BG6UFA Based on my practical experience, designing an APRS node doesn't offer much room for theoretical research. As a capstone project in engineering, the main challenges lie in the complex details of the protocols, and the limited support documentation makes it difficult to avoid pitfalls. Furthermore, building a complete system from the underlying hardware up is also a significant challenge from an engineering perspective.
Regarding your statement that "audio signal modulation and demodulation" is the "most critical" part, I don't quite understand. Is it because the transmitter maps 0/1 to 1200/2200 Hz audio frequencies, or is it because the receiver struggles to create a matching filter to distinguish between these two signals?