Method and Sensor Network for Feature Selection for Event Detection

Simple SummaryContent extracted from patent full text and abstract with AI.

This patent describes a method and sensor network system for improving event recognition in wireless sensor networks by intelligently selecting which sensor data (attributes) to use. Instead of transmitting and evaluating all available measured data, the system selects a relevant subset of attributes based on each sensor node's topological position and contribution. This optimizes the balance between accurate event detection and energy-efficient communication among sensors.

Use CasesContent extracted from patent full text and abstract with AI.

• Perimeter security systems (e.g., detecting attempts to climb or tamper with fences)
• Industrial installation monitoring (e.g., gas leak detection by distributed sensors)
• Environmental monitoring of remote or hazardous areas (e.g., forest fire detection via temperature sensors)
• Infrastructure monitoring (e.g., bridges, tunnels for structural integrity events)
• Smart building or facility monitoring for unauthorized intrusions or malfunctions
• Agricultural monitoring (e.g., distributed detection of weather or pest events)

BenefitsContent extracted from patent full text and abstract with AI.

• Reduces energy consumption and extends network lifetime by minimizing unnecessary data transmission
• Maintains or even increases event detection accuracy by selecting the most relevant sensor features
• Improves reliability by making event recognition less susceptible to communication errors or node failures
• Enables operation in large or complex sensor deployments (one- to three-dimensional topologies)
• Adapts to different event types and physical layouts by using topological information
• Decreases computational load on central systems by distributing analysis to sensor nodes

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

The invention relates to a method for attribute selection for an event recognition in sensor networks. The following steps occur in a configuration phase: providing a quantity (120) of attributes by sensor nodes (1-6) of a sensor network that characterize an event to be recognized, together with information for the topological origin of the attributes within the sensor network, and selecting (130) a sub-quantity from the quantity (120) of attributes, wherein the selection is made in consideration of the information on the topological origin of the attributes. In an execution phase, an event recognition occurs for an event to be currently recognized on the basis of the attributes that belong to a selected sub-quantity.