The marine traffic situation complex network (MTSCN) represents each ship in a monitored water area as a node and connects two ships by a weighted edge whenever their relative motion is converging, the edge weight combining pairwise density and conflict contributions. Built from Automatic Identification System (AIS) data for a single busy section of the Yangtze River (Zhenjiang Dagang), the source study reported that degree and vertex-strength distributions shift from scale-free (hub-dominated) to uniform as the aggregated situation-complexity index rises, and left two questions open: whether this construction, its density/conflict parameterizations, and the observed structural transitions hold across other waterways, traffic regimes, seasons, and AIS data-quality conditions; and how network-derived indicators should be operationally integrated into live traffic supervision. Subsequent work has produced improved variants of the construction - Voronoi-based, rule-based directed, directed, and multi-layer networks - together with node-importance and traffic-pattern methods for VTS-operator support, but these remain single-basin case studies. The systematic multi-area validation and sensitivity analysis the source requested, a replication of the scale-free-to-uniform transition in other waters, and deployment in live traffic supervision with real-time computation and alarm thresholds have not been reported. Both directions therefore remain open, with the operational-integration direction substantially advanced but unfinished.

Background

Marine traffic in busy waterways is supervised by Vessel Traffic Services (VTS), whose operators (VTSOs) watch large numbers of moving ships and must decide which vessels need attention or intervention. The principal data source is the Automatic Identification System (AIS), the shipborne VHF broadcast system by which vessels report identity, position, speed, and course at short intervals. Classical situation descriptors - ship count, traffic density, average speed - are scalar and local, and cannot distinguish situations in which the same number of ships produces very different multi-ship interaction structures. This motivated network-based treatments of 'marine traffic situation': earlier work by the same group had defined a pairwise traffic-complexity model from relative distance, relative speed, and intersecting trajectory, interpolated to an area complexity (Wen et al., 2015), and a parallel line combined density complexity with a multi-vessel collision-risk operator for regional near-miss assessment (Zhang et al., 2019).

The source study (Sui et al., 2020) introduced the marine traffic situation complex network (MTSCN): an undirected weighted graph whose vertices are the ships in a monitored area, with an edge between ships ii and jj exactly when their relative motion is converging, i.e. when the approaching rate

Rij=Dij→⋅Vij→∥Dij→∥R_{ij}=\frac{\overrightarrow{D_{ij}}\cdot\overrightarrow{V_{ij}}}{\left\|\overrightarrow{D_{ij}}\right\|}

is negative, where Dij→\overrightarrow{D_{ij}} is the relative position vector and Vij→\overrightarrow{V_{ij}} the relative velocity vector. Each edge carries a pairwise complexity weight, complexityij=denij+confij\mathrm{complexity}_{ij}=\mathrm{den}_{ij}+\mathrm{conf}_{ij}, the sum of a density contribution depending on ship separation and a conflict contribution depending on relative velocity and the track-crossing angle θij\theta_{ij}. From the weighted network the study derived node-level and area-level indicators: the vertex strength Si=∑jwijS_i=\sum_{j}w_{ij} (total interaction weight of ship ii); the per-node unit weight Ui=Si/kiU_i=S_i/k_i with kik_i the degree (number of interaction partners); the regional situation complexity M=∑iUiM=\sum_i U_i; and the network structure entropy E=−∑iIiln⁡IiE=-\sum_i I_i\ln I_i with Ii=ki/∑jkjI_i=k_i/\sum_j k_j, a normalized measure of how evenly influence is distributed over ships (larger EE means more homogeneous influence). Analyzing AIS snapshots of the Zhenjiang Dagang section of the Yangtze River (13 June 2018, observation intervals of 6, 9, and 12 minutes, two sub-areas), the study reported that when the computed complexity MM is low the degree and vertex-strength distributions follow power laws - the network is scale-free, meaning a few hub ships dominate the interactions - whereas as MM grows the distributions flatten toward uniform, many ships interacting with similar strength; the entropy EE was found to grow with MM, and clustering-coefficient behavior was consistent with this picture.

