386 Commits (e654062a7a21d78216e99ea7c56c86c6fdbb5c9c)

Author SHA1 Message Date
Francisco Paisana 55d62a2baa fix rlc buffer updating inside the scheduler to account for RLC headers 4 years ago
Francisco Paisana 980278fdec add comments regarding radio resource asn1 setting in the eNB 4 years ago
Francisco Paisana b40a5a958e use common function to fill rrc reconf message in handover 4 years ago
Francisco Paisana 836e7d7ede refactor scell addition in the srsenb::rrc to follow the same style of other rrc recfg fields 4 years ago
Francisco Paisana a8e11d7991 removal of the special case for rrc reconf update 4 years ago
Francisco Paisana f26b5ccde7 fix drb reestablishment during handover 4 years ago
Francisco Paisana 7bf196a6e0 removal of srb handling from srsenb::rrc::bearer_cfg_handler 4 years ago
Francisco Paisana 2ef1e286b2 fix set of rrc reconf message radio resource cfg fields 4 years ago
Francisco Paisana 890113fd32 now the rrc enb is using a common function for rr_cfg reconf for the first reconf message 4 years ago
Francisco Paisana ce78fa82ce now the rrc enb is using a common function for rr_cfg setup/reest 4 years ago
Francisco Paisana 97f2e4336e refactor of functions for filling the asn1 radio resource config structs. This functions are used during rrc setup/reest/reconf 4 years ago
Francisco Paisana 44a9ad76f1 avoid sched_ue dangling pointer
With the sched feature that allows scheduling in TTIs
ahead of time, there is no guarantee that when
the tti arrives to generate a sched result, the stored
raw sched_ue pointers are still valid. For this reason,
I now store the rnti and check if the rnti still exists.
4 years ago
Francisco Paisana 852c31c0bc log warnings when the sched ue cfg is not valid 4 years ago
Francisco Paisana 1ce928ab2e use of current / future ue sched config structs in the srsenb::rrc::ue::mac_controller as an approach to deal with the multi-step mac reconfiguration of the UE 4 years ago
Francisco Paisana 1e6d902873 cleanup of srsenb::rrc::ue::mac_controller api 4 years ago
Francisco Paisana f1bdf33d10 addition of cells provided in cell_meas_list of rr.conf to cellsToAddModList in the rrc reconfiguration message 4 years ago
Francisco Paisana c5c7700cb7 implemented DL logical channel prioritization test for finite PBRs 4 years ago
Francisco Paisana 9b10acff06 implemented logical channel prioritization in DL scheduler 4 years ago
Francisco Paisana b7ed8b1858 - added react method to to_state<...> transitions
- s1 handover composite state simplified
- the eNB now starts a HO cancellation when it receives an invalid
Handover Command
- the FSM log now prints the current state when it receives an unhandled
event
4 years ago
Francisco Paisana a761762ec8 extended cc_sched_ue fsm to account for more harmonious transition of an scell to deactivated state 4 years ago
Francisco Paisana 74c18ecf6c implementation of SCell Deactivation 4 years ago
Francisco Paisana eafc003671 Altered the way the generation of the scheduling decision for a given
TTI is made. Instead of generating one CC in each dl_sched()/ul_sched()
call, all CC decisions are generated at once. This avoids race
conditions with calls from the RRC to the MAC to configure UEs.
4 years ago
Francisco Paisana c46402c836 fix phy configuration during handover on srsenb. The function srsenb::rrc::ue::fill_scell_to_addmod_list assumes that the phy_cfg of the PCell is up-to-date, which was not the case. With this PR, the srsenb rrc updates the phy_cfg before the ho cmd is sent, but, in the case of intra-enb ho, the updates are not submitted to the phy before the reception of the crnti mac ce 4 years ago
Francisco Paisana 0f51d70aac fix measObj addition in srsenb. Now the enb considers both scells and meas_cells in the rr.conf 4 years ago
