summaryrefslogtreecommitdiff
path: root/cesar/ce/rx/bitloading/test
AgeCommit message (Collapse)Author
2013-06-28cesar/ce/rx: change NSR margin and CE restart default values, closes #3570Jean-Philippe SAVE
We change these parameters to avoid the following case: the PBER is too high while the BER_target is at it minimum. See ticket #3942 for more details.
2013-06-14cesar/ce/rx/bl: remove unused functions, refs #4018Cyril Jourdan
Those functions are no longer necessary since it is the PBProc that does the matching between frames received and intervals.
2013-06-14cesar/ce/rx/bl/fsm: use PBProc interval matching computation, refs #4018Cyril Jourdan
Instead of re-computing frame date and trying to match it on an interval. In the same time, correct management of sound frame SRC in the tests when we want to test the data case and update ce_rx_bl_fsm_measure_to_event stub in cesar/ce/rx/test.
2013-06-14cesar/ce/rx/bl/test: fix intervals' test cases titlesCyril Jourdan
2013-06-06cesar/ce/rx/bl: enable snr collect with interval, refs #4007Thierry Carré
2013-06-06cesar/ce/rx/bl: send new default tonemap after computation, closes #4026Cyril Jourdan
2013-05-24cesar/ce/rx/bl: compute a worst case tonemap when needed, closes #3880Cyril Jourdan
This tonemap will be used as default tonemap when intervals are activated.
2013-05-24cesar/ce/rx/bl: give CE RX context to update_default_tm function, refs #3880Cyril Jourdan
This is needed for the next commits.
2013-05-24cesar/ce/rx/bl: remove useless global variablesCyril Jourdan
A structure has been created for this parameters, this globals should not be used anymore.
2013-05-24cesar/ce/rx/bl: handle interval_sound events in global FSM, closes #3957Cyril Jourdan
The goal is to avoid that sound directed to an interval we can not match are treated by global FSM as sound events, what causes a global CE restart. This can lead to too many CE restarts and to an accumulation of measure in CE's mailbox, which leads to the observed assert.
2013-05-16cesar/ce/rx/bl: fix intervals bitloading contexts init, refs #3936Cyril Jourdan
The table of next restart dates was not initialized. Random values could make the CE believe that it is not allowed to restart before a very long time...
2013-04-25cesar/ce, cesar/mac/common: move CE globals to CENicolas Schodet
2013-04-25cesar/ce, cesar/mac/common: add group interval nb in mac tonemap configurationNicolas Schodet
2013-04-25cesar/mac/common, cesar/ce/rx: get rid of mac_interval_measure_shift_tolerance_Nicolas Schodet
2013-04-18cesar/ce/rx/bl: update default tonemap when restarting an interval, refs #2588Cyril Jourdan
The idea is to keep the default tonemap when we have only one interval active that is restarting, if it is doing so to make it better. If the update function seems a bit complex, it is because it is ready to be modified to handle the case where there is more than one interval (see #3880).
2013-04-18cesar/ce/rx/bl: add a function to check tonemaps usage, refs #2588Cyril Jourdan
The functions returns a bit field where each bit stands for a TMI and indicates if corresponding tonemap is currently used as default one or on at least one interval.
2013-04-18cesar/ce/rx/bl: indicate if we should use ROBO as default TM, refs #2588Cyril Jourdan
The goal is to help us to update the default tonemap when an interval is restarting.
2013-04-18cesar/ce/rx/bl/fsm: check SRC to generate sound events, closes #3886Cyril Jourdan
With this modification, sound frames with non interval specific SRCs are directed to global FSM. When in INTERVALS state, this causes a complete CE restart, what avoid sync issues between transmitter and receiver.
2013-04-18cesar/ce: use tonemap index definesCyril Jourdan
2013-03-07cesar/ce/rx/bitloading: return NULL if no NSR blocks are available, refs #3776Nélio Laranjeiro
When no NSR blocks are available the function ce_rx_bl_get_nsr should return NULL. This can happen if the CP receive a vs_get_nsr.req MME and the CP preempts the CE before it starts using the NSR received.
2013-01-18cesar/ce/rx/bl: keep measure of the frames before beacon period, refs #2362Jalil Chemseddine
For frames crossing the beacon period, the receiver sees frame date before beacon period start date. The measures for those frames were dropped whereas they are usable...
2013-01-18cesar/ce/rx/bl: return the correct fsm id when only one interval, refs #2362Jalil Chemseddine
2013-01-18cesar/ce/rx/bl/test/fsm: correct test, refs #2366Cyril Jourdan
The test needed an override for ce_rx_bl_nsr_clean function and there was some inits missing.