The study was thus a single-site empirical demonstration, and its authors left two directions open. First, geographical and dataset generalizability: whether the MTSCN construction, the parameterizations of denij\mathrm{den}_{ij} and confij\mathrm{conf}_{ij}, and the observed scale-free-to-uniform structural transitions hold across other waterways and sea areas, traffic regimes, seasons, and AIS data-quality conditions, which requires comprehensive multi-area validation and sensitivity analysis. Second, operational integration: how MTSCN-derived node-level and network-level indicators should be used in live traffic supervision and decision-making, including real-time computation, human-interpretable visualization, thresholds or alarms tied to control actions, and evaluation of how controllers or collision-avoidance systems should use the outputs to prioritize interventions.

Problem Statement

The problem, as stated by the source study and left open, has two linked parts serving one objective - establishing the MTSCN situation-complexity framework as a reliable and usable instrument for waterborne traffic management:

(1) Geographical and dataset generalization. Determine whether the MTSCN construction - the approaching-rate edge criterion, the pairwise weight complexityij=denij+confij\mathrm{complexity}_{ij}=\mathrm{den}_{ij}+\mathrm{conf}_{ij} with its density and conflict parameterizations, and the derived indicators MM and EE - and its reported structural behavior (scale-free degree and vertex-strength distributions at low situation complexity, tending toward uniform distributions as complexity rises) hold across other waterways and sea areas, different traffic regimes and seasons, and varying AIS data-quality conditions; these dimensions describe the class of settings the validation must span rather than a per-item checklist, and this requires comprehensive multi-area validation and sensitivity analyses of the parameterizations.

(2) Operational integration. Develop and evaluate procedures for applying MTSCN-derived node-level and network-level indicators in live traffic supervision and decision-making. Such procedures should cover, for example, real-time computation, human-interpretable visualization, thresholds or alarms tied to control actions, and assessment of how controllers and collision-avoidance systems should use the outputs to prioritize interventions; these facets illustrate the procedure scope rather than defining a mandatory checklist. This part is bounded to an evaluated end-to-end prototype on AIS data; a live operational deployment is not required.

Later literature has replaced the original construction with improved variants rather than validating it, so the generalization question is properly posed over the MTSCN family of situation-complexity networks, with validating the original construction as the required core of part (1).

