diff --git a/ci-scripts/conf_files/gnb.band78.tm1.106PRB.usrpn300.conf b/ci-scripts/conf_files/gnb.band78.tm1.106PRB.usrpn300.conf
index f125dcd3b2..43e4c6b493 100644
--- a/ci-scripts/conf_files/gnb.band78.tm1.106PRB.usrpn300.conf
+++ b/ci-scripts/conf_files/gnb.band78.tm1.106PRB.usrpn300.conf
@@ -231,7 +231,7 @@ RUs = (
          att_rx         = 0;
          bands          = [7];
          max_pdschReferenceSignalPower = -27;
-         max_rxgain                    = 75;
+         max_rxgain                    = 50;
          eNB_instances  = [0];
          ## beamforming 1x2 matrix: 1 layer x 2 antennas
          bf_weights = [0x00007fff, 0x0000];
diff --git a/executables/nr-gnb.c b/executables/nr-gnb.c
index 4047ff1386..8b3d6fc64e 100644
--- a/executables/nr-gnb.c
+++ b/executables/nr-gnb.c
@@ -355,7 +355,9 @@ void init_gNB_Tpool(int inst) {
   // ULSCH decoding threadpool
   gNB->threadPool = (tpool_t*)malloc(sizeof(tpool_t));
   int numCPU = sysconf(_SC_NPROCESSORS_ONLN);
+  LOG_I(PHY,"Number of threads requested in config file: %d, Number of threads available on this machine: %d\n",gNB->pusch_proc_threads,numCPU);
   int threadCnt = min(numCPU, gNB->pusch_proc_threads);
+  if (threadCnt < 2) LOG_E(PHY,"Number of threads for gNB should be more than 1. Allocated only %d\n",threadCnt);
   char ul_pool[80];
   sprintf(ul_pool,"-1");
   int s_offset = 0;
diff --git a/executables/nr-ue.c b/executables/nr-ue.c
index 2adf373196..137fcdfec3 100644
--- a/executables/nr-ue.c
+++ b/executables/nr-ue.c
@@ -27,6 +27,8 @@
 #include "SCHED_NR_UE/phy_frame_config_nr.h"
 #include "SCHED_NR_UE/defs.h"
 #include "PHY/NR_UE_TRANSPORT/nr_transport_proto_ue.h"
+#include "executables/softmodem-common.h"
+#include "SCHED_NR_UE/pucch_uci_ue_nr.h"
 
 /*
  *  NR SLOT PROCESSING SEQUENCE
@@ -85,6 +87,14 @@
  *
  */
 
+#ifndef NO_RAT_NR
+  #define DURATION_RX_TO_TX           (NR_UE_CAPABILITY_SLOT_RX_TO_TX)  /* for NR this will certainly depends to such UE capability which is not yet defined */
+#else
+  #define DURATION_RX_TO_TX           (6)   /* For LTE, this duration is fixed to 4 and it is linked to LTE standard for both modes FDD/TDD */
+#endif
+#define RX_JOB_ID 0x1010
+#define TX_JOB_ID 100
+
 typedef enum {
   pss = 0,
   pbch = 1,
@@ -101,6 +111,7 @@ void init_nr_ue_vars(PHY_VARS_NR_UE *ue,
   ue->Mod_id      = UE_id;
   ue->mac_enabled = 1;
   ue->if_inst     = nr_ue_if_module_init(0);
+  ue->dci_thres   = 0;
 
   // Setting UE mode to NOT_SYNCHED by default
   for (gNB_id = 0; gNB_id < nb_connected_gNB; gNB_id++){
@@ -123,10 +134,7 @@ void init_nr_ue_vars(PHY_VARS_NR_UE *ue,
  * \param arg is a pointer to a \ref PHY_VARS_NR_UE structure.
  */
 
-typedef struct syncData_s {
-  UE_nr_rxtx_proc_t proc;
-  PHY_VARS_NR_UE *UE;
-} syncData_t;
+typedef nr_rxtx_thread_data_t syncData_t;
 
 static void UE_synch(void *arg) {
   syncData_t *syncD=(syncData_t *) arg;
@@ -273,7 +281,11 @@ static void UE_synch(void *arg) {
   }
 }
 
-void processSlotTX( PHY_VARS_NR_UE *UE, UE_nr_rxtx_proc_t *proc) {
+void processSlotTX(void *arg) {
+
+  nr_rxtx_thread_data_t *rxtxD = (nr_rxtx_thread_data_t *) arg;
+  UE_nr_rxtx_proc_t *proc = &rxtxD->proc;
+  PHY_VARS_NR_UE    *UE   = rxtxD->UE;
   fapi_nr_config_request_t *cfg = &UE->nrUE_config;
   int tx_slot_type = nr_ue_slot_select(cfg, proc->frame_tx, proc->nr_slot_tx);
   uint8_t gNB_id = 0;
@@ -304,10 +316,14 @@ void processSlotTX( PHY_VARS_NR_UE *UE, UE_nr_rxtx_proc_t *proc) {
   }
 }
 
-void processSlotRX( PHY_VARS_NR_UE *UE, UE_nr_rxtx_proc_t *proc) {
+void processSlotRX(void *arg) {
 
