[Bug] Accumulation of unprocessed samples in channels #47
Milestone
Comments
This was referenced Mar 29, 2023
Merged
agrenier-gnss
added a commit
that referenced
this issue
Mar 29, 2023
Code debugged and working. Known issues: - A-GNSS not working yet - 10-20 m bias in south-est direction, probably due to bad handling of unprocessed samples number. See issue #47 and #48 --------- * Restructure Receiver class #36 Started restructuration of receiver, saving. * Saving * Saving * Debug new code structure #40 Acquisition and tracking debugged and working. Decoding and positioning still need to be tested. Code excecution is much slower than previous structure, needs to be reviewed why. * Debug new code structure #40 Code accelerated by doing a vector version of EPL * Debug new code structure #40 Added report maker * Implement SNR / CN0 estimation #32 (#43) Started implemented CN0 estimation. Co-authored-by: Antoine Grenier <[email protected]> * Debug new code structure #40 Simplified the configuration system for visualisation. File cleanup * [Code cleanup #29] * [Code cleanup #29] * [Debug new code structure #40] Debugging decoding, not working yet. * [Debug new code structure #40] Decoding still not working * [Debug new code structure #40] Preambule decoding working, error was an error in preambule bits inverted constant. * [Debug new code structure #40] Positioning running but not correct. * [Debug new code structure #40] Positioning debugged and working. Apparently, the issue [#27] has been fixed in the restructure. However, the actual reason why is still unclear. This needs to be documented. * [Code cleanup #29] Cleaning up the code with new architecture. * [Debug new code structure #40] Removing the two bit shift because it is not yet fully understood. * [Debug new code structure #40] Trying to add A-GNSS but now there is a shit north * [Debug new code structure #40] Normal positioning working again, bias is still not fixed. A-GNSS create another bias still. --------- Co-authored-by: Antoine Grenier <[email protected]>
agrenier-gnss
added a commit
that referenced
this issue
May 29, 2023
Major updated, implementing clearer and more modular code structure.
Tracking phase has been improved with different stages.
A Channel Manager class been implemented, dealing with sharing the data amongst the channels.
Known issues have been submitted in the related section.
---
* Started adding C implementation for tracking function.
* Saving
* saving
* tracking conversion update
* tracking conversion final
* acquisition conversion update
* acquisition conversion final
* Added unit test for generateReplica
* Unit test for tracking in C working.
* Tracking tested on real data and working.
delay and phase lock loop c implementation have been
commented for further testing.
* Fixed delay and phase lock loop in C
* Try to improve positioning without BRDC assist
* Code cleanup
* Saving
* Added plt plots generation and processing time estimation for ICL-GNSS publication.
* Enhance channel #31
Starting the channel update
* [Enhance channel #31]
Added tracking functions in new channel. Not tested yet.
* Adding ChannelManager class (#39)
Updated code structure to inlude a ChannelManager class. Code has not been debug/tested yet, thus the main implementation and old class are in used and will be changed in the next update.
The code has also been upgraded from Python 3.8 to 3.11, with possible consequences on the C implementation. Code has been tested with the main.py in Python3.11 and pure Python (no C calling) and debugged, it is in working state.
---------
* Enhance channel #31
Saving some changes.
* Add ChannelManager class #33
Saving before weekend
* Add ChannelManager class #33
Saving current state
* Add ChannelManager class #33
ChannelManager class created, need to be integrated within a new receiver object.
* Add ChannelManager class #33
Adapted Channel classes to fit the new ChannelManager class.
* Migrate to Python 3.11 #37
Updated requirements
* Add ChannelManager class #33
Updated Example
* #35 CircularBuffer and linked files updated.
---------
Co-authored-by: Antoine Grenier <[email protected]>
* Update licence with authors #34 (#41)
* Restructure receiver (#49)
Code debugged and working.
Known issues:
- A-GNSS not working yet
- 10-20 m bias in south-est direction, probably due to bad handling of unprocessed samples number. See issue #47 and #48
---------
* Restructure Receiver class #36
Started restructuration of receiver, saving.
* Saving
* Saving
* Debug new code structure #40
Acquisition and tracking debugged and working. Decoding and positioning still need to be tested.
Code excecution is much slower than previous structure, needs to be reviewed why.
* Debug new code structure #40
Code accelerated by doing a vector version of EPL
* Debug new code structure #40
Added report maker
* Implement SNR / CN0 estimation #32 (#43)
Started implemented CN0 estimation.
Co-authored-by: Antoine Grenier <[email protected]>
* Debug new code structure #40
Simplified the configuration system for visualisation.
File cleanup
* [Code cleanup #29]
* [Code cleanup #29]
* [Debug new code structure #40]
Debugging decoding, not working yet.
* [Debug new code structure #40]
Decoding still not working
* [Debug new code structure #40]
Preambule decoding working, error was an error in preambule bits inverted constant.
* [Debug new code structure #40]
Positioning running but not correct.
* [Debug new code structure #40]
Positioning debugged and working.
Apparently, the issue [#27] has been fixed in the restructure. However, the actual reason why is still unclear.
This needs to be documented.
* [Code cleanup #29]
Cleaning up the code with new architecture.
* [Debug new code structure #40]
Removing the two bit shift because it is not yet fully understood.
* [Debug new code structure #40]
Trying to add A-GNSS but now there is a shit north
* [Debug new code structure #40]
Normal positioning working again, bias is still not fixed.
A-GNSS create another bias still.
---------
Co-authored-by: Antoine Grenier <[email protected]>
* Update README.md
* Update README.md
Tested the project in Ubuntu as well as MacOS, design can not run in MacOS yet
* Update README.md
missing .result folder
* 50 coherent tracking (#51)
Implemented a new tracking state machine, for easier transition between tracking modes.
Coherent tracking is not implemented. As it mostly favors low SNR satellites, implementation is canceled for the moment.
Positioning has a known offset of 20m in the East
* [Implement coherent integration in tracking #50]
Replaced the correlator buffer, working.
* [Implement coherent integration in tracking #50]
Saving
* [Implement coherent integration in tracking #50]
Saving, not working yet.
* [Implement coherent integration in tracking #50]
Saving
* [Implement coherent integration in tracking #50]
Added new filters implementations.
Trying to implement FLL, not working.
* [Implement coherent integration in tracking #50]
Saving
* [Implement coherent integration in tracking #50]
Saving
* [Implement coherent integration in tracking #50]
New tracking state machine working with Kaplan filters. Coherent tracking is still not possible.
* [Implement coherent integration in tracking #50]
Tracking working, not positioning
* [Implement coherent integration in tracking #50]
saving
* [Implement coherent integration in tracking #50]
Tracking, decoding and positioning working.
Offset of 20m in the east for unknown reasons.
* [Implement coherent integration in tracking #50]
Coherent tracking still not working.
---------
Co-authored-by: Antoine Grenier <[email protected]>
* Quick fix of Assisted GNSS
* [Code cleanup #29]
* Small fixes.
---------
Co-authored-by: Antoine Grenier <[email protected]>
Co-authored-by: Hans Jakob Damsgaard <[email protected]>
Co-authored-by: Jie Lei <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
This is a follow-up issue to #27.
Unprocessed samples in channels keep growing and can reach the size of two codes. This need to be fixed to be sure all samples are processed accordingly.
This might be the reason for the large bias present in the positioning.
The text was updated successfully, but these errors were encountered: