed using these commands:
|
|||||||
docker logs -f oai-gnb-aerial
|
docker logs -f oai-gnb-aerial
|
||||||
docker logs -f nv-cubb
|
docker logs -f nv-cubb
|
||||||
```
|
```
|
||||||
|
|
||||||
#### Running with multiple L2s
|
#### Running with multiple L2s
|
||||||
One L1 instance can support multiple L2 instances. See also the [aerial documentation](https://developer.nvidia.com/docs/gputelecom/aerial-sdk/text/cubb_quickstart/running_cubb-end-to-end.html#run-multiple-l2-instances-with-single-l1-instance) for more details.
|
|
||||||
|
One L1 instance can support multiple L2 instances. See also
|
||||||
|
the [aerial
|
||||||
|
documentation](https://docs.nvidia.com/aerial/cuda-accelerated-ran/latest/quickstart_guide/running_cubb-end-to-end.html#run-multiple-l2-instances-with-single-l1-instance)
|
||||||
|
for more details.
|
||||||
|
|
||||||
In OAI the shared memory prefix must be configured in the configuration file.
|
In OAI the shared memory prefix must be configured in the configuration file.
|
||||||
|
|
||||||
@@ -274,11 +359,12 @@ In OAI the shared memory prefix must be configured in the configuration file.
|
|||||||
```
|
```
|
||||||
|
|
||||||
|
|
||||||
### Stopping the setup
|
#### Stopping the setup
|
||||||
|
|
||||||
|
Run the following command to stop and remove both containers, leaving the
|
||||||
|
system ready to be restarted later:
|
||||||
|
|
||||||
Run the following command to stop and remove both containers, leaving the system ready to be restarted later:
|
|
||||||
```bash
|
```bash
|
||||||
cd ci-scripts/yaml_files/sa_gnb_aerial/
|
cd ci-scripts/yaml_files/sa_gnb_aerial/
|
||||||
docker compose down
|
docker compose down
|
||||||
|
|
||||||
```
|
```
|
||||||
|
|||||||
@@ -3,7 +3,6 @@ Active_gNBs = ( "gNB-OAI");
|
|||||||
Asn1_verbosity = "none";
|
Asn1_verbosity = "none";
|
||||||
|
|
||||||
nfapi = "AERIAL";
|
nfapi = "AERIAL";
|
||||||
pdcp-drop = 1;
|
|
||||||
|
|
||||||
gNBs =
|
gNBs =
|
||||||
(
|
(
|
||||||
@@ -14,49 +13,33 @@ gNBs =
|
|||||||
|
|
||||||
// Tracking area code, 0x0000 and 0xfffe are reserved values
|
// Tracking area code, 0x0000 and 0xfffe are reserved values
|
||||||
tracking_area_code = 1;
|
tracking_area_code = 1;
|
||||||
plmn_list = ({
|
plmn_list = ({ mcc = 208; mnc = 99; mnc_length = 2; snssaiList = ({ sst = 1; }) });
|
||||||
mcc = 001;
|
|
||||||
mnc = 01;
|
|
||||||
mnc_length = 2;
|
|
||||||
snssaiList = (
|
|
||||||
{
|
|
||||||
sst = 1;
|
|
||||||
}
|
|
||||||
);
|
|
||||||
});
|
|
||||||
|
|
||||||
nr_cellid = 51L;
|
nr_cellid = 12345678L;
|
||||||
|
|
||||||
////////// Physical parameters:
|