2013-01-18cesar/ce/rx/bl/test: move fsm test to its own folder, refs #2771Cyril Jourdan
2013-01-18cesar/ce/rx/bl: adjust global fsm transitions for intervals, refs #2366Jérôme Jutteau
2013-01-18cesar/ce/rx/bl: create sub-fsm transitions, closes #2364Cyril Jourdan
2013-01-18cesar/ce/rx: include bsu_aclf context pointer in CE RX, refs #2366Cyril Jourdan
2013-01-18cesar/ce/rx/bl: use bsu_aclf for interval construction, refs #2366Jérôme Jutteau
2013-01-11cesar/ce/rx/bl: add pber high parameters to pber structure, refs #3251Jalil Chemseddine
2013-01-11cesar/ce/rx/bl: create structure for ber configuration, refs #3251Jalil Chemseddine
2013-01-11cesar/ce/rx/bl: group nsr parameters in ce_rx_param.h, refs #3251Jalil Chemseddine
2012-12-12cesar/ce/rx/bl: update tests to handle new polynomials BER vs NSR, closes #2560Jalil Chemseddine
2012-11-07cesar/ce/rx/bl: remove tonemap update tests, refs #3310Jalil Chemseddine
2012-10-18cesar/ce/rx/bl: take into account wrong PBs while computing BER, closes #2750Jalil Chemseddine
2012-09-12cesar/ce/rx/bl: disable BER margin update, closes #2376Jérémy Dufour
For the moment being BMU is disabled by setting the biggest tunnel possible. See #3351 for more information.
2012-09-12cesar/ce/rx: improve BMU by adding different mechanisms, refs #2376Jérémy Dufour
Replace frames counter by PB counter to trigger BMU: PBER sliding mean is computed by PB now, BMU should be triggerd by PB too. Add an oscillation counter to prevent too much restart: this is a security measure in case of the BMU go wild. Prevent BER margin to go over limits of BER target: after many BMU, the margin can in/decrease so much that it will go beyond possible BER target values. Move initialization of BMU config in a function rather than in global. Update test vectors to be in concordance with latest configuration. Improve handling of tone map at maximum
2012-09-12cesar/ce/rx/bl: do not restart BMU for low PBER when BL is at max, refs #2376Jérémy Dufour
2012-09-12cesar/ce/rx/bl: add a stat when BL can not optimize tone map, refs #2376Jérémy Dufour
2012-09-12cesar/ce/rx/bl: integrate BER margin update, refs #2376Jérémy Dufour
2012-09-12cesar/ce/rx: use an enum for restart dates table, refs #2376Jérôme Jutteau
2012-09-12cesar/ce/rx/bl: add BER margin update, refs #2376Jérémy Dufour
2012-09-07cesar/ce/rx: take tonemask from MAC configuration, closes #3248Nicolas Schodet
2012-08-22cesar/ce/rx/bl: send TM after some time with no measure on it, closes #1926Jérémy Dufour
If we receive some measurements on a tone map which differs from the default one we have configured, wait some times (3s) and send it again.
2012-07-19cesar/ce/rx: add missing update of carrier_nb, closes #3226Yacine Belkadi
Disabling the first two carriers changes the number of used carriers. So carrier_nb needs to be updated. Without that, in ce_rx_bl_compute_tone_map_iterative(), ce_rx_bl_update_tone_map_under_ber_consign() will put r_nb elements in opti[][], but ce_rx_bl_sort_optimization() will sort i_nb elements of opti[][], where r_nb is the real number of used carriers, and i_nb is the initial (non updated) number of used carriers. So if, for example, i_nb is 1355 and r_nb is 1353, this will lead to the use of whatever random values are present at the end of opti[][]. Depending on the values, this may result in: - "../../mac/common/src/tonemap.c:233: assertion failure in tonemap_get_tone: index < PHY_CARRIER_NB" - incorrect CE computation results
2012-06-28cesar/ce/rx/bl: compute PBER sliding mean by PB not by frame, closes #3215Jalil Chemseddine
2012-06-28cesar/ce/rx/bl: refactor PBER module, refs #3215Jérémy Dufour
This commit includes: - refactoring of the PBER function to create a module to handle the PBER, - propose a new re-initialisation system - an impossible value means using the value from the first frame, - a valid value will be used as the initialization. - set default value of coefficient to last value defined by TNS team, - reset the PBER sliding mean after each CE restart, - isolate corresponding tests.
2012-06-28cesar/ce/rx/bl: remove FEC rate from BERvsNSR polynomials, closes #3101Jalil Chemseddine
BER vs NSR polynomials are not FEC rate dependent. Only BER target polynomials depends on FEC rate.
2012-06-28cesar/ce/rx/bl: new polynomials BER vs NSR, refs #3101Jalil Chemseddine
This new polynomials: * are not FEC rate dependant, * does not cross each other when modulation increases.
2012-06-28cesar/ce/rx/bl: evaluate BER polynomials on all NSR ranges, refs #3101Jalil Chemseddine
2012-06-20cesar/ce/rx add new tone mask to handle faulty carriers, closes #3095Jalil Chemseddine
Add a new tone mask which disable faulty carriers in CE bit-loading computation. This tone mask differs from the general tone mask in the station and is only used in bit-loading.