Dual Compressive Sensing for Wireless Sensor Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional distributed compression technologies in wireless sensor networks require prior knowledge of correlation between sensor nodes and are not suitable for low-complexity sensor nodes with high operation throughput, and existing compressive sensing methods require additional information to determine active sensor nodes, increasing complexity and traffic.
Innovation Solution
A dual compressive sensing method using Gaussian codes and measurement matrices to transmit and receive signals, where sensor nodes determine operation based on allocated codes, and a fusion center recovers signals using these codes without additional advance information, reducing complexity and traffic by performing compressive sensing twice.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional distributed compression technologies (entropy coding, Slepian-Wolf coding) are used, then compression can be achieved, but prior knowledge of correlation between sensor nodes is required and device complexity increases
Solution Approach 1:
The patent changes the fundamental parameter of compression approach from correlation-based coding to compressive sensing based on sparsity. Instead of requiring knowledge of inter-node correlation, the system exploits the sparsity of the signal in some domain, fundamentally changing how compression is achieved and eliminating the need for complex correlation analysis
Solution Approach 2:
The patent replaces the mechanical/coded compression systems (entropy coding, Slepian-Wolf coding) with a mathematical transformation approach based on compressive sensing. This substitution eliminates the need for complex coding algorithms and replaces them with simpler linear measurements and sparsity-exploiting reconstruction
2Loss of information
If conventional distributed compression technologies are applied, then compression is achieved, but additional information about correlation must be obtained in advance, increasing traffic
Solution Approach 1:
The patent extracts only the essential measurement data at sensor nodes without requiring extraction or transmission of correlation information. By taking out only the necessary compressed measurements and relying on sparsity at the fusion center, the system eliminates the need for additional correlation information traffic
Solution Approach 2:
The system makes the fusion center self-sufficient by enabling it to reconstruct signals using only the received measurements and sparsity knowledge. The fusion center serves itself by performing the reconstruction without needing additional correlation information from sensor nodes, eliminating the need for extra information exchange
3Ease of operation
If compressive sensing is used for distributed compression, then easy distributed compression is achieved, but additional information is required to determine active sensor nodes, increasing complexity
Solution Approach 1:
The patent merges the activity detection function with the compressive sensing measurement process. Instead of separately determining which nodes are active and then compressing their data, the system combines both functions into a single measurement step, where the measurement matrix inherently handles both compression and activity indication
Solution Approach 2:
The measurement matrix serves multiple functions simultaneously: it performs compression, indicates active sensor nodes, and enables signal reconstruction. This multi-functionality eliminates the need for separate mechanisms to determine active nodes, reducing overall system complexity while maintaining ease of operation
4Loss of information
If additional packets are transmitted to indicate sensor operation, then active node information is available, but network traffic increases
Solution Approach 1:
The measurement matrix acts as an intermediary that encodes active node information within the compressed measurements themselves. Instead of transmitting separate packets to indicate active nodes, the measurement matrix mediates by embedding this information in the measurement process, eliminating the need for additional traffic
Data Source
AI summary
A method for transmitting signals based on dual sensing in a wireless communication system is disclosed. One or more sensor nodes receive Gaussian codes corresponding respectively to the one or more sensor nodes, allocated from a fusion center. The one or more sensor nodes determine whether to operate at a specific time. At least one sensor node that has determined to operate among the one or more sensor nodes multiplies the Gaussian codes by a transmission signal and transmits the multiplied signal to the fusion center.