Physical Unclonable Functions (PUF) for IoTs

IoTs play a pivotal role in digitising economy by collecting data, communicating with servers, and issuing instructions for action. Its growth, however, is not as fast as once predicted. One of the main constrains is security. The popular security protocol of network is software based and uses one-way mathematical functions, such as Rivest–Shamir–Adleman (RSA). This is acceptable for laptops and smart phones, which have adequate computing capability for its implementation. The main growth of IoTs in future, however, will be driven by low-cost edge units, such as sensors and actuators, and the connected logistics is forecasted as the fastest growth segment. Securing such low-cost IoTs is challenging: on one hand, their cost/power-restricted computing resource prevents running the full-fledged conventional security protocol; on the other hand, they are often mobile and physically unprotected. To overcome this challenge, intensive efforts have been made to use low-cost hardware security, such as the physical unclonable functions (PUF).  The success of PUFs, however, is limited, because of their weaknesses, including poor reliability, vulnerability to hacking, and difficulties in key renewal and revocation. The aim of this project is to overcome these diffculties through innovation, such as injecting dynamic entropy into PUFs. The project will use industrial grade test samples and will be in collaboration with industry, including companies specializing in PUF design. Liverpool John Moores University has the required test facilities and has worked on VLSI device qualification for over 30 years, in colloaboration with world lead orgnizations, such as IMEC.