Forward Error Correction with a Cauchy MDS block erasure codec is used to prevent block loss. This can make the UDP transmission more robust particularly over WiFi links.
Please note that there is no provision for handling out of sync UDP blocks. It is assumed that frames and block numbers always increase with possible blocks missing. Such out of sync situation has never been encountered in practice.
The remote SDRangel instance that sends the data stream is controlled via its REST API using a separate control software for example [SDRangelcli](https://github.com/f4exb/sdrangelcli)
A sample size conversion takes place if the stream sample size sent by the remote instance and the Rx sample size of the local instance do not match (i.e. 16 to 24 bits or 24 to 16 bits). Best performance is obtained when both instances use the same sample size.
The plugin will be built only if the [CM256cc library](https://github.com/f4exb/cm256cc) is installed in your system. For CM256cc library you will have to specify the include and library paths on the cmake command line. Say if you install cm256cc in `/opt/install/cm256cc` you will have to add `-DCM256CC_DIR=/opt/install/cm256cc` to the cmake commands.
This is the main buffer (writes from UDP / reads from DSP engine) length in units of time (seconds). As read and write pointers are normally about half the buffer apart the nominal delay introduced by the buffer is the half of this value.
Read and write pointers should always be a half buffer distance buffer apart. This is the difference in percent of the main buffer size from this ideal position.
- When positive it means that the read pointer is leading
- When negative it means that the write pointer is leading (read is lagging)
This is the current timestamp of the block of data sent from the receiver. It is refreshed about every second. The plugin tries to take into account the buffer that is used between the data received from the network and the data effectively used by the system however this may not be extremely accurate. It is based on the timestamps sent from the Remote sink channel at the other hand that does not take into account its own buffers.
There are two gauges separated by a dot in the center. Ideally these gauges should not display any value thus read and write pointers are always half a buffer apart. However due to the fact that a whole frame is reconstructed at once up to ~10% variation is normal and should appear on the left gauge (write leads).
- The left gauge is the negative gauge. It is the value in percent of buffer size from the write pointer position to the read pointer position when this difference is less than half of a buffer distance. It means that the writes are leading or reads are lagging.
- The right gauge is the positive gauge. It is the value in percent of buffer size of the difference from the read pointer position to the write pointer position when this difference is less than half of a buffer distance. It means that the writes are lagging or reads are leading.
The system tries to compensate read / write unbalance however at start or when a large stream disruption has occurred a delay of a few tens of seconds is necessary before read / write reaches equilibrium.
This is the total number of frames and number of FEC blocks separated by a slash '/' as sent in the meta data block thus acknowledged by the remote server. When you set the number of FEC blocks with (4.1) the effect may not be immediate and this information can be used to monitor when it gets effectively set in the remote server.
A frame consists of 128 data blocks (1 meta data block followed by 127 I/Q data blocks) and a variable number of FEC blocks used to protect the UDP transmission with a Cauchy MDS block erasure correction.
Using the Cauchy MDS block erasure correction ensures that if at least the number of data blocks (128) is received per complete frame then all lost blocks in any position can be restored. For example if 8 FEC blocks are used then 136 blocks are transmitted per frame. If only 130 blocks (128 or greater) are received then data can be recovered. If only 127 blocks (or less) are received then none of the lost blocks can be recovered.
This is the minimum total number of blocks per frame during the last polling period. If all blocks were received for all frames then this number is the nominal number of original blocks plus FEC blocks (Green lock icon). In our example this is 128+8 = 136.
Maximum number of FEC blocks used for original blocks recovery during the last polling timeframe. Ideally this should be 0 when no blocks are lost but the system is able to correct lost blocks up to the nominal number of FEC blocks (Neutral lock icon).
This counter counts the unrecoverable error conditions found (i.e. 4.4 between 128 and 128 plus the number of FEC blocks) since the last counters reset.
When the return key is hit within the address (5.2) or port (5.3) the changes are effective immediately. You can also use this button to set again these values. Clicking on this button will send a request to the API to get the remote SDRangel instance information that is displayed in the API message box (8)
When the return key is hit within the interface address (8.2), port (8.3), multicast group address (10) and multicast group join/leave (9) the changes of parameters for data reception are ready for commit and this button turns green. You then push this button to commit the changes.
Use this toggle button to join or leave the multicast group which address is specified next (10). Effective when the validation button (8.3) is pressed.
<h3>10: Multicast group address</h3>
This is the address of the multicast group. Effective when the validation button (8.3) is pressed.
