Abstract—One of the challenges of Mobile Ad Hoc Networks (MANET) is black hole attack. This is a form of destructive attack, causing very heavy damage to network performance once successfully implemented. By replying to the route with HC=1 and the largest SN, the malicious node fools the source node into thinking that it has the best and freshest cost-effective route to the destination node. As a result, all the data packets get caught in the malicious node and go missing without being able to reach the destination node. Most of the previous research was based on the characterization of black hole attacks or simple check mechanisms to detect cyberattacks. This leads to limitations that need to be overcome such as: Error rate in malicious node detection algorithm, routing waste, data routing efficiency in normal network scenario. This paper proposes a black hole attack detection algorithm (BDA) based on statistical theory. BDA collects information in real time so it can detect and prevent black hole attacks as they begin to act. The proposed solution uses a balance threshold value, calculated based on statistical theory, as the threshold for detecting black hole attack. A node that replies to the route with an SN value greater than the threshold is identified as a malicious node and isolated immediately upon attack. The article also proposes a black hole attack detection routing protocol (BDAODV) by improving the AODV protocol using BDA solution. The performance of the BDAODV protocol is evaluated and compared with related solutions on a network model with random mobile nodes. The simulation results have shown that the proposed protocol has very good performance in the network scenario under black hole attack with different number of malicious nodes.