This document proposes ShakeIn, a smartphone authentication system based on how users shake their phone with one hand. It extracts physiological and behavioral characteristics from accelerometer and gyroscope sensor readings during the shake to uniquely identify users. In a training phase, users perform a small number of shakes to establish an identification model. In the authentication phase, it verifies the user's identity by comparing a new shake to the model. Evaluation showed it achieved 1.2% equal error rate using only 35 training samples, even under shoulder surfing attacks. It aims to provide strong authentication that is also convenient for users.