+  nr_rxtx_thread_data_t *rxtxD = (nr_rxtx_thread_data_t *) arg;
+  UE_nr_rxtx_proc_t *proc = &rxtxD->proc;
+  PHY_VARS_NR_UE    *UE   = rxtxD->UE;
   fapi_nr_config_request_t *cfg = &UE->nrUE_config;
   int rx_slot_type = nr_ue_slot_select(cfg, proc->frame_rx, proc->nr_slot_rx);
+  int tx_slot_type = nr_ue_slot_select(cfg, proc->frame_tx, proc->nr_slot_tx);
   uint8_t gNB_id = 0;
 
   if (rx_slot_type == NR_DOWNLINK_SLOT || rx_slot_type == NR_MIXED_SLOT){
@@ -323,7 +339,7 @@ void processSlotRX( PHY_VARS_NR_UE *UE, UE_nr_rxtx_proc_t *proc) {
     phy_procedures_slot_parallelization_nrUE_RX( UE, proc, 0, 0, 1, no_relay, NULL );
 #else
     uint64_t a=rdtsc();
-    phy_procedures_nrUE_RX(UE, proc, gNB_id, get_nrUE_params()->nr_dlsch_parallel);
+    phy_procedures_nrUE_RX(UE, proc, gNB_id, get_nrUE_params()->nr_dlsch_parallel, &rxtxD->txFifo);
     LOG_D(PHY, "In %s: slot %d, time %lu\n", __FUNCTION__, proc->nr_slot_rx, (rdtsc()-a)/3500);
 #endif
 
@@ -341,34 +357,38 @@ void processSlotRX( PHY_VARS_NR_UE *UE, UE_nr_rxtx_proc_t *proc) {
         nr_pdcp_tick(proc->frame_rx, proc->nr_slot_rx / UE->frame_parms.slots_per_subframe);
       }
     }
+
+    // Wait for PUSCH processing to finish
+    notifiedFIFO_elt_t *res;
+    res = pullTpool(&rxtxD->txFifo,&(get_nrUE_params()->Tpool));
+    delNotifiedFIFO_elt(res);
+
+  } else {
+    processSlotTX(rxtxD);
   }
 
-}
+  if (tx_slot_type == NR_UPLINK_SLOT || tx_slot_type == NR_MIXED_SLOT){
+    if (UE->UE_mode[gNB_id] <= PUSCH) {
+      if (get_softmodem_params()->usim_test==0) {
+        pucch_procedures_ue_nr(UE,
+                               gNB_id,
+                               proc,
+                               FALSE);
+      }
 
-/*!
- * \brief This is the UE thread for RX subframe n and TX subframe n+4.
- * This thread performs the phy_procedures_UE_RX() on every received slot.
- * then, if TX is enabled it performs TX for n+4.
- * \param arg is a pointer to a \ref PHY_VARS_NR_UE structure.
- * \returns a pointer to an int. The storage is not on the heap and must not be freed.
- */
+      LOG_D(PHY, "Sending Uplink data \n");
+      nr_ue_pusch_common_procedures(UE,
+                                    proc->nr_slot_tx,
+                                    &UE->frame_parms,1);
+    }
 
-typedef struct processingData_s {
-  UE_nr_rxtx_proc_t proc;
-  PHY_VARS_NR_UE    *UE;
-}  processingData_t;
-
-void UE_processing(void *arg) {
-  processingData_t *rxtxD = (processingData_t *) arg;
-  UE_nr_rxtx_proc_t *proc = &rxtxD->proc;
-  PHY_VARS_NR_UE    *UE   = rxtxD->UE;
-  int slot_tx = proc->nr_slot_tx;
-  int frame_tx = proc->frame_tx;
-
-  processSlotRX(UE, proc);
-  processSlotTX(UE, proc);
-  ue_ta_procedures(UE, slot_tx, frame_tx);
+    if (UE->UE_mode[gNB_id] > NOT_SYNCHED && UE->UE_mode[gNB_id] < PUSCH) {
+      nr_ue_prach_procedures(UE, proc, gNB_id);
+    }
+    LOG_D(PHY,"****** end TX-Chain for AbsSubframe %d.%d ******\n", proc->frame_tx, proc->nr_slot_tx);
+  }
 
+  ue_ta_procedures(UE, proc->nr_slot_tx, proc->frame_tx);
 }
 
 void dummyWrite(PHY_VARS_NR_UE *UE,openair0_timestamp timestamp, int writeBlockSize) {
@@ -485,25 +505,44 @@ void *UE_thread(void *arg) {
   int start_rx_stream = 0;
   AssertFatal(0== openair0_device_load(&(UE->rfdevice), &openair0_cfg[0]), "");
   UE->rfdevice.host_type = RAU_HOST;
+  UE->lost_sync = 0;
+  UE->is_synchronized = 0;
   AssertFatal(UE->rfdevice.trx_start_func(&UE->rfdevice) == 0, "Could not start the device\n");
+
   notifiedFIFO_t nf;
   initNotifiedFIFO(&nf);
-  int nbSlotProcessing=0;
-  int thread_idx=0;
+
   notifiedFIFO_t freeBlocks;
   initNotifiedFIFO_nothreadSafe(&freeBlocks);
+
+  int nbSlotProcessing=0;
+  int thread_idx=0;
   NR_UE_MAC_INST_t *mac = get_mac_inst(0);
   int timing_advance = UE->timing_advance;
 
-  for (int i=0; i<RX_NB_TH+1; i++)  // RX_NB_TH working + 1 we are making to be pushed
-    pushNotifiedFIFO_nothreadSafe(&freeBlocks,
-                                  newNotifiedFIFO_elt(sizeof(processingData_t), 0,&nf,UE_processing));
-
   bool syncRunning=false;
   const int nb_slot_frame = UE->frame_parms.slots_per_frame;
   int absolute_slot=0, decoded_frame_rx=INT_MAX, trashed_frames=0;
 
