µxrce-dds: add parameter services to get and set parameters

[azerupi]

Solved Problem

At the moment if you want to get and set parameters from a companion computer the only real interface you have is MAVLink. There is no convenient way to change parameters over µXRCE-DDS from ROS 2.

Solution

This PR adds two new services, one to get and one to set parameters over µXRCE-DDS either by name or by index.

To achieve this, 4 new versioned messages where created that are used in two service definitions. The µXRCE-DDS services follow the implementation of the Vehicle Command service closely but differ because they don't need to provide a response async.

One of the inconvenient parts is that we have to use char[16] for the name and can't use string as far as I know. We would need to add a wrapper around this on the ROS 2 side in e.g. px4-ros2-interface-lib.

If you want to test this, I have a simple ROS 2 python package to test getting a parameter: px4_parameter.zip

Alternatives

Instead of using services we could use topics directly with the messages added here. It is unclear to me if that would have a positive performance impact when getting or setting a lot of parameters at once. With the services from my Laptop with a Python test node, connected to a Pixhawk 6X over ethernet I'm seeing about 15ms round-trip on average while a ping is 1.2 - 1.7ms and I've measured the µXRCE-DDS packet round-trip to be between 5 and 11ms in a wireshark capture.

ros2 run px4_parameter px4_parameter_get
Starting benchmark for parameter 'MPC_XY_VEL_MAX' with 500 iterations...
Progress: 500/500

========================================
Benchmark Results (500 samples)
========================================
Average Time:        15.1521 ms
Standard Deviation:  5.4943 ms
Min Time:            5.3019 ms
Max Time:            36.4480 ms
========================================

Open questions

• I haven't looked at how to implement this for Zenoh. I don't think Zenoh should block this PR but we should probably have a path to get feature parity in follow-up PRs.

@beniaminopozzan we discussed this feature at the PX4 dev summit, not sure if this is exactly what you had in mind. I'm open to feedback on how to improve this.

[azerupi]

Also at some point we probably should look into how we can consolidate the 3 (or more?) different messages to change parameters. The ones I know of are:

• The ones used by ModalAI
• The ones for UAVCAN
• The ones introduced in this PR