////////// Physical parameters:
|
||||||
|
|
||||||
pdsch_AntennaPorts_XP = 2;
|
pdsch_AntennaPorts_XP = 2;
|
||||||
#pdsch_AntennaPorts_N1 = 2; #Uncomment for 4 layers
|
pdsch_AntennaPorts_N1 = 2;
|
||||||
pusch_AntennaPorts = 2;
|
pusch_AntennaPorts = 2;
|
||||||
do_CSIRS = 1;
|
do_CSIRS = 1;
|
||||||
do_SRS = 0;
|
do_SRS = 1;
|
||||||
min_rxtxtime = 3;
|
min_rxtxtime = 2;
|
||||||
|
|
||||||
servingCellConfigCommon = (
|
servingCellConfigCommon = (
|
||||||
{
|
{
|
||||||
#spCellConfigCommon
|
#spCellConfigCommon
|
||||||
|
|
||||||
physCellId = 51;
|
physCellId = 0;
|
||||||
|
|
||||||
# downlinkConfigCommon
|
# downlinkConfigCommon
|
||||||
#frequencyInfoDL
|
#frequencyInfoDL
|
||||||
# this is 3300.24 + 134*12*30e3 = 3348.48 MHz (5G NR GSCN: 7741)
|
# center frequency = 3350.01 MHz
|
||||||
absoluteFrequencySSB = 649920;
|
# selected SSB frequency = 3349.92 MHz
|
||||||
#652860; #should be 649920
|
absoluteFrequencySSB = 653376;
|
||||||
#647520;
|
dl_frequencyBand = 77;
|
||||||
#647540;
|
# frequency point A = 3300.87 MHz
|
||||||
#649920;
|
dl_absoluteFrequencyPointA = 650056;
|
||||||
#650208;
|
|
||||||
#647540;
|
|
||||||
dl_frequencyBand = 78;
|
|
||||||
# this is 3300.24 MHz
|
|
||||||
dl_absoluteFrequencyPointA = 646724;
|
|
||||||
#649104;
|
|
||||||
#646724;
|
|
||||||
#scs-SpecificCarrierList
|
#scs-SpecificCarrierList
|
||||||
dl_offstToCarrier = 0;
|
dl_offstToCarrier = 0;
|
||||||
# subcarrierSpacing
|
# subcarrierSpacing
|
||||||
@@ -77,7 +60,7 @@ gNBs =
|
|||||||
|
|
||||||
#uplinkConfigCommon
|
#uplinkConfigCommon
|
||||||
#frequencyInfoUL
|
#frequencyInfoUL
|
||||||
ul_frequencyBand = 78;
|
ul_frequencyBand = 77;
|
||||||
#scs-SpecificCarrierList
|
#scs-SpecificCarrierList
|
||||||
ul_offstToCarrier = 0;
|
ul_offstToCarrier = 0;
|
||||||
# subcarrierSpacing
|
# subcarrierSpacing
|
||||||
@@ -157,13 +140,13 @@ gNBs =
|
|||||||
# pattern1
|
# pattern1
|
||||||
# dl_UL_TransmissionPeriodicity
|
# dl_UL_TransmissionPeriodicity
|
||||||
# 0=ms0p5, 1=ms0p625, 2=ms1, 3=ms1p25, 4=ms2, 5=ms2p5, 6=ms5, 7=ms10
|
# 0=ms0p5, 1=ms0p625, 2=ms1, 3=ms1p25, 4=ms2, 5=ms2p5, 6=ms5, 7=ms10
|
||||||
dl_UL_TransmissionPeriodicity = 6;
|
dl_UL_TransmissionPeriodicity = 6; #5;
|
||||||
nrofDownlinkSlots = 6;
|
nrofDownlinkSlots = 6; #3;
|
||||||
nrofDownlinkSymbols = 10;
|
nrofDownlinkSymbols = 10; #6;
|
||||||
nrofUplinkSlots = 3;
|
nrofUplinkSlots = 3; #1;
|
||||||
nrofUplinkSymbols = 0;
|
nrofUplinkSymbols = 0;
|
||||||
|
|
||||||
ssPBCH_BlockPower = -34;
|
ssPBCH_BlockPower = -15;
|
||||||
}
|
}
|
||||||
|
|
||||||
);
|
);
|
||||||
@@ -179,12 +162,13 @@ gNBs =
|
|||||||
