Scan Chain-based Technique For Detection Of Aging ICs

The electronics industry has become a primary target in the global market of counterfeiting mainly including the following several types: recycling,overproduction,cloning and tampering.It is reported that more than 80% of counterfeit parts are recycled,such recycled parts have poor performance,short life span,and are more prone to accidental errors that can lead to system failures.What's more,if the electronic components are implanted with malicious virus during the recycling process,it will cause huge loss to our country's vital industries such as aviation,aerospace,medical,military and infrastructure.By detecting the recycled ICs,these problems can be avoided in time.From the perspective of market economy and system security,it is of great significance to identify the recycled ICs and develop technologies that can detect the aging of integrated circuits.There are four cast of methods to detect recycled ICs at present,the first one is physical and electrical tests,but this method can be made ineffective by simply forging the appearance,and this method is very time-consuming and expensive.And the second one is DNA marking,it can realize the traceability of electronic devices,but the scheme has a complex certification process and expensive testing costs,making it difficult to commercialize.The third statistical methods,they are proposed to extract the changes of some parameters before and after the circuit aging for analysis and modeling,so as to carry out active detection on the unknown chip.And the last one is on-chip sensors,they have been proposed as an alternative to the conventional test methods for efficient detection of these ICs.To our knowledge,we are the first to implement recycling detection at printed circuit board(PCB)level.We utilize a boundary scan architecture(BSA)-based path delay technique to identify recycled PCBs.We put forward the recycled PCB detection process including the registration and authentication stage: In the stage of registration,we obtain a large number of fresh boundary scan architecture(BSA)delay information and store it in tamper-proof memory inside the PCB,and then measure the delay information after different days of aging.We calculate degradation rate of the delay time before and after PCB aging,and determine optimal threshold through intensive data analysis.In the authentication phase,compare the original delay information with the current BSA delay information.If the degradation of path delay is greater than threshold,the PCB will be detected as recycled;otherwise,it will be marked as new.This experiment is performed on HSPICE simulation platform.Spice language is used to describe the structure of the boundary scan chain,Monte Carlo analysis is used to simulate the manufacturing process error of the scanning unit,and the reliability analysis model(MOSRA)is added to simulate the aging of the circuit.Simulation results show that the proposed mechanism can accurately detect the recycled circuit board and the overhead required is only non-volatile memory for storing a small amount of delay information.In order to enlarge the application scenario,we propose a new method of recycled chip detection based on the time delay variation between normal stressed circuit structure and reference circuit structure.It is mainly divided into two parts: aging sensor and authentication mechanism.We design the normal stressed circuit-A unit whose structure is a single D flip-flop and some combinational logic circuit,and a similar reference circuit-ref unit structure as circuit-A unit,as well as a SR-latch as an arbiter to compare circuit-A and circuit-ref's transmission speed.We also introduce the authentication module to generate the unlocking signal of the reference circuit clock.When the user enters the correct key,the authentication module will generate the ‘ACT' signal to the clock port of the circuit-ref,thus the circuit-ref can perform the comparison analysis.According to the HSPICE simulation experiment data,the functional relationship between the aging time and the aging rate(ar)is fitted.Simulation results show that the proposed mechanism can accurately detect the recycled IC.The overhead is very small,which can be ignored in VLSI.In conclusion,in this topic,we first put forward the BSA-based structure to detect recycled PCB.Next,we proposed an aging sensor to detect recycled ICs,the designs both have the very good performance in safety and cost,in the future work,we will explore the use of machine learning and classification methods to estimate the aging time of PCB and IC.Applying more excellent machine algorithms to the field of hardware security will be the focus of future research.