To address the issues such as wide coverage and limited power, a real-time network status collection and fault diagnosis system is required for the electrical wireless private network. In this paper, the fixed terminals are utilized as the network status acquisition devices to monitor the key performance indicators of the network in a real-time manner, and a statistical analysis framework for network fault is proposed based on the data gathered by the minimizing driving test. And then, this paper also proposed a fault intelligent diagnosis model based on softmax neural network. The proposed statistical analysis framework can effectively identify the deterioration of network, and the fault identification accuracy is over 89%. The proposed system can effectively improve the timeliness of network fault diagnosis, and as a result, reduce the cost of network operation and maintenance, which makes a practical guiding sense.
