Abstract: It was found in this paper that the sensor node model used in the previous underwater sensor network simulation has the defects that are difficult to overcome. Boolean model has sacrificed more coverage performance to enhance the efficiency of simulation calculation, while continuous probability model(CPM) has given up the computational efficiency to approach the actual coverage performance simulation.In order to find a sensor node model which can not only meet the need of the covering performance but improve the calculation efficiency, the paper first proposes the concept of perception factor (PF) from the point of view of the effective detection of targets; and then puts forward a sensor node discrete model based on PF. After that, it is hypothesized that the nodes can be divided without limit; discrete perceptual factor model (DPFM) is extended to continuous perceptual factor model (CPFM), and the validity of CPFM is proved. Finally,CPFM is further optimized for multi ring perceptual factor model (MRPFM) to avoid the calculation of the logarithm.In this paper, the coverage performance and computational efficiency of MRPFM are simulated and compared with other models. The same model parameters and environmental parameters are set up by simulation.First, the coverage performance of CPFM in the condition of positive triangle, regular and regular hexagon deployment is simulated. It is found that the best way is not a regular hexagon deployment, but a positive triangle deployment.Then, this paper simulates the positive triangle optimal deployment of MRPFM and continuous probability model, and compares the coverage performance of the three models. It is obvious that the performance of MRPFM is a bit lower than that of continuous exponential probability model(CEPM), but significantly higher than Boolean model;Finally, this paper simulates the time of the deployment of Boolean model, CEPM, multi ring probability model(MRPM) and MRPFM, based on different node numbers. The simulation map clearly shows that, compared with using CEPM, the time to complete the deployment calculation of using the MRPFM is obviously reduced, and the more the number of nodes, the more obvious the effect of time reduction; and compared with using Boolean model and MRPM, the time is also reduced to some extent. Therefore, the MRPFM has practical value in underwater sensor network simulation