RF24Ethernet w/lwIP - Reaching a Point of Long Term Stability
Stable & Secure Communications with the RF24 Communication Stack
Since I started on this journey integrating the lwIP TCP/IP stack into the RF24 Comm Stack, I was unsure if I would ever get to this point. Its been about 6 months since I started, and at this point I can say I am glad I did. Using lwIP really brings the stack up to standard, as previously it used only the uIP TCP/IP stack, which works well, but is no longer maintained. Now, newer micro-controllers will use the newer, up-to-date lwIP stack, while smaller, older MCUs will use the uIP stack.
Over the past week or so, I've put in quite a lot of work on the RF24 Comm. Stack, including RF24Ethernet. In that time, it has become more secure, more stable, and is working better than it has in a long, long time.
Constant attacks & jamming targeting my networks has been taking place for years, and with the rise of AI, I decided to do some security & code reviews involving the RF24 Comm Stack. This was hugely beneficial, since I was able to work out a number of bugs and security flaws affecting the system. I would assume that most open source projects are beginning to do this as well, since there is a huge benefit from having a computer scan and verify your code.
In the end, I was able to fix what I see as the "Final Major Issues" affecting the communication stack, from buffer overflows, to memory leaks, to a problem with millis() timer usage, involving rollover un-safe millis() calls. The system is now more resilient and stable than ever before, still susceptible when jamming overwhelms parts of the system, but able to bring itself back online, recovering all functionality as soon as the jamming stops. The constant attacks against my network have slowed significantly, with very little to no impact on the system when they do occur, since there are now less if any avenues to exploit remotely.
I am now just running some final tests regarding the interaction of RF24Ethernet w/RF24Network when using lwIP, since there seemed to be some timing related issues with delay() vs network.update() calls, which should be addressed in the RF24Network master branch, but require testing before a release is made. These issues cause WDT resets on my ESP8266 while using WiFi in testing, but only occur once every 3-4 days or more, so tracking them down was difficult.
All in all, I am very happy with the progress made over the last week, satisfied that my security concerns have been addressed, functionality is back to where it should be, and everything just keeps working.
