A Platform for Adaptive Interference Mitigation and Intent Analysis Using OpenLANE

Researchers have long studied cognitive processes, particularly through analyzing physiological signals. This work has informed the development of human-machine interfaces, crucially expanding interaction opportunities, especially for disabled individuals, enabling them to operate devices like exoskeletons. EEG technique has been verified as a useful and safe method due to its relatively high temporal resolution and non-invasive characteristics. At this stage, however, the field is facing many practical challenges and there are gaps emerging in deploying a system to make the most of physiological signals, as the foresight of increasing spatial resolutions and the requirements of resistance to environmental noise as well as adaptive capability to variant context are proposed, which lead to extreme computing burden and power consumption. In response to these issues, after exploring related fields we propose a trade-off approach with four characteristics inspired by EMD and HDC. This approach tries to utilize the EMD decomposition property under simulated non-stationary interference or stationary interference to target the interference problem and contribute to reconstructing the valid information latent in the original signal. Meanwhile, it inherits the HDC theory's properties of normal distribution of mapping information and simplicity of mapping computation, leading to fault-Tolerance and less computation, increases the number of channels that can be captured, and helps to better meet the increasing spatial resolution requirements with the same total computational resources. For cost-efficiency and time, the total system wants to take an open-source EDA as routing and layout support and finally be elaborated to a tape-out-ready GDSII. © 2023 IEEE.