A Simulator Driven by Trajectory Big Data for Network Feature Extraction and Data Transmission

Due to the rapidly growing communication require-ments among mobile real-world communication devices, e.g., smartphones, vehicles, and vessels, and the resulted exponentially increasing complexity of network structures, network simulators have attracted increasing attention for the design, analysis, and evaluation of network protocols in both academia and industry. However, existing simulators fail to integrate critical function modules of network feature extraction and visualization. While significant learning burden is inevitable but unbearable for users. In this paper, we propose a trajectory big data driven network simulator with interactive visualization for large-scale networking, feature extraction and data transmission. Our simulator consists of three key modules of trajectory data preprocessing, network feature extraction, and routing protocol evaluation. The integration of multiple function modules with intuitive performance visualization greatly enhances usage convenience. In particular, diverse social network features are available for users to achieve in-depth perception on practical networks and facilitate better design of realistic protocols. A novel social com-munity discovering algorithm dedicated for ocean vessel network is proposed as a supplement to land-based mobile networks. The developed simulator is extensively evaluated with a real dataset of vessel trajectories over East China Sea. Experimental results illustrate the feasible functions, step-by-step visualization, and convenient interfaces.