Review of Security with Noisy Data by Pim Tuyls , Boris Skoric and Tom

This book describes how to use biometric features such as finger prints or iris patterns to construct private keys for cryptography. It is a very interesting topic, as, nowadays, cryptographic protocols can store/produce thousands of bits of secure information, while the user needs a password to make the exchange secure. The major weakness in a cryptographic protocols is the user himself. Biometric features are methods to efficiently provide a large unique private key for a user. Unfortunately, the measurement equipment to acquire biometric data is inherently noisy and consequently a biometric matching operation is only approximate. The aim of this book is to study how the presence of noise has an impact on information security and how it can be dealt with.