the in_sync flag has been misused in some parts of the code
receiving a PHY measurement for a cell does not automatically
mean that we are "in sync", as it is used in RRC as a condition
to indicate if we are camping on cell
the in_sync/out_of_sync condition should only be altered by the
corresponding functions sent from the PHY for the currently selected
cell (and not automatically for all cells)
the commit also removes the deletion of cells which are not "in sync"
from the list of neighbor cells. Instead, RRC calls a clean-up function
peridically that deletes old cells eventually.
move non-type-specific methods of proc_t to its base class.
procedure state machine was simplified via a future-type. Now procedures dont get stuck until the user reads the procedure outcome.
made the NAS procedures more event trigger/reaction-based.
before cell reselection was postponed to the next TTI but this is not possible in
the case where the serving cell changes.
if the "old" serving cell is no longer suitable, we would directly go to RRC IDLE below.
Therefore we need to select the new serving cell before sending the reestablishment
request.
this is tested in TC_8_5_1_1
the main issue was the plmn_selected variable. the solution
was to create a function to enter the deregistered state that resets the variable
and performs all other actions that need to be performed, like clearing
EPS bearers.
The patch also extends the SS to support AT commands to enable/disable
data services and now succesfully passes TC_9_2_2_1_7
* Calling tx_end() from radio_reset() causes long execution time in B210. Using boolean to call later instead. Fixes RLF failing in B210. Checked in X300
* Fix issue #655
if phy measurements are received and processes before a serving cell
is selected, an invalid neighbor cell is added because the uninitlized
serving cell is used.
fix the issue by checking if the serving cell is valid
RRC-PHY interface refactor. Moved RRC-MAC interface to rrc_asn1_utils and created RRC-PHY interface also in rrc_asn1_utils. All ASN1 includes should be made from rrc_asn1_utils only keeping ue_interfaces clean of ASN1
Tested with different common and dedicated configurations (64QAM UL, 256QAM, CA, SRS enabled/disabled, etc)
* first step towards moving MAC timers to stack. Mac is still using its own timers. srslte::timers class can be restructured.
* moved timers out of the UE MAC
this fixes the issue when the stack is torn down if, for example,
the radio couldn't be loaded correctly. it will hence call stop() on all stack
components which are not initialzized yet, and logging therefore doesn't work.
the log object is know during contruction time and therefore can be passed
in as soon as possible.
- add interface to RRC to allow NAS to query the DRB for a given
EPS bearer id
- extend interface for NAS to setup GW to also pass LCID of bearer
- in NAS, use this new interface to pass actual LCID of the default
DRB when creating the TUN device
this will also extend all classes that use srslte::thread
to specify the name of the thread in the ctor as well
as to set the name of the worker threads in the thread pool
the thread name will be displayed in gdb.
- abstract UE object now consists of a radio, a PHY, and a stack layer
- add new stack abstraction layer that combines MAC, RLC, RRC, PDCP, NAS and GW
- PHY layer now has a single stack interface and does not talk to MAC and RRC seperatly