A Trojan horse circuit detection and error recovery method when equipment is running

Abstract

The present invention proposes a Trojan horse circuit detection and error recovery method during device operation, which includes the following steps: establishing original calculation and recalculation for Trojan horse circuit detection, and recovery calculation when the Trojan horse circuit is triggered and makes an error; Bind task nodes to IP cores; calculate the number of IP core manufacturers required by IP cores; match the optimal IP core manufacturer combination; determine task scheduling time and the number of IP cores; construct positioning and Replace the structure; check in real time whether the original calculation and recalculation results are the same, if not, locate the IP core where the activated Trojan horse circuit is located; replace the IP core where the activated Trojan horse circuit is located. The method can ensure that an output error is detected after the Trojan horse circuit is triggered and provide recovery calculations so that the system can run normally, and at the same time it can quickly locate the infected IP core and replace it before the system runs next time.

Description

technical field [0001] The invention relates to the field of computers, in particular to a Trojan horse circuit detection and error recovery method when a device in an integrated circuit is running, and in particular to a Trojan horse circuit detection and error recovery method when an untrustworthy third-party IP core is running. Background technique [0002] The globalization of integrated circuit design and manufacturing has made chip security and reliability an important issue, and the detection of Trojan horse circuits in integrated circuit design has always been the top priority. [0003] There are four main detection methods for general Trojan horse circuits, which are physical detection, bypass analysis, functional testing, and trusted design. They are all non-runtime detection methods and do not provide recovery calculations. We analyze four Trojan horse circuit detection methods respectively. (1) Physical detection, which is essentially a technology based on fail...

AI Technical Summary

Problems solved by technology

To sum up, the existing four Trojan horse circuit detection mechanisms are not universally applicable, and do not provide error recovery calculations, and cannot meet the runtime requirements

[0004] At the same time, the infected IP core cannot be located and replaced in time

Images