SOFTWARE IMPLEMENTATION AND SIMULATION MODELING OF A MODIFIED MEASUREMENT ALGORITHM FOR SENSOR SYSTEMS

Koibash Aleksandr Andreevich

The paper presents research results on applying postbinary computing to sensor system data processing. The objective is the theoretical substantiation of postbinary encoding applicability for representing measurements with uncertainty and creating a demonstration software implementation on modern binary platforms. A modified algorithm for sensor data processing is described: interval symmetrization relative to the expected value followed by postbinary encoding. Two formats are analyzed: PostBinary 64/32 p (minimal memory consumption) and PostBinary 128/32 ip (unconditional accuracy of interval boundaries). An object-oriented C++ library with encoding/decoding procedures has been developed. Simulation modeling of sensors with various error levels revealed format characteristics: 64/32p is effective for accurate sensors under stable conditions, while 128/32ip is preferable for large errors and sharp changes. The results establish a theoretical foundation for the sensor component of computational infrastructure compatible with prospective postbinary architectures. The library is applicable in research projects studying postbinary data representation.

Keywords:post-binary computing, sensor systems, tetralogic, sensor data processing, four-valued logic