+  for (int i=0; i<NR_RX_NB_TH+1; i++) {// NR_RX_NB_TH working + 1 we are making to be pushed
+    notifiedFIFO_elt_t *newElt = newNotifiedFIFO_elt(sizeof(nr_rxtx_thread_data_t), RX_JOB_ID,&nf,processSlotRX);
+    nr_rxtx_thread_data_t *curMsg=(nr_rxtx_thread_data_t *)NotifiedFifoData(newElt);
+    initNotifiedFIFO(&curMsg->txFifo);
+    pushNotifiedFIFO_nothreadSafe(&freeBlocks, newElt);
+  }
+
   while (!oai_exit) {
+    if (UE->lost_sync) {
+      int nb = abortTpool(&(get_nrUE_params()->Tpool),RX_JOB_ID);
+      nb += abortNotifiedFIFO(&nf, RX_JOB_ID);
+      LOG_I(PHY,"Number of aborted slots %d\n",nb);
+      for (int i=0; i<nb; i++)
+        pushNotifiedFIFO_nothreadSafe(&freeBlocks, newNotifiedFIFO_elt(sizeof(nr_rxtx_thread_data_t), RX_JOB_ID,&nf,processSlotRX));
+      nbSlotProcessing = 0;
+      UE->is_synchronized = 0;
+      UE->lost_sync = 0;
+    }
+
     if (syncRunning) {
       notifiedFIFO_elt_t *res=tryPullTpool(&nf,&(get_nrUE_params()->Tpool));
 
@@ -516,6 +555,7 @@ void *UE_thread(void *arg) {
           decoded_frame_rx=(decoded_frame_rx + (!UE->init_sync_frame) + trashed_frames) % MAX_FRAME_NUMBER;
         }
         delNotifiedFIFO_elt(res);
+        start_rx_stream=0;
       } else {
         readFrame(UE, &timestamp, true);
         trashed_frames+=2;
@@ -562,13 +602,13 @@ void *UE_thread(void *arg) {
 
     // whatever means thread_idx
     // Fix me: will be wrong when slot 1 is slow, as slot 2 finishes
-    // Slot 3 will overlap if RX_NB_TH is 2
+    // Slot 3 will overlap if NR_RX_NB_TH is 2
     // this is general failure in UE !!!
-    thread_idx = absolute_slot % RX_NB_TH;
+    thread_idx = absolute_slot % NR_RX_NB_TH;
     int slot_nr = absolute_slot % nb_slot_frame;
     notifiedFIFO_elt_t *msgToPush;
-    AssertFatal((msgToPush=pullNotifiedFIFO_nothreadSafe(&freeBlocks)) != NULL,"chained list failure");
-    processingData_t *curMsg=(processingData_t *)NotifiedFifoData(msgToPush);
+    AssertFatal((msgToPush=pullTpool(&freeBlocks,&(get_nrUE_params()->Tpool))) != NULL,"chained list failure");
+    nr_rxtx_thread_data_t *curMsg=(nr_rxtx_thread_data_t *)NotifiedFifoData(msgToPush);
     curMsg->UE=UE;
     // update thread index for received subframe
     curMsg->proc.thread_id   = thread_idx;
@@ -596,18 +636,18 @@ void *UE_thread(void *arg) {
 
     for (int i=0; i<UE->frame_parms.nb_antennas_tx; i++)
       txp[i] = (void *)&UE->common_vars.txdata[i][UE->frame_parms.get_samples_slot_timestamp(
-               ((slot_nr + DURATION_RX_TO_TX - RX_NB_TH)%nb_slot_frame),&UE->frame_parms,0)];
+               ((slot_nr + DURATION_RX_TO_TX - NR_RX_NB_TH)%nb_slot_frame),&UE->frame_parms,0)];
 
     int readBlockSize, writeBlockSize;
 
     if (slot_nr<(nb_slot_frame - 1)) {
       readBlockSize=get_readBlockSize(slot_nr, &UE->frame_parms);
-      writeBlockSize=UE->frame_parms.get_samples_per_slot((slot_nr + DURATION_RX_TO_TX - RX_NB_TH) % nb_slot_frame, &UE->frame_parms);
+      writeBlockSize=UE->frame_parms.get_samples_per_slot((slot_nr + DURATION_RX_TO_TX - NR_RX_NB_TH) % nb_slot_frame, &UE->frame_parms);
     } else {
       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 - NR_RX_NB_TH) % nb_slot_frame, &UE->frame_parms)- UE->rx_offset_diff;
     }
 
     AssertFatal(readBlockSize ==
@@ -639,14 +679,14 @@ void *UE_thread(void *arg) {
 
     notifiedFIFO_elt_t *res;
 
