|
|
A Novel Application of Compressed Sensing for the Accelerometer Data from Wireless Body Area Network with Low Energy Consumption |
1 School of Mathematics and Computer Science of Fujian Normal University, Fuzhou 350007, China
2 Key Laboratory of Biomedical Information Engineering of Education Ministry, Xi’an Jiaotong University, Xi’an 710049, China
|
|
|
Abstract This paper proposed a novel scheme of energy-efficient compressed sensing framework for processing the acceleration-based data acquired from wireless body area networks (WBANs), in order to save the energy of WBANs-based system. With the optimal scheme of sparse binary matrices, the raw accelerometer data with nonsparse is compressed by linearly projected at sensor node before their transmission, and then the compressed data is reconstructed by the novel block Bayesian learning algorithm (BSBL) at remote terminal. The acceleration data from USC-HAD dataset of Southern California was used to evaluate the effectiveness of our proposed technique. The experimental results showed that the optimal scheme of sparse binary matrix could obtain the same reconstruction error (0.0045) as Gaussian or Bernouilli random matrix when a number of nonzero values were selected as 6 in each column of the designed sparse binary matrix and the ratio of compression was 50%. Besides, compared with the traditional CS-based reconstruction algorithms, our proposed BSBL algorithm for reconstruction of acceleration data could increase by 17 dB of signalnoise ratio, significantly improving the performance of reconstruction of acceleration data. These results suggested that, with the optimal design of sparse binary matrix, the designed compressed sensing framework could acquired the acceleration data at sub-Nyquist sampling rate and greatly reduce the number of transmitted data by simple linear transform at sensor node for saving energy. It also can contribute to improving the performance of reconstruction of non-sparse acceleration data by using BSBL. Our work can provide a novel approach for further practical implementation such as the design of simple hardware of sensor node, improvement of the performance of reconstruction of accelerationdata and the development of WBANs\|based system with lower energy consumption for remote monitoring physical activity.
|
|
|
|
|
[1]Hadjidj A, Souil M, Bouabdallah A, et al. Wireless sensor networks for rehabilitation applications: challenges and opportunities [J]. Journal of Network and Computer Applications, 2013, 36(1): 1-15.
[2]石欣, 张涛. 一种可穿戴式跌倒检测装置设计 [J]. 仪器仪表学报, 2012, 33(3): 576-580.
[3]Balouchestani M, Raahemifar K, Krishnan S. Compressed sensing in wireless sensor networks: survey [J]. Canadian Journal on Multimedia and Wireless Networks, 2011,2(1): 231-239.
[4]王天荆, 张宝玉, 杨震. 基于滤波的压缩感知信号采集方案 [J]. 仪器仪表学报, 2013, 34(3): 573-580.
[5]Anumbhuti K, Manish S, Vijay BN. ECG data compression using dwt [J]. International Journal of Engineering and Advanced Technology, 2011, 1(1): 11-13.
[6]Fauvel S, Ward R. An energy efficient compressed sensing framework for the compression of electroencephalogram signals [J]. Sensors, 2014, 14(1): 1474-1496.
[7]Mamaghanian H, Khaled N, Atienza D, et al. Compressed sensing for realtime energyefficient ECG compression on wireless body sensor nodes[J]. IEEE Transactions on Biomedical Engineering, 2011, 58(9): 2456-2466.
[8]Dixon AMR, Allstot EG, Gangopadhyay D, et al. Compressed sensing system considerations for ECG and EMG wireless biosensors [J]. IEEE Transactions on Biomedical Circuits and Systems, 2012, 6(2): 156-166.
[9]Zhang Z, Jung T, Makeig S, et al. Compressed sensing for energyefficient wireless telemonitoring of noninvasive fetal ECG via block sparse Bayesian learning [J]. IEEE Transactions on Biomedical Engineering, 2013, 60(2): 300-309.
[10]Zhang Z, Jung T, Makeig S, et al. Compressed sensing of EEG for wireless telemonitoring with low energy consumption and inexpensive hardware [J]. IEEE Transactions on Biomedical Engineering, 2013, 60(1): 221-224.
[11]Donoho D. Compressed sensing [J]. IEEE Transactions on Information Theory, 2006, 52(4): 1289-1306.
[12]Donoho D. For most large underdetermined systems of linear equations, the minimal l1-norm solution is also the sparsest solution [J]. Communications on Pure and Applied Mathematics, 2006, 59(6): 797-829.
[13]Gilbert A, Indyk P. Sparse recovery using sparse matrices [J]. Proceedings of the IEEE, 2010, 98(6): 937-947.
[14]Zhang Z, Rao B. Extension of SBL algorithms for the sensing framework of block sparse signals with intrablock correlation [J]. IEEE Transactions on Signal Processing, 2013,61(8): 2009-2015.
[15]Tropp J, Gilbert A. Signal recovery from random measurements via orthogonal matching pursuit [J]. IEEE Transactions on Information Theory, 2007, 53(12): 4655-4666.
[16]Berg E, Friedlande M. Probing the pareto frontier for basis pursuit solutions [J]. SIAM Journal on Scientific Computing, 2008, 31(2): 890-912.
[17]Mohimani H, Massoud B, Jutten C. A fast approach for overcomplete sparse decomposition based on smoothed l0 norm [J]. IEEE Transactions on Signal Processing, 2009, 51(1): 289-301.
[18]Zou H, Hastie T. Regularization and variable selection via the elastic net [J]. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 2005, 67(2): 301-320.[19]Eldar Y, Kuppinger P, Bolcskei H. Blocksparse signals: uncertainty relations and efficient recovery [J]. IEEE Transactions on Signal Processing, 2010, 58(6): 3042-3054.
[20]Yu L, Sun H, Barbot J, et al. Bayesian compressive sensing for cluster structured sparse signals [J]. IEEE Transactions on Signal Processing, 2012, 92(1): 259-269.
[21]Huang J, Zhang T, Metaxas D. Learning with structured sparsity [J]. The Journal of Machine Learning Research, 2011, 〖STHZ〗12(2): 3371-3412.
[22]Zhang M, Sawchuk A. USCHAD: a daily activity dataset for ubiquitous activity recognition using wearable sensors [C] // Anind KD, eds. Proceedings of the 2012 ACM Conference on Ubiquitous Computing. Pittsburgh: ACM, 2012: 1036-1043. |
|
|
|