Francisco Paisana d42cdca729 fix scell activation in the phy during intra-enb handover. 4 years ago
Francisco Paisana 688dda30a4 favour carriers with best channel conditions for UL grants due to SRs 4 years ago
Francisco Paisana 137e554eb3 reestablish pdcp bearers is required in case of AM bearers 4 years ago
Francisco Paisana 4c88404801 fix measurement report handling in case of intra enb handover 4 years ago
Francisco Paisana 942c28881d converted member method of list of cells in the enb to a free function 4 years ago
Francisco Paisana 3a924a1d80 avoid sending cellsToAddModList if cell offsets are zero 4 years ago
Francisco Paisana da70c0fdaf fix cellsToAddModList handling. Now the eNB always adds PCell+SCells as measObjects in the RRC Reconfiguration messages. It may also further include cellsToAddModList if they are provided in the rr.conf file 4 years ago
Francisco Paisana 141960fbda remove unnecessary shared_ptrs and use value semantics instead for the var_meas_cfg of the different enb cells 4 years ago
Francisco Paisana 7e6487b6fa perform PUCCH+PRACH PRB reservation right at the start of a new TTI in the scheduler 4 years ago
Ismael Gomez 59b6301cc8 Remove legacy arbritrary 1s timeout for eNB requests 4 years ago
Ismael Gomez 38e2ffe414 Use Format1A when using CA and 15 PRB due to DCI size ambiguity 4 years ago
Francisco Paisana fddcc3d8ea In the case of Handover, postpone the activation of the CQI and SR config present in the HO command for after the RA proc completion to the target eNB 4 years ago
Francisco Paisana 13caf287f9 fix S1 Handover with UE CA. The HandoverPreparation message was not being filled correctly 4 years ago
Ismael Gomez a4835dd2c8
More accurate MCS reduction when PUSCH carries UCI (#1630) 4 years ago
Francisco Paisana cc469fc7a3 fix reestablishment after S1 handover by setting ncc accordingly 4 years ago
Francisco Paisana 0036941af4 added extra state to S1 target enb mobility FSM. Accounts for out-of-order arrival of MMEStatusTransfer. Also, now we avoid creating DRBs before the MMEStatusTransfer 4 years ago
Francisco Paisana bb5dd92dca implemented s1 handover cancellation procedure to force the target enb to release the ue ctxt 4 years ago
Francisco Paisana 8c9e596f89 refactored s1ap handover request interface. cleaned up rrc tenb s1 handover code 4 years ago
Francisco Paisana 7cd4f45d62 fsm refactor.
- avoid triggering directly a subfsm
- improved metafunctions
- other cleanups
4 years ago
Francisco Paisana 5d1343fe75 implemented S1 bearer status transfer 4 years ago
Francisco Paisana db8af399f8 extended rrc tenb s1 handover test. implemented layer reconfig during s1 handover 4 years ago
Francisco Paisana 7605c7df23 added s1ap handover notification, fixed s1ap asn1 bug, and fixed bad fields for admitted erabs in ho req ack 4 years ago
Francisco Paisana 20b69fb22e substitution of ul_alloc_t for prb_interval 4 years ago
Francisco Paisana 26f46ea067 add S1 Handover Request handling in target eNB 4 years ago
Francisco Paisana caf862c99f moved expected, bounded_bitset, move_callback to adt folder. created a test for bitset 4 years ago
Ismael Gomez 382811e10e
Update S1AP RNTI context when doing Reestablishment (#1582) 4 years ago
Francisco Paisana 35fa5346b9 only reduce mcs if the current tti requires uci 4 years ago
Francisco Paisana 58c4bcf288 reduce mcs for UL grants carrying UCI 4 years ago
Andre Puschmann a7911f0dab pdcp,rlc: refactor SDU queueing policy
this patch refactors the SDU queuing and dropping policy of the RLC and PDCP layer.
the previous design had issues when packets have been generated at a higher
rate above the PDCP than they could be consumed below the RLC.

