The circuit prototyping unit (CPU): From theoretical science to device prototyping based on complex and nonlinear systems

Besides the theoretical research, the activities of the CPU focus on the electronic implementation of mathematical models of complex systems and algorithms for real-time analysis of real-world time series, as well as software development. It thus acts as an incubator for prototyping electronic devices and developing data encryption techniques and data analysis methods. To date, the CPU group has nonlinear oscillators and discrete-time systems such as Arnold’s Cat map and its large-period variants. Given the difficulty of experimentally analyzing the behavior of real-world systems, the group has developed and implemented on microcontrollers time series analysis methods such as permutation largest slope entropy (PLSE), matching energy (ME) and complexity from ordinal pattern positioned slopes (COPPS). A prototype device implementing PLSE has been produced and allows us to characterize the stability of a DC voltage source. This device can help detect faults in a charge regulator and thus prevent battery mischarging. COPPS has been used for the classification of signals, including heartbeats. The group has also developed a multimedia data encryption application using the modeled dynamical systems as pseudo-random number generators. The above application called JSApp allows secure data exchange over the Internet and can be easily deployed in an organization thanks to its hierarchical key management. In addition, the CPU group designs and implements intelligent systems such as alarms and motion detectors to optimize energy consumption.