Abstract
The concept of hybrid in space-time Cellular Automata is introduced, for the first time, in this paper, and it is suggested that non-linear hybrid in space-time Cellular Automata can be used as pseudorandom pattern generators for VLSI systems, because they can produce patterns with various densities of “1”, distributed at will in space and time. The cycle lengths of non-linear hybrid Cellular Automata can be estimated using Lyapunov exponents.
Highlights
Pseudorandom pattern generators are mainly used for the generation of input test vectors for a VLSI system under test
The concept of hybrid in space-time Cellular Automata has been introduced, for the first time and it is in this paper suggested that non-linear hybrid in space-time Cellular Automata can be suitable for pseudorandom pattern generators for VLSI systems, because they can produce patterns with various densities P(1) distributed at will in space and time
The cycle length of non-linear hybrid in space-time Cellular Automata depends on the initial global state and the time period in which the second rule is activated
Summary
Pseudorandom pattern generators are mainly used for the generation of input test vectors for a VLSI system under test. Pseudorandom Number Generators (PRNG’s) based on Linear Cellular Automata have been found to generate more random patterns than those generated by Linear Feedback Shift Registers (LFSR’s). Cellular Automata do not have long feedback loops, and they operate at higher speeds. It has been shown that only by the rules 90 and 150, linear Cellular Automata can generate exhaustive and pseudoexhaustive test patterns. Linear Cellular Automata have been extensively investigated, but a little effort has gone into the non-linear. Cellular Automata, they generate more random patterns and have more complex behaviour than the linear ones [8]. Knowledge non-linear hybrid Cellular Automata have never been investigated.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have