Method for collecting low information density data with scalable quality based on compressed sensing

Abstract

In order to solve the problem that a distributed sensing network communication channel is unstable, and has low network information density, the invention provides a method for collecting low information density data with scalable quality based on compressed sensing by maximally optimizing the data collecting quality through sensing data space-time correlation and the scalability of compressed sensing measuring matrix. The dimension of the measuring matrix is initially set based on the information density of a signal source, the dimension of the measuring matrix is adjusted based on real-time effective communication bandwidth, and the scalable control of data collecting quality is realized. Through expending and reducing of the measuring matrix, the data collecting quality can be dynamically adjusted according to the information density of the signal source and the network bandwidth state. The method is widely applicable.

Description

technical field [0001] The invention relates to a method for collecting data with scalable quality and low information density based on compressed sensing, and belongs to the fields of communication technology and data collection. Background technique [0002] With the continuous development of perception technology, the amount of data in the perception network is growing at a high speed in a geometric progression, and the continuously growing information resources contain a huge amount of valuable information, and people have entered the era of "big data". However, perception data has obvious characteristics of low information density—mainly manifested as a large amount of data and a small amount of valuable data. Therefore, individuals and enterprises are increasingly eager for data analysis and other services. Without effective source detection and data collection methods, users often cannot extract truly effective information and cannot realize the effective use of low-i...

AI Technical Summary

Problems solved by technology

However, ordinary clustering, distributed spatiotemporal correlation and other methods only simply apply compressed sensing technology to data collection, and there is no method to expand or reduce the dimension of the measurement matrix in combination with the network bandwidth status to dynamically adjust the quality of data collection.

[0006] In summary, for how to combine CS Measurement Matrix and Network Bandwidth Status To achieve quality scalable low information density data collection methods from low information density data and to achieve the purpose of dynamically adjusting the quality of data collection, there is currently no scientific solution

Images