In order to solve the problems that the traditional aquatic water quality monitoring system could not give early warning and the communication delay of the system was high, a low delay wireless sensing water quality monitoring platform with dissolved oxygen prediction was proposed. In this study, four water quality monitoring nodes were set up, and the real-time monitoring of water quality data was realized through the communication between the LoRa module and the convergence module. To reduce the system delay, the strategy of edge computing is applied to offload the computing and system control tasks of cloud server to the upper computer, which updates the data of local and cloud server, and realizes the prediction function of dissolved oxygen based on the wavelet transform and the Long Short-Term Memory (WT-LSTM) model. The experimental results show that, compared with other prediction models, WT-LSTM model has better effect; the relative errors of pH, temperature, dissolved oxygen, conductivity and ammonia nitrogen in monitoring data are less than 1.4%, 0.7%, 0.2%, 12% and 5%, respectively;based on the evaluation coefficient analysis, the prediction result of dissolved oxygen at 1 h is more accurate, which can be used as a reference for dissolved oxygen early warning. The platform can realize the real-time monitoring of water quality data and complete the prediction of dissolved oxygen in 1 h under the condition of low cost and low delay, which makes the control of the aerator more reasonable and intelligent.