better handling of PRACH TX power to improve separation between PHY and MAC
This commit is contained in:
@@ -1085,6 +1085,14 @@ int handle_dci(module_id_t module_id, int cc_id, unsigned int gNB_index, frame_t
|
||||
|
||||
}
|
||||
|
||||
void handle_ssb_meas(NR_UE_MAC_INST_t *mac, uint8_t ssb_index, int16_t rsrp_dbm, int16_t pathloss) {
|
||||
|
||||
mac->phy_measurements.ssb_index = ssb_index;
|
||||
mac->phy_measurements.ssb_rsrp_dBm = rsrp_dbm;
|
||||
mac->phy_measurements.pathloss = pathloss;
|
||||
|
||||
}
|
||||
|
||||
// L2 Abstraction Layer
|
||||
// Note: sdu should always be processed because data and timing advance updates are transmitted by the UE
|
||||
int8_t handle_dlsch(nr_downlink_indication_t *dl_info, NR_UL_TIME_ALIGNMENT_t *ul_time_alignment, int pdu_id){
|
||||
@@ -1225,7 +1233,10 @@ int nr_ue_dl_indication(nr_downlink_indication_t *dl_info, NR_UL_TIME_ALIGNMENT_
|
||||
|
||||
switch(dl_info->rx_ind->rx_indication_body[i].pdu_type){
|
||||
case FAPI_NR_RX_PDU_TYPE_SSB:
|
||||
mac->ssb_rsrp_dBm = (dl_info->rx_ind->rx_indication_body+i)->ssb_pdu.rsrp_dBm;
|
||||
handle_ssb_meas(mac,
|
||||
(dl_info->rx_ind->rx_indication_body+i)->ssb_pdu.ssb_index,
|
||||
(dl_info->rx_ind->rx_indication_body+i)->ssb_pdu.rsrp_dBm,
|
||||
(dl_info->rx_ind->rx_indication_body+i)->ssb_pdu.pathloss);
|
||||
ret_mask |= (handle_bcch_bch(dl_info->module_id, dl_info->cc_id, dl_info->gNB_index, dl_info->phy_data,
|
||||
(dl_info->rx_ind->rx_indication_body+i)->ssb_pdu.pdu,
|
||||
(dl_info->rx_ind->rx_indication_body+i)->ssb_pdu.additional_bits,
|
||||
|
||||
Reference in New Issue
Block a user