maki/openairinterface5g
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Use nr_generate_Msg4() in the case of Msg.3 with C-RNTI

Browse Source 
After removing RA.crnti, there is no need to use a separate function
nr_generate_Msg3_dcch_dtch_response() to generate Msg.4 (yes, the
function mentions Msg3 but it does Msg3 response == Msg.4).

Also, using nr_generate_Msg4() has the advantage that it waits until we
have the reconfiguration available in RLC to be sent to the UE. Finally,
we can delete the old nr_generate_Msg3_dcch_dtch_response() as we don't
need it anymore.
This commit is contained in:
Robert Schmidt
2023-08-18 12:25:51 +02:00
parent 93f59a3c6e
commit 27f9d6f743
1 changed files with 3 additions and 186 deletions
Download Patch File Download Diff File Expand all files Collapse all files
openair2/LAYER2/NR_MAC_gNB/gNB_scheduler_RA.c
+3 -186
View File
@@ -1615,188 +1615,6 @@ static void prepare_dl_pdus(gNB_MAC_INST *nr_mac,
LOG_D(NR_MAC,"numDlDci: %i\n", pdcch_pdu_rel15->numDlDci);
}
static void nr_generate_Msg3_dcch_dtch_response(module_id_t module_idP,
int CC_id,
frame_t frameP,
sub_frame_t slotP,
NR_RA_t *ra,
nfapi_nr_dl_tti_request_t *DL_req,
nfapi_nr_tx_data_request_t *TX_req)
{
gNB_MAC_INST *nr_mac = RC.nrmac[module_idP];
// proceed only if it is a DL slot
if (!is_xlsch_in_slot(nr_mac->dlsch_slot_bitmap[slotP / 64], slotP))
return;
// UE is known by the network, C-RNTI to be used instead of TC-RNTI
int rnti = ra->rnti;
NR_UE_info_t *UE = find_nr_UE(&nr_mac->UE_info, rnti);
if (!UE) {
LOG_E(NR_MAC, "Received Msg3 with C-RNTI but rnti %04x not in the table\n", rnti);
return;
}
NR_COMMON_channels_t *cc = &nr_mac->common_channels[CC_id];
NR_ServingCellConfigCommon_t *scc = cc->ServingCellConfigCommon;
NR_SearchSpace_t *ss = ra->ra_ss;
NR_ControlResourceSet_t *coreset = ra->coreset;
AssertFatal(coreset!=NULL,"Coreset cannot be null for RA-Msg4\n");
// Only need to schedule DCI with and empty DL
NR_UE_DL_BWP_t *dl_bwp = &ra->DL_BWP;
long BWPStart = 0;
long BWPSize = 0;
NR_Type0_PDCCH_CSS_config_t *type0_PDCCH_CSS_config = NULL;
if(*ss->controlResourceSetId!=0) {
BWPStart = dl_bwp->BWPStart;
BWPSize = dl_bwp->BWPSize;
} else {
type0_PDCCH_CSS_config = &nr_mac->type0_PDCCH_CSS_config[ra->beam_id];
BWPStart = type0_PDCCH_CSS_config->cset_start_rb;
BWPSize = type0_PDCCH_CSS_config->num_rbs;
}
// get CCEindex, needed also for PUCCH and then later for PDCCH
uint8_t aggregation_level;
int CCEIndex = get_cce_index(nr_mac,
CC_id, slotP, 0,
&aggregation_level,
ss,
coreset,
&ra->sched_pdcch,
true);
if (CCEIndex < 0) {
LOG_E(NR_MAC, "%s(): cannot find free CCE for RA RNTI 0x%04x!\n", __func__, rnti);
