The intrusion detection system based on feature selection deals with huge amount of data which contains redundant and noisy features causing slow training and testing process, high resource consumption as well as poor detection rate. Feature selection, therefore, is an important issue in intrusion detection and it can delete redundant and noisy features. In order to improve performances of intrusion detection system in terms of detection speed and detection rate, a survey of intrusion detection system based on feature selection is necessary. This paper introduces the concepts and algorithms of feature selection, surveys the existing lightweight intrusion detection systems based on feature selection algorithms, groups and compares different systems in three broad categories: filter, wrapper, and hybrid. This paper concludes the survey by identifying trends of feature selection research and development in intrusion detection system. Feature selection is not only useful for intrusion detection system, but also helpful to provide a new research direction for intrusion detection system.