After gNB dropping the UE connection the UE doesn't trigger new RA. But when I try to stop the UE by pressing ctrl+C, I see the message saying SR counter reached `sr_MaxTransmissions` and it triggers the new RA before stopping. So there is an issue with update of the counter or somewhere. UE logs: ``` [NR_PHY] [UE 0] RSRP = -37 dBm [NR_PHY] [UE 0] RSRP = -40 dBm [NR_PHY] [UE 0] RSRP = -37 dBm [NR_PHY] [UE 0] RSRP = -40 dBm [NR_PHY] [UE 0] RSRP = -38 dBm [NR_PHY] ============================================ [NR_PHY] [UE 0] Harq round stats for Downlink: 147/2/0 [NR_PHY] ============================================ [NR_PHY] [UE 0] RSRP = -37 dBm [NR_PHY] [UE 0] RSRP = -39 dBm [NR_PHY] [UE 0] RSRP = -37 dBm [NR_PHY] [UE 0] RSRP = -40 dBm [NR_PHY] [UE 0] RSRP = -37 dBm [NR_PHY] [UE 0] RSRP = -40 dBm [NR_PHY] [UE 0] RSRP = -37 dBm ^CPress ^C again to trigger immediate shutdown [NAS] [UE 0] Received NAS_DEREGISTRATION_REQ mac 6f 1a 7 28 [NR_PHY] [UE 0] RSRP = -39 dBm [NR_PHY] ============================================ [NR_PHY] [UE 0] Harq round stats for Downlink: 147/2/0 [NR_PHY] ============================================ [NR_PHY] [UE 0] RSRP = -38 dBm [NR_PHY] [UE 0] RSRP = -37 dBm [NR_PHY] [UE 0] RSRP = -39 dBm [NR_MAC] [UE 0] SR not served! SR counter 64 reached sr_MaxTransmissions 64 [NR_MAC] Triggering new RA procedure for UE with RNTI 4174 [NR_MAC] Initialization of 4-step contention-based random access procedure [NR_MAC] PRACH scheduler: Selected RO Frame 359, Slot 19, Symbol 0, Fdm 0 [PHY] PRACH [UE 0] in frame.slot 359.19, placing PRACH in position 1804, msg1 frequency start 0 (k1 0), preamble_offset 0, first_nonzero_root_idx 0 [NR_MAC] [UE 0][RAPROC][RA-RNTI 0005] Got BI RAR subPDU 267 ms [NR_MAC] [UE 0][RAPROC][RA-RNTI 010b] Got RAPID RAR subPDU [NR_MAC] [UE 0][RAPROC][360.7] Found RAR with the intended RAPID 3 [MAC] received TA command 31 [PHY] RAR-Msg2 decoded [NR_MAC] [RAPROC][360.17] RA-Msg3 transmitted [NR_PHY] ============================================ [NR_PHY] [UE 0] Harq round stats for Downlink: 148/2/0 [NR_PHY] ============================================ [NR_PHY] ============================================ [NR_PHY] [UE 0] Harq round stats for Downlink: 148/2/0 [NR_PHY] ============================================ [NR_PHY] ============================================ [NR_PHY] [UE 0] Harq round stats for Downlink: 148/2/0 [NR_PHY] ============================================ [NR_PHY] ============================================ [NR_PHY] [UE 0] Harq round stats for Downlink: 148/2/0 [NR_PHY] ============================================ [NR_PHY] ============================================ [NR_PHY] [UE 0] Harq round stats for Downlink: 148/2/0 [NR_PHY] ============================================ [NR_PHY] ============================================ [NR_PHY] [UE 0] Harq round stats for Downlink: 148/2/0 [NR_PHY] ============================================ [NR_PHY] ============================================ [NR_PHY] [UE 0] Harq round stats for Downlink: 148/2/0 [NR_PHY] ============================================ Returned from ITTI signal handler oai_exit=1 [HW] releasing USRP [HW] [recv] received 8163 samples out of 23040 D[HW] Time: 252.594 s Fragmentation offset: 60364 ERROR_CODE_OVERFLOW (Out of sequence error) ``` ## To reproduce I see this when I run with USRPs and sometime the gNB machine has real-time issues which causes the UE to skip some PUSCH orders. When it happens the gNB drops the UE and the UE won't trigger RA till press ctrl+C. Maybe this can be reproduced with rfsimulator and changing the UL channel SNR via telnet?