-    while (nbSlotProcessing >= RX_NB_TH) {
+    while (nbSlotProcessing >= NR_RX_NB_TH) {
       res=pullTpool(&nf, &(get_nrUE_params()->Tpool));
       nbSlotProcessing--;
-      processingData_t *tmp=(processingData_t *)res->msgData;
+      nr_rxtx_thread_data_t *tmp=(nr_rxtx_thread_data_t *)res->msgData;
 
       if (tmp->proc.decoded_frame_rx != -1)
         decoded_frame_rx=(((mac->mib->systemFrameNumber.buf[0] >> mac->mib->systemFrameNumber.bits_unused)<<4) | tmp->proc.decoded_frame_rx);
-      else 
+      else
          decoded_frame_rx=-1;
 
       pushNotifiedFIFO_nothreadSafe(&freeBlocks,res);
@@ -659,7 +699,7 @@ void *UE_thread(void *arg) {
     // use previous timing_advance value to compute writeTimestamp
     writeTimestamp = timestamp+
       UE->frame_parms.get_samples_slot_timestamp(slot_nr,&UE->frame_parms,DURATION_RX_TO_TX
-      - RX_NB_TH) - firstSymSamp - openair0_cfg[0].tx_sample_advance -
+      - NR_RX_NB_TH) - firstSymSamp - openair0_cfg[0].tx_sample_advance -
       UE->N_TA_offset - timing_advance;
 
     // but use current UE->timing_advance value to compute writeBlockSize
@@ -669,7 +709,7 @@ void *UE_thread(void *arg) {
     }
 
     int flags = 0;
-    int slot_tx_usrp = slot_nr + DURATION_RX_TO_TX - RX_NB_TH;
+    int slot_tx_usrp = slot_nr + DURATION_RX_TO_TX - NR_RX_NB_TH;
 
     if (openair0_cfg[0].duplex_mode == duplex_mode_TDD) {
 
@@ -701,26 +741,9 @@ void *UE_thread(void *arg) {
       memset(txp[i], 0, writeBlockSize);
 
     nbSlotProcessing++;
-    msgToPush->key=slot_nr;
+    LOG_D(PHY,"Number of slots being processed at the moment: %d\n",nbSlotProcessing);
     pushTpool(&(get_nrUE_params()->Tpool), msgToPush);
 
-    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
-      res=pullTpool(&nf, &(get_nrUE_params()->Tpool));
-      nbSlotProcessing--;
-      processingData_t *tmp=(processingData_t *)res->msgData;
-
-      if (tmp->proc.decoded_frame_rx != -1)
-        decoded_frame_rx=(((mac->mib->systemFrameNumber.buf[0] >> mac->mib->systemFrameNumber.bits_unused)<<4) | tmp->proc.decoded_frame_rx);
-      else 
-        decoded_frame_rx=-1;
-        //decoded_frame_rx=tmp->proc.decoded_frame_rx;
-
-      pushNotifiedFIFO_nothreadSafe(&freeBlocks,res);
-    }
-
   } // while !oai_exit
 
   return NULL;
diff --git a/executables/nr-uesoftmodem.c b/executables/nr-uesoftmodem.c
index 174d5409bd..13a191a18a 100644
--- a/executables/nr-uesoftmodem.c
+++ b/executables/nr-uesoftmodem.c
@@ -232,9 +232,16 @@ nrUE_params_t *get_nrUE_params(void) {
 }
 /* initialie thread pools used for NRUE processing paralleliation */ 
 void init_tpools(uint8_t nun_dlsch_threads) {
-  char *params=calloc(1,(RX_NB_TH*3)+1);
-  for (int i=0; i<RX_NB_TH; i++) {
-    memcpy(params+(i*3),"-1,",3);
+  char *params = NULL;
+  if (IS_SOFTMODEM_RFSIM) {
+    params = calloc(1,2);
+    memcpy(params,"N",1);
+  }
+  else {
+    params = calloc(1,(NR_RX_NB_TH*NR_NB_TH_SLOT*3)+1);
+    for (int i=0; i<NR_RX_NB_TH*NR_NB_TH_SLOT; i++) {
+      memcpy(params+(i*3),"-1,",3);
+    }
   }
   initTpool(params, &(nrUE_params.Tpool), false);
   free(params);
@@ -258,7 +265,6 @@ static void get_options(void) {
     printf("%s\n",uecap_xer);
     uecap_xer_in=1;
   } /* UE with config file  */
-    init_tpools(nrUE_params.nr_dlsch_parallel);
 }
 
 // set PHY vars from command line
@@ -415,6 +421,7 @@ int main( int argc, char **argv ) {
   get_options (); //Command-line options specific for NRUE
 
   get_common_options(SOFTMODEM_5GUE_BIT );
+  init_tpools(nrUE_params.nr_dlsch_parallel);
   CONFIG_CLEARRTFLAG(CONFIG_NOEXITONHELP);
 #if T_TRACER
   T_Config_Init();
diff --git a/executables/softmodem-common.h b/executables/softmodem-common.h
index e4275a48bb..9fa96462fd 100644
--- a/executables/softmodem-common.h
+++ b/executables/softmodem-common.h
@@ -73,6 +73,7 @@ extern "C"
 #define CONFIG_HLP_STMON         "Enable processing timing measurement of lte softmodem on per subframe basis \n"
 #define CONFIG_HLP_256QAM        "Use the 256 QAM mcs table for PDSCH\n"
 
+#define CONFIG_HLP_NONSTOP       "Go back to frame sync mode after 100 consecutive PBCH failures\n"
 //#define CONFIG_HLP_NUMUES        "Set the number of UEs for the emulation"
 #define CONFIG_HLP_MSLOTS        "Skip the missed slots/subframes \n"
 #define CONFIG_HLP_ULMCS         "Set the maximum uplink MCS\n"
@@ -115,6 +116,7 @@ extern "C"
 #define USIM_TEST           softmodem_params.usim_test
 #define USE_256QAM_TABLE    softmodem_params.use_256qam_table
 #define NFAPI               softmodem_params.nfapi
+#define NON_STOP            softmodem_params.non_stop
 
