Compressed sensing method for human activity sensing using mobile phone accelerometers
Compressed sensing method for human activity sensing using mobile phone accelerometers
复制标题
DOI：
10.1109/inss.2012.6240525
复制
发表时间：
2012-06
期刊：
影响因子：
--
通讯作者：
Daito Akimura;Y. Kawahara;T. Asami
中科院分区：
文献类型：
--
作者：
Daito Akimura;Y. Kawahara;T. Asami
本文提出了首次将压缩感知(CS)理论应用于智能手机活动传感器数据采集的完整设计。如今，大多数手机都配备了多个传感器，如摄像头、GPS和加速计。通过利用感知功能，我们捕获了许多不同的事件，并在移动网络上共享它们。对于这种参与式传感系统来说，最重要的挑战之一是降低移动设备的电池消耗。我们通过减少通信数据量来克服这一挑战，而不会在移动终端引入密集的计算。CS技术在移动端由非常简单的矩阵运算组成，在网络侧资源丰富的机器上进行CPU密集型重构。由于CS是有损压缩技术，根据原始信号的稀疏程度，重建信号包含误差。我们通过使用由90名测试对象执行的六个基本活动组成的大量真实活动数据来评估所提出的方法。我们还在iPhone/iPod平台上实现了我们的方法，结果表明，与ZIP压缩相比，我们的方法可以减少大约16%的功耗，同时将误差保持在10%以下。
This paper presents the first complete design to apply the compressed sensing (CS) theory to activity sensor data gathering for smart phones. Today, most of the mobile phones are equipped with multiple sensors, such as cameras, GPS, and accelerometers. By exploiting the sensing features, we capture many different events and share them over the mobile network. One of the most important challenges for such a participatory sensing system is to reduce the battery consumption of the mobile device. We overcome this challenge by reducing the communication data, without introducing intensive computation at mobile terminals. The CS technique consists of very simple matrix operations at the mobile side, and CPU-intensive reconstruction is performed on the resource-rich machine on the network side. Since CS is a lossy compression technique, the reconstructed signal contains errors depending on the degree of sparseness of the original signal. We evaluated the proposed method by using a large amount of real activity data consisting of six basic activities performed by 90 test subjects. We also implemented our method on the iPhone/iPod platform and showed that our method can reduce power consumption by approximately 16% as compared with ZIP compression, while maintaining the error below 10%.