When the RLC SDU queues were full, we allowed two policies, one to block on the write
and the other to drop the SDU. Both options are not ideal because they either
lead to a blocking stack thread or to lost PDCP PDUs.

To avoid this, this patch makes the following changes:

* PDCP monitors RLC's SDU queue and drops packets on its north-bound SAP if queues are full
  * a new method sdu_queue_is_full() has been added to the RLC interface for PDCP
* remove blocking write from pdcp and rlc write_sdu() interface
  * all writes into queues need to be non-blocking
  * if Tx queues are overflowing, SDUs are dropped above PDCP, not RLC
* log warning if RLC still needs to drop SDUs
  * this case should be avoided with the monitoring mechanism
4 years ago
Francisco Paisana 5e06430455 removal of stack handler interface 4 years ago
Francisco Paisana 039977aeb5 renamed interfaces 4 years ago
Francisco Paisana 1a9c04bf2f use of task scheduler handle in pdcp 4 years ago
Francisco Paisana e9f34c7613 using task scheduler in enb stack 4 years ago
Francisco Paisana 4f5e65781f created task scheduler class to deal with timers, thread pool, multiqueue, internal tasks 4 years ago
Pedro Alvarez 89b6e0f714 Send paging message to all cells. 4 years ago
Francisco Paisana fe8f8be62d apply separate cqi configurations to scells 4 years ago
Francisco Paisana e550bf726d updated fsm to allow enter methods to receive event that caused them 4 years ago
Ismael Gomez e378a45db0 Reorganized priorities to avoid GW starve the CPU 4 years ago
Francisco Paisana 97d20ca579 remove obsolete bearer cfg methods 4 years ago
Francisco Paisana 0929a9de62 created rrc mac_controller class to handle the configuration of scheduler 4 years ago
Francisco Paisana 7ab4b5f723 fix long bsr handling 4 years ago
Francisco Paisana af58597647 fix mac tb building in scheduler 4 years ago
Francisco Paisana cb7f620254 created separate class to manage bearers buffers/status reports in the scheduler 4 years ago
Francisco Paisana 2cb80f9bf3 remove rnti in mac/phy in staged manner. First the "upper mac", and then the "lower mac" and phy 4 years ago
Ismael Gomez 493b7ea2da Remove SNR threshold based link failure detection 4 years ago
Francisco Paisana f0874b780b add radio_base/null classes 4 years ago
Francisco Paisana 64e8f5e3ed update license year 4 years ago
Pedro Alvarez 9c1a7d2cf1 Removed get_bearer_status in PDCP. Renamed get_state/set_state to
get_bearer_state/set_bearer_state.
4 years ago
Pedro Alvarez fb088a8c54 Restoring PDCP state only when using RLC AM. 4 years ago
Pedro Alvarez 66d4f320e5 Added intereface to set PDCP state after reestablishment. 4 years ago
Pedro Alvarez 217f16d137 Getting PDCP state on re-establishment. 4 years ago
Pedro Alvarez 33567d81db Added interface to get internal PDCP state. 4 years ago
Andre Puschmann 06d31f46ba enb: set maximum allowed TTI difference between PHY and Stack TTI clocks for ZMQ
this fixes the issue where PHY and MAC threads, and the processing
of events therein, are drifting apart too much.

This patch effectivly enforces a strong wait on the PHY if the
stack thread is too slow in processing its tasks.

For RF hardware the default is still 8192 TTIs, for ZMQ 1 TTI is used.
4 years ago
Andre Puschmann a1d64c1efe enb: improve error message when RAR cannot be sent 4 years ago
Francisco Paisana 3c2e3dbdbf added latest mac nr updates 4 years ago
Francisco Paisana d859c33035 added check for UCI collision during DL scheduling for CA 4 years ago
Francisco Paisana 53b85691b5 moved sf result to sched class. This way each carrier has a view to the results of the remaining carriers in the same subframe 4 years ago
Francisco Paisana 5d33acdb53 changed name of sched_ue_carrier class 4 years ago
Ismael Gomez 60b5fb9a74 Refactor variable name 4 years ago
Pedro Alvarez 20c5ea7d17 Fix warning with GCC 10.1 with EEA/EIA prefrence list. 4 years ago
Francisco Paisana 7ce23e0afb extend handover intraenb handover fsm to account for multiple C-RNTI CEs 5 years ago
Francisco Paisana 92c4b3f226 fix non-final sdap/mac/rlc dtor error 5 years ago
Francisco Paisana e4a91a207d remove unused variables 5 years ago
Francisco Paisana b843585abc add srsgnb stack class 5 years ago
Francisco Paisana f5174415f8 add srsenb::sdap class 5 years ago
Francisco Paisana 613003ac53 add srsenb::rrc class 5 years ago
Francisco Paisana 0916e29de4 add srsenb::rlc_nr class 5 years ago
Francisco Paisana 9001117b2d add srsenb::pdcp_nr class 5 years ago
Pedro Alvarez 960c0e97cd Added handling of RRC re-establishment to srsENB.
* Added the appropriate code for handling and sending the
   re-establishment procedure messages to rrc_ue.c/.h.
 * Triggered RRC reconfiguration after the reception of RRC
   re-establishment complete
 * Refreshed K_eNB at the reception of re-establishment
   request
 * Changed the mapping of TEIDs to RNTIs in the GTP-U layer,
   as the RNTI might change with reestablishment.
5 years ago
Francisco Paisana 452c043370 created a global map of lcids to srbs/drbs for NR 5 years ago
Francisco Paisana 6a164e032d added srsue::rrc_nr 5 years ago
Francisco Paisana 026628a84f use of reestablishment method to reset the RLC bearers during handover 5 years ago
Francisco Paisana 9a0ea6e08c use log_helper for log macros 5 years ago
Francisco Paisana edba7f57e4 corrected mac_nr_pdu includes 5 years ago
Francisco Paisana d7b35ba7d1 added nr mac 5 years ago
Andre Puschmann a0e4b18507 enb_mac: pre-allocate UE object for quicker PRACH/RAR handling
we've seen long delays in handling PRACH on low-power devices
such as the RPi4. This was mainly caused by long delays
in creating the UE object on the fly during PRACH handling.

This patch pre-allocates one UE object that is then used
during the PRACH.
5 years ago