 #define DEFAULT_RFCONFIG_FILE    "/usr/local/etc/syriq/ue.band7.tm1.PRB100.NR40.dat";
 
@@ -148,6 +150,7 @@ extern int usrp_tx_thread;
     {"use-256qam-table",     CONFIG_HLP_256QAM,       PARAMFLAG_BOOL, iptr:&USE_256QAM_TABLE,             defintval:0,           TYPE_INT,    0},                     \
     {"usrp-tx-thread-config", CONFIG_HLP_USRP_THREAD, 0,              iptr:&usrp_tx_thread,               defstrval:0,           TYPE_INT,    0},        \
     {"nfapi",                CONFIG_HLP_NFAPI,        0,              u8ptr:&nfapi_mode,                       defintval:0,           TYPE_UINT8,  0},                     \
+    {"non-stop",            CONFIG_HLP_NONSTOP,      PARAMFLAG_BOOL, iptr:&NON_STOP,                       defintval:0,           TYPE_INT,  0},                     \
   }
 
   
@@ -238,6 +241,7 @@ typedef struct {
   uint32_t       send_dmrs_sync;
   int            use_256qam_table;
   uint8_t        nfapi;
+  int            non_stop;
 } softmodem_params_t;
 
 extern uint64_t get_softmodem_optmask(void);
diff --git a/openair1/PHY/NR_UE_TRANSPORT/dci_nr.c b/openair1/PHY/NR_UE_TRANSPORT/dci_nr.c
index 8176514eb4..c1cebc64a8 100644
--- a/openair1/PHY/NR_UE_TRANSPORT/dci_nr.c
+++ b/openair1/PHY/NR_UE_TRANSPORT/dci_nr.c
@@ -874,6 +874,33 @@ void nr_pdcch_unscrambling(int16_t *z,
 
 
 #ifdef NR_PDCCH_DCI_RUN
+/* This function compares the received DCI bits with
+ * re-encoded DCI bits and returns the number of mismatched bits
+ */
+uint16_t nr_dci_false_detection(uint64_t *dci,
+                            int16_t *soft_in,
+                            const t_nrPolar_params *polar_param,
+                            int encoded_length,
+                            int rnti) {
+
+  uint32_t encoder_output[NR_MAX_DCI_SIZE_DWORD];
+  polar_encoder_fast(dci, (void*)encoder_output, rnti, 1, polar_param);
+  uint8_t *enout_p = (uint8_t*)encoder_output;
+  uint16_t x = 0;
+
+  for (int i=0; i<encoded_length/8; i++) {
+    x += ( enout_p[i] & 1 ) ^ ( ( soft_in[i*8] >> 15 ) & 1);
+    x += ( ( enout_p[i] >> 1 ) & 1 ) ^ ( ( soft_in[i*8+1] >> 15 ) & 1 );
+    x += ( ( enout_p[i] >> 2 ) & 1 ) ^ ( ( soft_in[i*8+2] >> 15 ) & 1 );
+    x += ( ( enout_p[i] >> 3 ) & 1 ) ^ ( ( soft_in[i*8+3] >> 15 ) & 1 );
+    x += ( ( enout_p[i] >> 4 ) & 1 ) ^ ( ( soft_in[i*8+4] >> 15 ) & 1 );
+    x += ( ( enout_p[i] >> 5 ) & 1 ) ^ ( ( soft_in[i*8+5] >> 15 ) & 1 );
+    x += ( ( enout_p[i] >> 6 ) & 1 ) ^ ( ( soft_in[i*8+6] >> 15 ) & 1 );
+    x += ( ( enout_p[i] >> 7 ) & 1 ) ^ ( ( soft_in[i*8+7] >> 15 ) & 1 );
+  }
+  return x;
+}
+
 uint8_t nr_dci_decoding_procedure(PHY_VARS_NR_UE *ue,
                                   UE_nr_rxtx_proc_t *proc,
                                   fapi_nr_dci_indication_t *dci_ind) {
@@ -920,15 +947,23 @@ uint8_t nr_dci_decoding_procedure(PHY_VARS_NR_UE *ue,
         if (crc == n_rnti) {
           LOG_D(PHY, "(%i.%i) Received dci indication (rnti %x,dci format %d,n_CCE %d,payloadSize %d,payload %llx)\n",
                 proc->frame_rx, proc->nr_slot_rx,n_rnti,rel15->dci_format_options[k],CCEind,dci_length,*(unsigned long long*)dci_estimation);
-          dci_ind->SFN = proc->frame_rx;
-          dci_ind->slot = proc->nr_slot_rx;
-          dci_ind->dci_list[dci_ind->number_of_dcis].rnti        = n_rnti;
-          dci_ind->dci_list[dci_ind->number_of_dcis].n_CCE       = CCEind;
-          dci_ind->dci_list[dci_ind->number_of_dcis].dci_format  = rel15->dci_format_options[k];
-          dci_ind->dci_list[dci_ind->number_of_dcis].payloadSize = dci_length;
-          memcpy((void*)dci_ind->dci_list[dci_ind->number_of_dcis].payloadBits,(void*)dci_estimation,8);
-          dci_ind->number_of_dcis++;
-          break;    // If DCI is found, no need to check for remaining DCI lengths
+          uint16_t mb = nr_dci_false_detection(dci_estimation,tmp_e,currentPtrDCI,L*108,n_rnti);
+          ue->dci_thres = (ue->dci_thres + mb) / 2;
+          if (mb > (ue->dci_thres+20)) {
+            LOG_W(PHY,"DCI false positive. Dropping DCI index %d. Mismatched bits: %d/%d. Current DCI threshold: %d\n",j,mb,L*108,ue->dci_thres);
+            continue;
+          }
+          else {
+            dci_ind->SFN = proc->frame_rx;
+            dci_ind->slot = proc->nr_slot_rx;
+            dci_ind->dci_list[dci_ind->number_of_dcis].rnti        = n_rnti;
+            dci_ind->dci_list[dci_ind->number_of_dcis].n_CCE       = CCEind;
+            dci_ind->dci_list[dci_ind->number_of_dcis].dci_format  = rel15->dci_format_options[k];
+            dci_ind->dci_list[dci_ind->number_of_dcis].payloadSize = dci_length;
+            memcpy((void*)dci_ind->dci_list[dci_ind->number_of_dcis].payloadBits,(void*)dci_estimation,8);
+            dci_ind->number_of_dcis++;
+            break;    // If DCI is found, no need to check for remaining DCI lengths
+          }
         } else {
           LOG_D(PHY,"(%i.%i) Decoded crc %x does not match rnti %x for DCI format %d\n", proc->frame_rx, proc->nr_slot_rx, crc, n_rnti, rel15->dci_format_options[k]);
         }
diff --git a/openair1/PHY/defs_nr_UE.h b/openair1/PHY/defs_nr_UE.h
index bf0a7bc9e5..e3c0737838 100644
--- a/openair1/PHY/defs_nr_UE.h
+++ b/openair1/PHY/defs_nr_UE.h
@@ -763,6 +763,8 @@ typedef struct {
   int is_synchronized;
   /// \brief Indicates on which frame is synchronized in a two frame synchronization
   int is_synchronized_on_frame;
+  /// \brief Indicator that UE lost frame synchronization
+  int lost_sync;
   /// Data structure for UE process scheduling
   UE_nr_proc_t proc;
   /// Flag to indicate the UE shouldn't do timing correction at all
@@ -789,6 +791,8 @@ typedef struct {
   uint8_t ho_initiated;
   /// \brief indicator that Handover procedure has been triggered
   uint8_t ho_triggered;
+  /// threshold for false dci detection
+  int dci_thres;
   /// \brief Measurement variables.
   PHY_NR_MEASUREMENTS measurements;
   NR_DL_FRAME_PARMS  frame_parms;
@@ -1080,6 +1084,7 @@ typedef struct {
 typedef struct nr_rxtx_thread_data_s {
   UE_nr_rxtx_proc_t proc;
   PHY_VARS_NR_UE    *UE;
+  notifiedFIFO_t txFifo;
 }  nr_rxtx_thread_data_t;
 
