Brain-like infrastructure for embedded SoC diagnosis

TL;DR

This paper presents a high-speed multiprocessor architecture inspired by brain functions for diagnosing embedded SoC systems by analyzing complex associative relations in high-dimensional vector spaces.

Contribution

It introduces a novel multiprocessor architecture and vector-logical process models for efficient analysis and diagnosis of embedded systems.

Findings

Effective analysis of associative relations in high-dimensional spaces

New integral non-arithmetical metric for solution quality estimation

High-speed processing capabilities demonstrated

Abstract

This article describes high-speed multiprocessor architecture for the concurrent analyzing information represented in analytic, graph- and table forms of associative relations to search, recognize and make a decision in n-dimensional vector discrete space. Vector-logical process models of actual applications,for which the quality of solution is estimated by the proposed integral non-arithmetical metric of the interaction between Boolean vectors, are described.