Sensor deployment is one of the key issues in wireless sensor networks (WSNs). An optimal placement of sensors is propitious to the maximum possible utilization of the available sensors, balancing node energy consumption, and prolonging the network lifetime. In this paper, we consider the mobile sensor deployment optimization for k-coverage of sensor networks with limited mobility. Since a mobile sensor’s cost is higher than that of a static sensor, we want to allow partial sensors to move to fill the coverage holes and vacancies. Through the analysis of the node distribution of randomly deployed sensor networks, we present the density of the vacancy for k-coverage of sensor networks, and the number of mobile sensors required to move to fill the coverage holes and vacancies. Moreover, since sensor mobility consumes much more energy than sensing and communication, we restrict sensors to move in a limited mobility model, i.e., the disk-based mobility model, which can reduce the sensor moving distance greatly. In this paper, we attempt to improve k-coverage of the mobile sensor networks using the improved particle swarm optimization algorithm under limited mobility. Simulation results show that few mobile sensors in the disk-based mobility model can realize the k-coverage, which reduces the cost of the sensor networks and moving distance of mobile sensors.