 #include "SIMULATION/ETH_TRANSPORT/defs.h"
diff --git a/openair1/PHY/defs_nr_common.h b/openair1/PHY/defs_nr_common.h
index 398252c37b..d1d69ecabe 100644
--- a/openair1/PHY/defs_nr_common.h
+++ b/openair1/PHY/defs_nr_common.h
@@ -109,6 +109,8 @@
 
 #define MAX_NUM_NR_CHANNEL_BITS (14*273*12*8)  // 14 symbols, 273 RB
 #define MAX_NUM_NR_RE (14*273*12)
+#define NR_RX_NB_TH 1
+#define NR_NB_TH_SLOT 2
 
 extern const uint8_t nr_rv_round_map[4]; 
 extern const uint8_t nr_rv_round_map_ue[4]; 
diff --git a/openair1/SCHED_NR/phy_procedures_nr_gNB.c b/openair1/SCHED_NR/phy_procedures_nr_gNB.c
index 17aa2ec7d0..d68cacece2 100644
--- a/openair1/SCHED_NR/phy_procedures_nr_gNB.c
+++ b/openair1/SCHED_NR/phy_procedures_nr_gNB.c
@@ -564,6 +564,10 @@ void phy_procedures_gNB_uespec_RX(PHY_VARS_gNB *gNB, int frame_rx, int slot_rx)
           gNB->uci_pdu_list[num_ucis].pdu_size = sizeof(nfapi_nr_uci_pucch_pdu_format_0_1_t);
           nfapi_nr_uci_pucch_pdu_format_0_1_t *uci_pdu_format0 = &gNB->uci_pdu_list[num_ucis].pucch_pdu_format_0_1;
 
+          offset = pucch_pdu->start_symbol_index*gNB->frame_parms.ofdm_symbol_size + (gNB->frame_parms.first_carrier_offset+pucch_pdu->prb_start*12);
+          power_rxF = signal_energy_nodc(&gNB->common_vars.rxdataF[0][offset],12);
+          LOG_D(PHY,"frame %d, slot %d: PUCCH signal energy %d\n",frame_rx,slot_rx,power_rxF);
+
           nr_decode_pucch0(gNB,
 	                   slot_rx,
                            uci_pdu_format0,
diff --git a/openair1/SCHED_NR_UE/defs.h b/openair1/SCHED_NR_UE/defs.h
index f7f44a520d..2675e46a04 100644
--- a/openair1/SCHED_NR_UE/defs.h
+++ b/openair1/SCHED_NR_UE/defs.h
@@ -115,14 +115,17 @@ void phy_procedures_nrUE_TX(PHY_VARS_NR_UE *ue, UE_nr_rxtx_proc_t *proc, uint8_t
   @param proc           Pointer to proc information
   @param gNB_id         Local id of eNB on which to act
   @param dlsch_parallel use multithreaded dlsch processing
+  @param txFifo         Result fifo if PDSCH is run in parallel
 */
 int phy_procedures_nrUE_RX(PHY_VARS_NR_UE *ue,
                            UE_nr_rxtx_proc_t *proc,
                            uint8_t gNB_id,
-                           uint8_t dlsch_parallel);
+                           uint8_t dlsch_parallel,
+                           notifiedFIFO_t *txFifo);
 
