A high-efficiency and energy-saving data compression method for spaceborne wireless sensor network

Abstract

The invention discloses a high-efficiency and energy-saving spaceborne wireless sensor network data compression method, comprising the following steps: step 1, collecting flow data of each terminal sensing node of the wireless sensor network; step 2, preprocessing the flow data; step 3 1. Construct the calculation method of the D-CRBM network calculation layer; Step 4, combine the D-CRBM network calculation layer and the variational hybrid encoder to construct the CBN-VAE network; Step 5, train the CBN-VAE network, obtain model parameters, and build compression model; step 6, compressing the data of the wireless sensor network by using the compression model. The invention effectively reduces the energy consumption of node communication, storage energy consumption and calculation energy consumption of the wireless sensor network.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence. More specifically, the present invention relates to a high-efficiency and energy-saving space-borne wireless sensor network data compression method. Background technique [0002] Wireless Sensor Networks (WSN for short) is a smart device that integrates sensing, computing, and communication capabilities. Because of its simple deployment, automatic data collection, real-time data processing, multi-hop wireless communication of ad hoc network, and good adaptability to harsh environments, WSN is considered to be an important technology for future aerial detection. Due to the distance from the earth and the complex environment, deep space exploration is more technically difficult and risky than near-earth or even lunar exploration activities. It needs to further break through a number of new core technologies, among which autonomous navigation control, energy and propulsion, Measurem...

AI Technical Summary

Problems solved by technology

An important reason for the lack of research is that the calculation of deep convolutional networks is very expensive, and it is difficult to apply to sensor nodes with limited computing power. Yildirim et al. used deep convolutional networks to compress ECG signals, but the network requires a lot of calculations. Comprehensive As mentioned above, it is necessary to find a high-efficiency and energy-saving compression model calculated by convolutional neural network for WSN data compression.

Images