A Local 3D Voronoi-Based Optimization Method for Sensor Network Deployment in Complex Indoor Environments.

Area of Science:

• Computer Science
• Geographic Information Science
• Robotics

Background:

• Sensor network deployment in built environments faces challenges with coverage gaps.
• Existing optimization methods often oversimplify complex 3D indoor spaces.
• Indoor applications like navigation and evacuation require precise sensor placement.

Purpose of the Study:

• To propose a novel local optimization algorithm for sensor network deployment in 3D indoor environments.
• To address limitations of empirical methods and oversimplified environmental models.
• To improve coverage and efficiency in sensor network applications.

Main Methods:

• Developed a local optimization algorithm utilizing a 3D Voronoi diagram for proximity analysis.
• Integrated the algorithm with an IndoorGML model for managing indoor environment components and sensor relations.
• Compared performance against Genetic Algorithm (GA) and Covariance Matrix Adaptation Evolution Strategy (CMA-ES).

Main Results:

• Achieved 98.86% coverage, comparable to GA and CMA-ES.
• Demonstrated approximately six times greater computational efficiency.
• Successfully defined proximity relations in complex 3D indoor settings.

Conclusions:

• The proposed 3D Voronoi diagram-based algorithm offers an efficient and effective solution for sensor network deployment in 3D indoor environments.
• This method overcomes limitations of existing approaches by considering architectural complexity.
• The integration with IndoorGML enhances management of indoor spaces and sensor networks.