With the integration of information technologies, power system operations are increasingly threatened by cyber-attacks. It has even been revealed that an attacker can inject false data into real-time measurements stealthily without knowing the full configuration (e.g., network topology) of a power system. In this paper, we present a comprehensive review on false data injection attacks which utilize barrier conditions, blind identification techniques and data driven approaches to overcome limitations of incomplete network information. We also point out future research topics for facilitating the detection and prevention of such false data attacks.