A Context-Aware Voronoi-Based Algorithm for Optimal Placement of Point-Based Sensor Networks for Spatiotemporal Field Monitoring

Accurately monitoring the spatio-temporal dynamics of environmental fields using point-based Wireless Sensor Networks (WSNs) is critical for applications such as urban microclimate monitoring and building energy assessment. However, existing deployment strategies often rely on oversimplified sensing models, discrete grid sampling, or neglect temporal variability, limiting their effectiveness in real-world sensor systems. This paper proposes an Extended Voronoi-based Optimal Sensor Deployment Algorithm (E-VOSDA) for point-based sensor networks that integrates spatial context and temporal variability in the target field. E-VOSDA leverages Voronoi diagram characteristics to manage the sensor network, assess error distribution, and enhance coverage through sensor insertion and movement strategies. It incorporates a dimensionality reduction mechanism to extract dominant temporal patterns. E-VOSDA’s effectiveness was evaluated through controlled experiments on synthetic heatmaps. The results confirm E-VOSDA’s reliability, particularly under diverse spatial constraints. When a preferred area constraint was designated, it achieved a lower RMSE using even fewer sensors compared to the ordinary VOSDA, and in the scenario with a restricted area constraint, it maintained performance while avoiding restricted zones. To assess spatiotemporal performance, E-VOSDA was also applied to a simulated temperature dataset generated by a microclimate model over Laval University campus. In this scenario, E-VOSDA optimized sensor deployment on a heatmap generated through temporal dimensionality reduction, achieving an RMSE of 2.49 with 20 sensors. When the same sensor configuration was applied to a given time epoch during the hottest week of the year, E-VOSDA achieved an RMSE of 0.49, demonstrating its practical effectiveness in capturing key temporal variations and adapting sensor placement to shifting spatial patterns.