A Digital Compressed Sensing-Based Energy-Efficient Single-Spot Bluetooth ECG Node

Figure 1. Hardware framework of proposed wireless ECG node. (a) System framework (b) structure, and prototype of the proposed node.

Figure 2. Dual-clock source-based periodic sleep/wake-up scheme. (a) Flowchart of the operation modes. (b) Sequence diagram of the operation modes.

Figure 3. Scheme of digital CS compression on node.

Figure 4. The ECG signals for compression tests.

Figure 5. The prototype of single-spot Bluetooth ECG node.

Figure 6. The relationships of signal distortion (PRD), compression efficiency (M), and computation complexity (K).

Figure 7. The visual inspection results. The recovery algorithm and sparse bases are BSBL and DCT, respectively: N = 512, M = 256, and K = 4; Pan-Tompkins method [30] was used for R-peak detection in the recovery signals. The red dotted lines are the segmented indicators of every frame, and the signal quality of each recovered frame was evaluated by PRD.

Figure 8. Power traces of the proposed node during different schemes. (a) The normal scheme, no energy saving strategy is used. (b) The S/W scheme. (c) The S/W scheme + CS. (d) The detailed figure of (c). The sampling energy consumption is marked by the red ellipses in (a–c).

Figure 9. Energy consumption of the proposed node during different schemes.

Figure 10. The recovered signal with 10 points delay.