Train safety monitoring and fault diagnosis are critically important because of the disastrous results caused by train collisions and derailments. Train safety protection sensors network is capable of autonomously monitoring the working condition and actively control faults. A number of strategically placed sensors in the vehicles form a network that can monitor various vital parameters and provide real-time prompt to train driver and dispatcher. Designing such networks faces a number of challenging tasks, as one needs to address some conflicting requirements for quick diagnosis, collaborative decision making, achieving high precision and reliability. This paper presents an on-line Train Safety Sensor Network (TSSN) architecture, discusses its hardware and software structure for ambulatory failure status monitoring. The network consists of multiple sensor layers that monitor train's electrical and mechanical activities, a train data center and a ground data analysis server. The server implements fault diagnosis based on a Fault Tree Analysis method (FTA). The results shows that the sensor network contributes to higher train safety guarantee.