Multi-hop wireless mesh networks (WMNs) are increasingly growing interest as a promising technology for high-speed network access. WMNs are integrated with other networks such as Internet, sensor networks, and cellular networks through gateways. Therefore gateways in WMNs are prone to various security threads and can be easily exploited by the attacker from any part of the Internet to launch a distributed flooding attack to compromise computer systems, affect the network performance, and destruct the services. Many approaches have been proposed by researchers in this regard but still more efforts are needed in terms of accuracy and efficiency. In this research paper, we will propose an artificial neural network-based technique for detection of distributed flooding attacks to things such as sensors or actuators in WMNs called the distributed flood attack detector. In our simulation, sample dataset used to train and test the artificial neural network is generated using NS-2 network simulator. Simulation results and real system implementation proved that the distributed flood attack detector can be used in a real network environment to detect the intermediate and severe distributed flood attacks with low-false positive and false-negative rates.