The receiver autonomous integrity monitoring(RAIM)technology,which with the advantage of quick response and completed independent,plays an important role in the integrity monitoring system of GNSS.With the development of BDS and other navigation satellite systems,the dual-mode and the multi-mode satellite navigation receivers are paid more and more attention,which provide not only superior positioning accuracy,but also sufficient redundancy for RAIM.Among the fault detection algorithms,the "snapshot" detection methods show satisfied rapidity for detecting large fault in pseudorange measurement,while their abilities to detect a minor fault is poor.Although minor faults usually not serious enough to cause unavailable positioning of GNSS receiver immediately,but their influence on positioning error accuracy will accumulate with time to a significant level.In order to overcome the shortcoming of the traditional RAIM algorithm in detecting minor faults,an improved minor fault detection algorithm based on sliding-window accumulated parity vector is proposed in this paper.In the proposed algorithm,a new statistics for fault detection is obtained by accumulating the parity vector in a sliding window,so as to make an optimized fault diagnosis results online.A simulation system is established for validating the proposed algorithm.The simulation results show that the proposed algorithm could shorten the time delay in detecting minor faults.