References

  1. Sui, Z., Wen, Y., Huang, Y., Zhou, C., Xiao, C., Chen, H. (2020). Empirical analysis of complex network for marine traffic situation. Ocean Engineering 214:107848. DOI: 10.1016/j.oceaneng.2020.107848. https://doi.org/10.1016/j.oceaneng.2020.107848
  2. Sui, Z., Wen, Y., Zhou, C., Xiao, C., Zhang, F., Huang, L., Chen, H. (2020). Empirical analysis of complex network for marine traffic situation. In: Proceedings of the International Seminar on Safety and Security of Autonomous Vessels (ISSAV) and European STAMP Workshop and Conference (ESWC) 2019. De Gruyter, pp. 24-36. DOI: 10.2478/9788395669606-003. https://doi.org/10.2478/9788395669606-003
  3. Sui, Z., Huang, Y., Wen, Y., Zhou, C., Huang, X. (2021). Marine traffic profile for enhancing situational awareness based on complex network theory. Ocean Engineering 241:110049. DOI: 10.1016/j.oceaneng.2021.110049. https://doi.org/10.1016/j.oceaneng.2021.110049
  4. Sui, Z., Wen, Y., Huang, Y., Zhou, C., Du, L., Piera, M. A. (2022). Node importance evaluation in marine traffic situation complex network for intelligent maritime supervision. Ocean Engineering 247:110742. DOI: 10.1016/j.oceaneng.2022.110742. https://doi.org/10.1016/j.oceaneng.2022.110742
  5. Sui, Z., Wen, Y., Zhou, C., Huang, X., Zhang, Q., Liu, Z., Piera, M. A. (2022). An improved approach for assessing marine traffic complexity based on Voronoi diagram and complex network. Ocean Engineering 266:112884. DOI: 10.1016/j.oceaneng.2022.112884. https://doi.org/10.1016/j.oceaneng.2022.112884
  6. Zhang, F., Liu, Y., Du, L., Goerlandt, F., Sui, Z., Wen, Y. (2023). A rule-based maritime traffic situation complex network approach for enhancing situation awareness of vessel traffic service operators. Ocean Engineering 284:115203. DOI: 10.1016/j.oceaneng.2023.115203. https://doi.org/10.1016/j.oceaneng.2023.115203
  7. Tong, Y., Zhen, R., Dong, H., Liu, J. (2023). Identifying influential ships in multi-ship encounter situation complex network based on improved WVoteRank approach. Ocean Engineering 284:115192. DOI: 10.1016/j.oceaneng.2023.115192. https://doi.org/10.1016/j.oceaneng.2023.115192
  8. Xin, X., Yang, Z., Liu, K., Zhang, J., Wu, X. (2023). Multi-stage and multi-topology analysis of ship traffic complexity for probabilistic collision detection. Expert Systems with Applications 213:118890. DOI: 10.1016/j.eswa.2022.118890. https://doi.org/10.1016/j.eswa.2022.118890
  9. Cheng, X., Sui, Z., Wen, Y., Han, D. (2024). An improved maritime traffic situation complexity model for intelligent maritime management in the inland ferry area. Computers and Electrical Engineering 119:109612. DOI: 10.1016/j.compeleceng.2024.109612. https://doi.org/10.1016/j.compeleceng.2024.109612
  10. Xiao, Z., Sun, Q., Feng, L., Wang, X., Lu, X. (2025). A novel method for complexity analysis of marine traffic based on complex networks. Proceedings of the Institution of Mechanical Engineers, Part M: Journal of Engineering for the Maritime Environment 239:899-923. DOI: 10.1177/14750902251330035. https://doi.org/10.1177/14750902251330035
  11. Sui, Z., Wang, S., Wen, Y. (2025). A multi-layer complex network approach for evaluating ship behaviour complexity. Transportmetrica A: Transport Science (online first, 1-41). DOI: 10.1080/23249935.2025.2483800. https://doi.org/10.1080/23249935.2025.2483800
  12. Ji, Z., Zhang, Y., Wang, F., Wang, S., Yang, J. (2026). A dual-layer multilevel complex network approach for maritime traffic situation envelope assessment towards autonomous ship navigation. Ships and Offshore Structures (online first, 1-18). DOI: 10.1080/17445302.2026.2713182. https://doi.org/10.1080/17445302.2026.2713182
  13. Xin, X., Liu, K., Liu, J., Wang, W., Yang, Z. (2025). Modeling, Evaluation, and Mitigation of Maritime Traffic Complexity in Complex Waters. IEEE Transactions on Intelligent Transportation Systems 26:13275-13292. DOI: 10.1109/TITS.2025.3584705. https://doi.org/10.1109/TITS.2025.3584705
  14. Zhang, L., Ge, J., Goerlandt, F., Du, L., Chen, T., Gu, T., Gan, L., Li, X. (2025). A Method for Enhancing the Traffic Situation Awareness of Vessel Traffic Service Operators by Identifying High Risk Ships in Complex Navigation Conditions. Journal of Marine Science and Engineering 13:379. DOI: 10.3390/jmse13020379. https://doi.org/10.3390/jmse13020379
  15. Wen, Y., Huang, Y., Zhou, C., Yang, J., Xiao, C., Wu, X. (2015). Modelling of marine traffic flow complexity. Ocean Engineering 104:500-510. DOI: 10.1016/j.oceaneng.2015.04.051. https://doi.org/10.1016/j.oceaneng.2015.04.051
  16. Zhang, W., Feng, X., Qi, Y., Shu, F., Zhang, Y., Wang, Y. (2019). Towards a Model of Regional Vessel Near-miss Collision Risk Assessment for Open Waters based on AIS Data. Journal of Navigation 72:1449-1468. DOI: 10.1017/S037346331900033X. https://doi.org/10.1017/S037346331900033X