 int phy_procedures_slot_parallelization_nrUE_RX(PHY_VARS_NR_UE *ue, UE_nr_rxtx_proc_t *proc, uint8_t eNB_id, uint8_t abstraction_flag, uint8_t do_pdcch_flag, relaying_type_t r_type);
 
+void processSlotTX(void *arg);
 
 #ifdef UE_SLOT_PARALLELISATION
   void *UE_thread_slot1_dl_processing(void *arg);
diff --git a/openair1/SCHED_NR_UE/fapi_nr_ue_l1.c b/openair1/SCHED_NR_UE/fapi_nr_ue_l1.c
index 54464446a5..c8ac871bbf 100644
--- a/openair1/SCHED_NR_UE/fapi_nr_ue_l1.c
+++ b/openair1/SCHED_NR_UE/fapi_nr_ue_l1.c
@@ -127,6 +127,8 @@ int8_t nr_ue_scheduled_response(nr_scheduled_response_t *scheduled_response){
               // dlsch0_harq->status not ACTIVE may be due to false retransmission. Reset the 
               // following flag to skip PDSCH procedures in that case.
               dlsch0->active = 0;
+              dlsch0_harq->harq_ack.ack = 1;
+              dlsch0_harq->harq_ack.send_harq_status = 1;
             }
             dlsch0_harq->harq_ack.vDAI_DL = dlsch_config_pdu->dai;
             /* PTRS */
diff --git a/openair1/SCHED_NR_UE/phy_procedures_nr_ue.c b/openair1/SCHED_NR_UE/phy_procedures_nr_ue.c
index 60fe897456..ff7e4280a1 100644
--- a/openair1/SCHED_NR_UE/phy_procedures_nr_ue.c
+++ b/openair1/SCHED_NR_UE/phy_procedures_nr_ue.c
@@ -275,26 +275,8 @@ void phy_procedures_nrUE_TX(PHY_VARS_NR_UE *ue,
         nr_ue_ulsch_procedures(ue, harq_pid, frame_tx, slot_tx, proc->thread_id, gNB_id);
     }
 
-    if (get_softmodem_params()->usim_test==0) {
-      LOG_D(PHY, "Generating PUCCH\n");
-      pucch_procedures_ue_nr(ue,
-                             gNB_id,
-                             proc,
-                             FALSE);
-    }
-
-    LOG_D(PHY, "Sending Uplink data \n");
-    nr_ue_pusch_common_procedures(ue,
-                                  slot_tx,
-                                  &ue->frame_parms,1);
-
   }
 
-  if (ue->UE_mode[gNB_id] > NOT_SYNCHED && ue->UE_mode[gNB_id] < PUSCH) {
-    nr_ue_prach_procedures(ue, proc, gNB_id);
-  }
-  LOG_D(PHY,"****** end TX-Chain for AbsSubframe %d.%d ******\n", frame_tx, slot_tx);
-
   VCD_SIGNAL_DUMPER_DUMP_FUNCTION_BY_NAME(VCD_SIGNAL_DUMPER_FUNCTIONS_PHY_PROCEDURES_UE_TX, VCD_FUNCTION_OUT);
 #if UE_TIMING_TRACE
   stop_meas(&ue->phy_proc_tx);
@@ -435,8 +417,13 @@ void nr_ue_pbch_procedures(uint8_t gNB_id,
     ue->pbch_vars[gNB_id]->pdu_errors++;
 
     if (ue->pbch_vars[gNB_id]->pdu_errors_conseq>=100) {
-      LOG_E(PHY,"More that 100 consecutive PBCH errors! Exiting!\n");
-      exit_fun("More that 100 consecutive PBCH errors! Exiting!\n");
+      if (get_softmodem_params()->non_stop) {
+        LOG_E(PHY,"More that 100 consecutive PBCH errors! Going back to Sync mode!\n");
+        ue->lost_sync = 1;
+      } else {
+        LOG_E(PHY,"More that 100 consecutive PBCH errors! Exiting!\n");
+        exit_fun("More that 100 consecutive PBCH errors! Exiting!\n");
+      }
     }
   }
 
@@ -1615,7 +1602,8 @@ int is_pbch_in_slot(fapi_nr_config_request_t *config, int frame, int slot, NR_DL
 int phy_procedures_nrUE_RX(PHY_VARS_NR_UE *ue,
                            UE_nr_rxtx_proc_t *proc,
                            uint8_t gNB_id,
-                           uint8_t dlsch_parallel
+                           uint8_t dlsch_parallel,
+                           notifiedFIFO_t *txFifo
                            )
 {                                         
   int frame_rx = proc->frame_rx;
@@ -1772,6 +1760,13 @@ int phy_procedures_nrUE_RX(PHY_VARS_NR_UE *ue,
   }
 
