Random value generator

Abstract

The present invention discloses a circuit for generating a random output value which comprises: a bistable circuit having two stable states in which a 0 or a 1 is output and having a balanced metastable state in which a floating value between 0 and 1 is output, said bistable circuit resolving from said metastable state to one of said stable states on being switched on, said state depending on a voltage level at a port on said bistable circuit; a voltage level control circuit for controlling a voltage level at said port on said bistable circuit; a time measuring circuit for measuring a switching time taken for said bistable circuit to switch from said metastable state to one of said stable states following switch on; and control logic for controlling said time measuring circuit, said voltage level control circuit and a switching off and on of said bistable circuit, said control logic being adapted to perform a following sequence: control said voltage level control circuit to set a predetermined voltage level at said port on said bistable circuit, switch said bistable circuit on, detect a measured switching time, and turn said bistable circuit off and if said measured switching timeis longer than a predetermined value, output said resolved stable state of said bistable circuit as said random output value.

Description

technical field [0001] The field of the invention relates to data processing and, in particular, to the generation of random numbers. Background technique [0002] Generating truly random numbers is a difficult task. Different approaches have attempted to employ metastable events to generate random output values. None of them seem to be ideal, and tend to process the output to restore randomness when they find nonexistent. It is essentially a non-random process. For example, some use von Neumann correctors to eliminate a large number of 0s and 1s (see for example R. Brederlow et al. "Low Power True Random Number Generator Using Random Telegraph Noise of Single Oxide Traps" (ISSCC Dig. Tech. Papers, pp. 536-532, February 2006)). Such a correction serves only to mask the fact that the original bit stream was not truly random, and true randomness cannot be recovered. "Design of On-Chip Random Number Generator Using Metastability" (Proc European Solid-State circuit conf, pp...

AI Technical Summary

Problems solved by technology

However, this jeopardizes the system by producing correlation flows that effectively lack randomness

Images