diff --git a/executables/nr-ue.c b/executables/nr-ue.c index d744b6e780..1ad0ddd50a 100644 --- a/executables/nr-ue.c +++ b/executables/nr-ue.c @@ -705,8 +705,7 @@ void *UE_thread(void *arg) { UE->rx_offset_diff = computeSamplesShift(UE); readBlockSize=get_readBlockSize(slot_nr, &UE->frame_parms) - UE->rx_offset_diff; - writeBlockSize=UE->frame_parms.get_samples_per_slot((slot_nr + DURATION_RX_TO_TX - RX_NB_TH) % nb_slot_frame, &UE->frame_parms)- - UE->rx_offset_diff; + writeBlockSize=UE->frame_parms.get_samples_per_slot((slot_nr + DURATION_RX_TO_TX - RX_NB_TH) % nb_slot_frame, &UE->frame_parms)- UE->rx_offset_diff; } AssertFatal(readBlockSize == @@ -765,14 +764,28 @@ void *UE_thread(void *arg) { timing_advance = UE->timing_advance; } - AssertFatal( writeBlockSize == - UE->rfdevice.trx_write_func(&UE->rfdevice, - writeTimestamp, - txp, - writeBlockSize, - UE->frame_parms.nb_antennas_tx, - 1),""); + int flags = 0; + int slot_tx_usrp = slot_nr + DURATION_RX_TO_TX - RX_NB_TH; + uint8_t tdd_period = mac->phy_config.config_req.tdd_table.tdd_period_in_slots; + uint8_t num_UL_slots = mac->scc->tdd_UL_DL_ConfigurationCommon->pattern1.nrofUplinkSlots + + (mac->scc->tdd_UL_DL_ConfigurationCommon->pattern1.nrofUplinkSymbols!=0); + uint8_t first_tx_slot = tdd_period - num_UL_slots; + if (slot_tx_usrp%tdd_period==first_tx_slot) + flags=2; + else if (slot_tx_usrp%tdd_period==first_tx_slot+num_UL_slots-1) + flags = 3; + else if (slot_tx_usrp%tdd_period>first_tx_slot) + flags = 1; + if (flags || IS_SOFTMODEM_RFSIM) + AssertFatal( writeBlockSize == + UE->rfdevice.trx_write_func(&UE->rfdevice, + writeTimestamp, + txp, + writeBlockSize, + UE->frame_parms.nb_antennas_tx, + flags),""); + for (int i=0; iframe_parms.nb_antennas_tx; i++) memset(txp[i], 0, writeBlockSize); @@ -780,7 +793,7 @@ void *UE_thread(void *arg) { msgToPush->key=slot_nr; pushTpool(&(get_nrUE_params()->Tpool), msgToPush); - if ( IS_SOFTMODEM_RFSIM || IS_SOFTMODEM_NOS1) { //getenv("RFSIMULATOR") + if (IS_SOFTMODEM_RFSIM) { //getenv("RFSIMULATOR") // FixMe: Wait previous thread is done, because race conditions seems too bad // in case of actual RF board, the overlap between threads mitigate the issue // We must receive one message, that proves the slot processing is done