 #endif //NR_PDCCH_SCHED
+  
+  // Start PUSCH processing here. It runs in parallel with PDSCH processing
+  notifiedFIFO_elt_t *newElt = newNotifiedFIFO_elt(sizeof(nr_rxtx_thread_data_t), proc->nr_slot_tx,txFifo,processSlotTX);
+  nr_rxtx_thread_data_t *curMsg=(nr_rxtx_thread_data_t *)NotifiedFifoData(newElt);
+  curMsg->proc = *proc;
+  curMsg->UE = ue;
+  pushTpool(&(get_nrUE_params()->Tpool), newElt);
 
 #if UE_TIMING_TRACE
   start_meas(&ue->generic_stat);
@@ -1904,6 +1899,7 @@ int phy_procedures_nrUE_RX(PHY_VARS_NR_UE *ue,
   VCD_SIGNAL_DUMPER_DUMP_FUNCTION_BY_NAME(VCD_SIGNAL_DUMPER_FUNCTIONS_PDSCH_PROC, VCD_FUNCTION_OUT);
 
  }
+ 
 #if UE_TIMING_TRACE
 start_meas(&ue->generic_stat);
 #endif
diff --git a/openair1/SIMULATION/NR_PHY/dlsim.c b/openair1/SIMULATION/NR_PHY/dlsim.c
index f247860e00..6fb03e46b7 100644
--- a/openair1/SIMULATION/NR_PHY/dlsim.c
+++ b/openair1/SIMULATION/NR_PHY/dlsim.c
@@ -202,6 +202,8 @@ int is_x2ap_enabled(void)
   return 0;
 }
 
+void processSlotTX(void *arg) {}
+
 //nFAPI P7 dummy functions
 
 int oai_nfapi_dl_tti_req(nfapi_nr_dl_tti_request_t *dl_config_req) { return(0);  }
@@ -286,6 +288,20 @@ void nr_dlsim_preprocessor(module_id_t module_id,
   AssertFatal(ps->mcsTableIdx >= 0 && ps->mcsTableIdx <= 2, "invalid mcsTableIdx %d\n", ps->mcsTableIdx);
 }
 
+typedef struct {
+  uint64_t       optmask;   //mask to store boolean config options
+  uint8_t        nr_dlsch_parallel; // number of threads for dlsch decoding, 0 means no parallelization
+  tpool_t        Tpool;             // thread pool 
+} nrUE_params_t;
+
+nrUE_params_t nrUE_params;
+
+nrUE_params_t *get_nrUE_params(void) {
+  return &nrUE_params;
+}
+
+void do_nothing(void *args) {
+}
 
 int main(int argc, char **argv)
 {
@@ -885,6 +901,7 @@ int main(int argc, char **argv)
   unsigned int errors_bit    = 0;
   uint32_t errors_scrambling = 0;
 
+  initTpool("N", &(nrUE_params.Tpool), false);
 
   test_input_bit       = (unsigned char *) malloc16(sizeof(unsigned char) * 16 * 68 * 384);
   estimated_output_bit = (unsigned char *) malloc16(sizeof(unsigned char) * 16 * 68 * 384);
@@ -1135,7 +1152,8 @@ int main(int argc, char **argv)
         phy_procedures_nrUE_RX(UE,
                                &UE_proc,
                                0,
-                               dlsch_threads);
+                               dlsch_threads,
+                               NULL);
         
         //printf("dlsim round %d ends\n",round);
         round++;
diff --git a/openair1/SIMULATION/NR_PHY/prachsim.c b/openair1/SIMULATION/NR_PHY/prachsim.c
index 26c1518188..54ec6ab9da 100644
--- a/openair1/SIMULATION/NR_PHY/prachsim.c
+++ b/openair1/SIMULATION/NR_PHY/prachsim.c
@@ -174,6 +174,20 @@ int nr_derive_key(int alg_type, uint8_t alg_id,
   return 0;
 }
 
+typedef struct {
+  uint64_t       optmask;   //mask to store boolean config options
+  uint8_t        nr_dlsch_parallel; // number of threads for dlsch decoding, 0 means no parallelization
+  tpool_t        Tpool;             // thread pool 
+} nrUE_params_t;
+
+nrUE_params_t nrUE_params;
+
+nrUE_params_t *get_nrUE_params(void) {
+  return &nrUE_params;
+}
+
+void processSlotTX(void *arg) {}
+
 int main(int argc, char **argv){
 
   char c;
diff --git a/openair1/SIMULATION/NR_PHY/ulsim.c b/openair1/SIMULATION/NR_PHY/ulsim.c
index 920994810d..e28f404f21 100644
--- a/openair1/SIMULATION/NR_PHY/ulsim.c
+++ b/openair1/SIMULATION/NR_PHY/ulsim.c
@@ -198,6 +198,19 @@ int nr_derive_key(int alg_type, uint8_t alg_id,
   return 0;
 }
 
+typedef struct {
+  uint64_t       optmask;   //mask to store boolean config options
+  uint8_t        nr_dlsch_parallel; // number of threads for dlsch decoding, 0 means no parallelization
+  tpool_t        Tpool;             // thread pool 
+} nrUE_params_t;
+
+void processSlotTX(void *arg) {}
+
+nrUE_params_t nrUE_params;
+
+nrUE_params_t *get_nrUE_params(void) {
+  return &nrUE_params;
+}
 // needed for some functions
 uint16_t n_rnti = 0x1234;
 openair0_config_t openair0_cfg[MAX_CARDS];