Computing complex functions using factorization in unipolar stochastic logic

This paper addresses computing complex functions using unipolar stochastic logic. Stochastic computing requires simple logic gates and is inherently fault-tolerant. Thus, these structures are well suited for nanoscale CMOS technologies. Implementations of complex functions cost extremely low hardware complexity compared to traditional two's complement implementation. In this paper an approach based on polynomial factorization is proposed to compute functions in unipolar stochastic logic. In this approach, functions are expressed using polynomials, which are derived from Taylor expansion or Lagrange interpolation. Polynomials are implemented in stochastic logic by using factorization. Experimental results in terms of accuracy and hardware complexity are presented to compare the proposed designs of complex functions with previous implementations using Bernstein polynomials.