Case ID: M23-179P

Published: 2023-12-06 14:16:51

Last Updated: 1701872211


Arunabha Sen
Kaustav Basu

Technology categories

Advanced Materials/NanotechnologyEnergy & PowerPhysical ScienceWireless & Networking

Licensing Contacts

Physical Sciences Team

Sensor Network Design for Uniquely Identifying Sources of Contamination in Water Distribution Networks


Sensors are being increasingly adopted for use in smart cities in order to monitor various parameters. The use of sensors allows for any anomalous behaviors in the deployment area to be easily detected. Once sensors are deployed, they have two functions, one of which is sensing/coverage of target parameters including temperature, pressure, and vibration. The other function involves transmitting the sensed data either directly or through multiple other sensor nodes to the control station for analysis of the sensed data.

In recent years, several coverage models have been proposed that utilize the Set Cover-based problem formulation. However, this method lacks unique identification capability for the location where anomalous behavior is sensed. This can be overcome through the utilization of Identifying Code. The optimal solution of the Identifying Code problem provides the minimum number of sensors that will be needed to uniquely identify the location where anomalous behavior is sensed.

Invention Description

Researchers at Arizona State University have developed a novel budget constrained method for identifying sources of anomalies in water distribution systems of smart cities. This design includes an Integer Linear Programming (ILP) formulation that provides an optimal solution for the problem. Typical ILPs are computationally expensive, but this method showed shorter computation times during initial tests, even for graphs with more than 2300 nodes and 4800 edges.

Potential Applications

  • Sensor network design for smart cities (e.g., monitoring anomalous behavior, identifying sources of contamination)

Benefits and Advantages

  • Reduced computation time
  • Less expensive
  • Does not require specialized equipment (e.g., can be run on typical processors)