return;
}
// Checking if the DCI allocation is feasible in current subframe
nfapi_nr_dl_tti_request_body_t *dl_req = &DL_req->dl_tti_request_body;
if (dl_req->nPDUs > NFAPI_NR_MAX_DL_TTI_PDUS - 2) {
LOG_I(NR_MAC, "[RAPROC] Subframe %d: FAPI DL structure is full, skip scheduling UE %d\n", slotP, rnti);
return;
}
uint8_t time_domain_assignment = get_dl_tda(nr_mac, scc, slotP);
int mux_pattern = type0_PDCCH_CSS_config ? type0_PDCCH_CSS_config->type0_pdcch_ss_mux_pattern : 1;
NR_tda_info_t tda = get_dl_tda_info(dl_bwp, ss->searchSpaceType->present, time_domain_assignment,
scc->dmrs_TypeA_Position, mux_pattern, NR_RNTI_C, coreset->controlResourceSetId, false);
NR_pdsch_dmrs_t dmrs_info = get_dl_dmrs_params(scc,
dl_bwp,
&tda,
1);
int subheader_len = sizeof(NR_MAC_SUBHEADER_SHORT);
int pdu_length = subheader_len + 7; // 7 is contetion resolution length
int mcsTableIdx = dl_bwp->mcsTableIdx;
int mcsIndex = 0;
int rbStart = 0;
int rbSize = 0;
int tb_scaling = 0;
uint32_t tb_size = 0;
// increase PRBs until we get to BWPSize or TBS is bigger than MAC PDU size
do {
if(rbSize < BWPSize)
rbSize++;
else
mcsIndex++;
LOG_D(NR_MAC,"Calling nr_compute_tbs with N_PRB_DMRS %d, N_DMRS_SLOT %d\n",dmrs_info.N_PRB_DMRS,dmrs_info.N_DMRS_SLOT);
tb_size = nr_compute_tbs(nr_get_Qm_dl(mcsIndex, mcsTableIdx),
nr_get_code_rate_dl(mcsIndex, mcsTableIdx),
rbSize, tda.nrOfSymbols, dmrs_info.N_PRB_DMRS * dmrs_info.N_DMRS_SLOT, 0, tb_scaling,1) >> 3;
} while (tb_size < pdu_length && mcsIndex<=28);
AssertFatal(tb_size >= pdu_length,"Cannot allocate response to MSG3 with DCCH\n");
int i = 0;
uint16_t *vrb_map = cc[CC_id].vrb_map;
while ((i < rbSize) && (rbStart + rbSize <= BWPSize)) {
if (vrb_map[BWPStart + rbStart + i]&SL_to_bitmap(tda.startSymbolIndex, tda.nrOfSymbols)) {
rbStart += i+1;
i = 0;
} else {
i++;
}
}
if (rbStart > (BWPSize - rbSize)) {
LOG_E(NR_MAC, "%s(): cannot find free vrb_map for RNTI %04x!\n", __func__, rnti);
return;
}
// If the UE used MSG3 to transfer a DCCH or DTCH message, then contention resolution is successful if the UE receives a PDCCH transmission which has its CRC bits scrambled by the C-RNTI
// Just send padding LCID
uint8_t buf[tb_size];
NR_MAC_SUBHEADER_FIXED *padding = (NR_MAC_SUBHEADER_FIXED *) &buf[0];
padding->R = 0;
padding->LCID = DL_SCH_LCID_PADDING;
for(int k = 1; k < tb_size; k++)
buf[k] = 0;
T(T_GNB_MAC_DL_PDU_WITH_DATA, T_INT(module_idP), T_INT(CC_id), T_INT(ra->rnti),
T_INT(frameP), T_INT(slotP), T_INT(0), T_BUFFER(buf, tb_size));
// SCF222: PDU index incremented for each PDSCH PDU sent in TX control message. This is used to associate control
// information to data and is reset every slot.
