| 摘要: |
| 在大规模无线传感器网络中的时延估计研究中，良好的数据压缩是实现节省带宽资源与尽量降低时延估计精度损失的一个重要方法。然而，传统的方法在信号压缩与重建过程中存在信息损失大、计算复杂度高等问题。为了解决这些问题，提出了一种使用卷积自编码器-注意力机制-广义互相关相位变换的模型，实现无线传感器网络中的二进制相移键控(Binary Phase Shift Keying,BPSK)信号数据压缩与时延估计的平衡。仿真实验表明，相较于使用奇异值分解与赫夫曼(Huffman)编码分别压缩解压后计算的时延，所提模型具有以下优势：一是可实现更高的信号压缩率；二是重构信号具有更小的均方误差；三是压缩比例相同时具有更高的时延估计精度。仿真实验数据进一步验证了模型在应用场景中的有效性，表明其适用于大规模无线传感器网络的实际应用。 |
| 关键词: 无线传感器网络(WSN) 数据压缩 时延估计 卷积自编码器-注意力机制 |
| DOI：10.20079/j.issn.1001-893x.240613005 |
|
| 基金项目: |
|
| Balanced Optimization of Data Compression and Delay Estimation Accuracy in Wireless Sensor Networks |
| REN Guodong,GAO Yong |
| (College of Electronics and Information Engineering,Sichuan University,Chengdu 610065,China) |
| Abstract: |
| In the research of delay estimation in large-scale wireless sensor networks(WSNs),effective data compression is an important approach to save bandwidth resources and minimize the loss of delay estimation accuracy.However,traditional methods suffer from problems such as large information loss and high computational complexity during signal compression and reconstruction.A model combining a convolutional autoencoder,attention mechanism,and generalized cross-correlation phase transfor-mation is proposed to achieve a balance between binary phase shift keying(BPSK) signal data compression and delay estimation in WSNs.Simulation experiments indicate that compared with the methods using singular value decomposition and Huffman coding for compression and decompression followed by delay computation,the proposed model offers such advantage as higher signal compression ratio,reconstructed signals with smaller mean square error and higher delay estimation accuracy at the same compression ratio.The simulation experimental data further verifies the effectiveness of the model in the application scenario and shows that the method is suitable for the practical application in large-scale WSNs. |
| Key words: wireless sensor network(WSN) data compression delay estimation convolutional autoencoder-attention mechanism |
《电讯技术》编辑部 