This paper proposes a faulty node detection scheme with a hybrid algorithm using a Markov chain model that performs collective monitoring for naval combat systems(NCS). With regard to distributed systems such as NCS, it is highly likely that a system's workload may be unfairly distributed among the master node and slave nodes. Therefore, when the controller executes an operation in a tactical environment, the concerned system's master node may not easily detect the faulty state which slave nodes have bad performance to smooth tactic. To solve these problems, this paper proposes the possibility more reliable faulty node detection scheme in distributed systems with the use of Markov chain. The proposed schemes use the weighting interval factors, which is based on the probabilistic computation to enhanced Bose-Chaudhuri-Hocquenghem(BCH) code for reliable decision of faulty node detection. Before each slave nodes send master node to single BCH code, the Markov chain is applied to processes of each slave nodes, where the probability between two-state such as Good, Warning, and Bad-state is calculated based on independence in master and slave nodes using catalyst matrix. Simulation results show that the proposed method can improve the performance of the faulty detection reliability and miss-detection rate for real-time distributed systems.