To protect the channel resources of automatic identification system (AIS), the paper proposes a margin optimization model of virtual AIS AtoN. The model is divided into two stages, in the first stage, the gray model is used to calculate the number of available time slots in the next period; in the second stage, the improved PSO-Non-dominated Sorting Genetic Algorithm-II (PSO-NSGA-II) is used to convert multi-objective optimization of time slot occupation and time slot resource conflict rate as the objective function to calculate the optimal margin of virtual AIS AtoN. The performance of the model is verified by comparing the simulated number of different ships with the real case of ships in and out of Shanghai port in 2023. The results show that this margin optimization model can reduce the time slot conflict rate of the system and balance the maritime resources between ships and virtual AIS AtoN.