| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |\ |
|
| | |
| |
| |
| | |
Luzpaz went ahead and found typos using `codespell -q 3`.
|
| | | |
|
| | |\
| | |
| | |
| | |
| | | |
Conflicts:
cmake/Modules/GrSwig.cmake
|
| | | |
| | |
| | |
| | | |
than the default payload length.
|
| | | | |
|
| |\| | |
|
| | |\| |
|
| | | |
| | |
| | |
| | |
| | |
| | | |
Add a test test that compares the additive scrambler operation with a
reference implementation to highlight the incorrect behavior when
reset is tag triggered.
|
| |\| | |
|
| | |\ \ |
|
| | | | | |
|
| | | |/ |
|
| |\| | |
|
| | |/
| |
| |
| |
| |
| |
| |
| | |
Added payload bits per symbol parameter to the header_format_default
block. This allows it to be used with payload modulations besides, for
example, BPSK.
Fixes #1162
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| |
| | |
* This also removes blks2 from gr-digital; however,
the old style OFDM code still uses this and are
broken until they are removed.
|
| |\| |
|
| | | |
|
| | | |
|
| |/
|
|
|
|
| |
CPack is not used, unmaintaned, and broken.
This does not eliminate any MSVC build functionality.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
See documentation for details. In brief:
- digital_protocol_* classes added (replace packet header blocks)
- header_buffer class added
- Lots of QA
- default packet formatter: contains access code and payload length
fields.
- counter packet formatter: also contains the bits/sym in the payload
modulation and a packet number counter.
- Also adds QA code and the infrastructure for C++ QA code in
gr-digital.
- New block to use the packet formatter objects.
- The header and payload are transmitted as PDU messages out of different ports.
- header_buffer class for header formatting.
- Used by packet_formatter_x classes to more easily add and format
header data for transmission.
- New blocks that parse headers.
- Handles either hard or soft bits. Produces an info message to
instruct payload demod.
- 'receive mode' to header_buffer class.
- Designed to make extracting fields from the received header bits
easier for packet formatter classes. Use extract_field[N] to get
the different fields out of a collected set of header bits. *
Docs describing both functions with small examples for how to use
them.
- Don't force access code to be a multiple of 8
- need to verify everything else handles this correctly.
- Throw an exception in the ctor if setting the code failed and emit a
log message.
Committed-By: Martin Braun <martin.braun@ettus.com>
|
| |\ |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
- Added padding feature (trigger/tag can now be off by some items)
- Payload offset can also be specified
- Fixed some index counting bugs
- More and better unit tests, cleaned up the unit test file
- Cleanups:
- Consistent whitespace
- Consistent use of size_t and other types
- Used more enums where it increases readability
|
| |\ \
| |/
|/| |
|
| | |
| |
| |
| |
| |
| |
| | |
- For unusual carrier allocations, the parser would report the wrong
number of payload OFDM symbols => Fixed.
- Updated QA code to track this use case.
- Minor whitespace/indentation fixes
|
| |\ \ |
|
| | | | |
|
| |\ \ \
| |_|/
|/| | |
|
| | |/ |
|
| |/
|
|
| |
documentation
|
| |
|
|
|
|
|
| |
The set_taps was redone a while ago to better handle the differential
taps as well as the normal ones, but this made externally resetting
them impossible. Adding a new update_taps function to properly handle
this case so as to not change the API. We should fix this in 3.8.
|
| | |
|
| | |
|
| |
|
|
| |
manual.
|
| |
|
|
| |
the bit stream
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
set_symbol_table didn't work with additonal qa_tests
|
| |\ |
|
| | |
| |
| |
| | |
Addresses issue #745. When rate < 1 (downsampling), take into account the amount of downsampling (decimation) and adjust the LPF's passband to reduce aliasing.
|
| | |
| |
| |
| |
| |
| |
| |
| | |
the decisions are calculated in the C++ code and some issues with the QAM16 constellation in particular.
This addresses issue #737. The patch attached to that issue is not actually valid and is only an ordering problem/confusion. I will be adding an example GRC flowgraph that compares the output of the hard decision and soft decision versions to the original input stream to show how they match.
Increased testing coverage in the QA to test certain known points as well as random samples.
|
| |\| |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
fixed a few typos in explanatory comments.
Also deleted code in
if 1:
...
else:
...
clauses.
Also changed a few [1.0 for i in range(...)] to in-line
[1.0] * (fft_length // 2) for superior readability.
|
| |\| |
|
| | | |
|