const int pduindex = nr_mac->pdu_index[CC_id]++;
prepare_dl_pdus(nr_mac, ra, dl_bwp, dl_req, NULL, dmrs_info, tda, aggregation_level, CCEIndex, tb_size, 0, 0, 0,
0, time_domain_assignment, CC_id, rnti, 0, mcsIndex, tb_scaling, pduindex, rbStart, rbSize);
// DL TX request
nfapi_nr_pdu_t *tx_req = &TX_req->pdu_list[TX_req->Number_of_PDUs];
memcpy(tx_req->TLVs[0].value.direct, buf, sizeof(uint8_t) * tb_size);
tx_req->PDU_length = tb_size;
tx_req->PDU_index = pduindex;
tx_req->num_TLV = 1;
tx_req->TLVs[0].length = tb_size + 2;
TX_req->SFN = frameP;
TX_req->Number_of_PDUs++;
TX_req->Slot = slotP;
// Mark the corresponding symbols and RBs as used
fill_pdcch_vrb_map(nr_mac,
CC_id,
&ra->sched_pdcch,
CCEIndex,
aggregation_level);
for (int rb = 0; rb < rbSize; rb++)
vrb_map[BWPStart + rb + rbStart] |= SL_to_bitmap(tda.startSymbolIndex, tda.nrOfSymbols);
// If the UE used MSG3 to transfer a DCCH or DTCH message, then contention resolution is successful upon transmission of PDCCH
LOG_A(NR_MAC, "(ue rnti 0x%04x) CBRA procedure succeeded!\n", ra->rnti);
nr_clear_ra_proc(module_idP, CC_id, frameP, ra);
UE->Msg4_ACKed = true;
UE->ra_timer = 0;
if(!UE->CellGroup) {
// In the specific scenario where UE correctly received MSG4 (assuming it decoded RRCsetup with CellGroup) and gNB didn't correctly received an ACK,
// the UE would already have CG but the UE-dedicated gNB structure wouldn't (because RA didn't complete on gNB side).
uper_decode(NULL,
&asn_DEF_NR_CellGroupConfig, //might be added prefix later
(void **)&UE->CellGroup,
(uint8_t *)UE->cg_buf,
(UE->enc_rval.encoded+7)/8, 0, 0);
}
process_CellGroup(UE->CellGroup, UE);
// switching to initial BWP
NR_UE_sched_ctrl_t *sched_ctrl = &UE->UE_sched_ctrl;
configure_UE_BWP(nr_mac, scc, sched_ctrl, NULL, UE, 0, 0);
// Reset uplink failure flags/counters/timers at MAC so gNB will resume again scheduling resources for this UE
nr_mac_reset_ul_failure(sched_ctrl);
}
static void nr_generate_Msg4(module_id_t module_idP,
int CC_id,
frame_t frameP,
@@ -1810,8 +1628,7 @@ static void nr_generate_Msg4(module_id_t module_idP,
NR_UE_DL_BWP_t *dl_bwp = &ra->DL_BWP;
// if it is a DL slot, if the RA is in MSG4 state
if (is_xlsch_in_slot(nr_mac->dlsch_slot_bitmap[slotP / 64], slotP) &&
ra->state == Msg4) {
if (is_xlsch_in_slot(nr_mac->dlsch_slot_bitmap[slotP / 64], slotP)) {
NR_ServingCellConfigCommon_t *scc = cc->ServingCellConfigCommon;
NR_SearchSpace_t *ss = ra->ra_ss;
@@ -1838,7 +1655,7 @@ static void nr_generate_Msg4(module_id_t module_idP,
if (srb_status.bytes_in_buffer == 0) {
lcid = DL_SCH_LCID_DCCH;
// Check for data on SRB1 (RRCReestablishment)
// Check for data on SRB1 (RRCReestablishment, RRCReconfiguration)
srb_status = mac_rlc_status_ind(module_idP, ra->rnti, module_idP, frameP, slotP, ENB_FLAG_YES, MBMS_FLAG_NO, lcid, 0, 0);
}
@@ -2266,7 +2083,7 @@ void nr_schedule_RA(module_id_t module_idP,
nr_generate_Msg3_retransmission(module_idP, CC_id, frameP, slotP, ra, ul_dci_req);
break;
case Msg3_dcch_dtch:
nr_generate_Msg3_dcch_dtch_response(module_idP, CC_id, frameP, slotP, ra, DL_req, TX_req);
/* fallthrough */
case Msg4:
nr_generate_Msg4(module_idP, CC_id, frameP, slotP, ra, DL_req, TX_req);
break;