Compressed Sensing in Wireless Sensor Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing compressed sensing (CS) methods for wireless sensor networks (WSNs) face challenges in effectively interplaying with routing topology to reduce transmission costs and designing a suitable representation basis for real-world signals, leading to high energy consumption and inefficiencies in data acquisition, especially in multi-hop large-scale deployments.
Innovation Solution
The proposed method employs routing topology tomography to determine dynamic routing paths and constructs a measurement matrix, combined with graph wavelets via deep learning to find an optimized representation basis, allowing for efficient data acquisition without storing the measurement matrix in sensor nodes, thus reducing energy consumption and improving data recovery fidelity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional CS methods use dense measurement matrices for data acquisition in WSNs, then measurement accuracy is maintained, but transmission costs and energy consumption increase significantly
Solution Approach 1:
The patent extracts and utilizes the routing path information that already exists in WSN data packets as the measurement matrix, eliminating the need for separate dense measurement matrices. This extraction of useful information from existing routing structures reduces transmission overhead and energy consumption while maintaining measurement capability through routing topology tomography
Solution Approach 2:
The routing paths in WSNs serve dual purposes: traditional data transmission and compressed sensing measurement. By making the routing topology multi-functional, the system uses existing data packets for both their original purpose and as measurement vectors, eliminating redundant transmissions and reducing energy consumption without sacrificing measurement precision
2Measurement precision
If CS methods store parts of measurement matrix in sensor nodes, then measurement capability is improved, but device complexity and memory requirements increase
Solution Approach 1:
The patent extracts measurement information from the routing topology itself rather than storing measurement matrices in sensor nodes. The routing paths that naturally exist in the network are used as measurement vectors, eliminating the need for additional memory storage and reducing device complexity at resource-constrained sensor nodes
Solution Approach 2:
The routing infrastructure serves itself by providing measurement capability through its topology structure. The routing paths automatically function as measurement vectors without requiring sensor nodes to store or process measurement matrices, making the system self-sufficient and reducing device complexity
3Loss of information
If per-packet routing paths are recorded in data packets for CS measurement, then routing information is captured, but packet overhead and bandwidth consumption increase
Solution Approach 1:
The routing path information already embedded in data packets for traditional routing purposes is repurposed as measurement matrices for compressed sensing. This multi-functional use of existing routing data eliminates the need for separate path recording mechanisms, capturing routing information without additional bandwidth consumption
Solution Approach 2:
The patent extracts measurement capability from the existing routing topology structure rather than adding separate path recording overhead. By taking out and utilizing the routing paths that already exist in the data transmission process, the system captures complete routing information without increasing packet size or bandwidth consumption
4Measurement precision
If random walk algorithms are used for data gathering in multi-hop WSNs, then measurements are collected along random walks, but additional transmissions increase energy consumption
Solution Approach 1:
The patent makes the existing routing paths multi-functional by using them simultaneously for data transmission and compressed sensing measurements. This eliminates the need for separate random walk transmissions, as the routing infrastructure already provides sufficient measurement coverage without additional energy-consuming transmissions
Data Source
AI summary
A system and method for data acquisition in a wireless sensor network include receiving measurements made by a plurality of sensor nodes in the wireless sensor network during a given time period, the measurements being carried in data packets that are routed from the plurality of sensor nodes to a base node. The system and method also include determining a plurality of routing paths for the data packets based on routing topology tomography. The system and method further include determining a measurement matrix and a representation basis, and acquiring sensor signals in the wireless sensor network based on the measurement matrix and the representation basis.