一种基于自编码器的无线传感网数据收集方案

An Autoencoder-Based Data Collection Scheme for Wireless Sensor Networks

利用自编码器设计了一种高效的数据收集方案.该方案包括模型训练和数据收集两个阶段,在模型训练阶段利用历史监测数据进行模型训练从而获得测量矩阵和重构矩阵,在数据收集阶段首先利用测量矩阵实现分布式数据压缩,然后利用重构矩阵重构网络中的所有监测数据.实验结果表明,该方案不但具有较高的数据压缩率,而且具有较高的数据重构精度和较快的数据重构速度.

Data collection is one of the key operations in wireless sensor networks. In this paper, we propose an energy efficient data collection scheme for wireless sensor networks by using an autoencoder. It includes the model training process and the data collection process. In the model training process, historical dataset is utilized to train the autoencoder with the goal of obtaining a measurement matrix and a reconstruction matrix. In the data collection process, the measurement matrix is utilized to compress the sensed data in a distributed manner and the reconstruction matrix is utilized to reconstruct the surveillant data of the whole sensor network. The experiment results show that the proposed scheme presents higher data compression ratio and higher data reconstruction accuracy as well as faster data reconstruction speed than existed data collection schemes.