|
|
||||||
|
|
||||||
////////// AMF parameters:
|
////////// AMF parameters:
|
||||||
amf_ip_address = ({ ipv4 = "192.168.70.132"; });
|
amf_ip_address = ({ ipv4 = "172.21.6.5"; });
|
||||||
|
|
||||||
|
|
||||||
NETWORK_INTERFACES :
|
NETWORK_INTERFACES :
|
||||||
{
|
{
|
||||||
GNB_IPV4_ADDRESS_FOR_NG_AMF = "192.168.70.129/24";
|
GNB_IPV4_ADDRESS_FOR_NG_AMF = "172.21.8.120";
|
||||||
GNB_IPV4_ADDRESS_FOR_NGU = "192.168.70.129/24";
|
GNB_IPV4_ADDRESS_FOR_NGU = "172.21.8.120";
|
||||||
GNB_PORT_FOR_S1U = 2152; # Spec 2152
|
GNB_PORT_FOR_S1U = 2152; # Spec 2152
|
||||||
};
|
};
|
||||||
|
|
||||||
@@ -193,25 +177,19 @@ gNBs =
|
|||||||
|
|
||||||
MACRLCs = (
|
MACRLCs = (
|
||||||
{
|
{
|
||||||
num_cc = 1;
|
num_cc = 1;
|
||||||
|
remote_s_address = "127.0.0.1"; // pnf addr [!]
|
||||||
tr_s_preference = "aerial";
|
local_s_address = "127.0.0.2"; // vnf addr
|
||||||
tr_s_shm_prefix = "nvipc";
|
local_s_portc = 50001; // vnf p5 port
|
||||||
tr_s_poll_core = 10;
|
remote_s_portc = 50000; // pnf p5 port [!]
|
||||||
remote_s_address = "127.0.0.1"; // pnf addr [!]
|
local_s_portd = 50011; // vnf p7 port [!]
|
||||||
local_s_address = "127.0.0.2"; // vnf addr
|
remote_s_portd = 50010; // pnf p7 port [!]
|
||||||
local_s_portc = 50001; // vnf p5 port
|
tr_s_preference = "aerial";
|
||||||
remote_s_portc = 50000; // pnf p5 port [!]
|
tr_n_preference = "local_RRC";
|
||||||
local_s_portd = 50011; // vnf p7 port [!]
|
pusch_TargetSNRx10 = 280; # 150;
|
||||||
remote_s_portd = 50010; // pnf p7 port [!]
|
pucch_TargetSNRx10 = 200; #200;
|
||||||
|
dl_max_mcs = 28;
|
||||||
tr_n_preference = "local_RRC";
|
ul_max_mcs = 28;
|
||||||
pusch_FailureThres = 10;
|
|
||||||
pusch_TargetSNRx10 = 280; # 150;
|
|
||||||
pucch_TargetSNRx10 = 100; #200;
|
|
||||||
ulsch_max_frame_inactivity = 10;
|
|
||||||
dl_max_mcs = 25;
|
|
||||||
ul_max_mcs = 25;
|
|
||||||
}
|
}
|
||||||
);
|
);
|
||||||
|
|
||||||
@@ -235,13 +213,13 @@ security = {
|
|||||||
|
|
||||||
log_config :
|
log_config :
|
||||||
{
|
{
|
||||||
global_log_level ="info";
|
global_log_level ="info";
|
||||||
hw_log_level ="info";
|
hw_log_level ="info";
|
||||||
phy_log_level ="info";
|
phy_log_level ="info";
|
||||||
mac_log_level ="info";
|
mac_log_level ="info";
|
||||||
rlc_log_level ="info";
|
rlc_log_level ="info";
|
||||||
pdcp_log_level ="info";
|
pdcp_log_level ="info";
|
||||||
rrc_log_level ="info";
|
rrc_log_level ="info";
|
||||||
ngap_log_level ="debug";
|
ngap_log_level ="debug";
|
||||||
f1ap_log_level ="debug";
|
f1ap_log_level ="debug";
|
||||||
};
|
};
|
||||||
|
|||||||
Reference in New Issue
Block a user