Optimized compression and recovery of electrocardiographic signal for IoT platform

Internet of Things (IoT) promises to a continuous, remote recording and monitoring of electrocardiogram (ECG). Thus it creates large volumes of data for healthcare purposes. The huge recordings result in the heavy burden of the communication, and the wearable devices require frequent charging since the huge data consumes energy quickly. To ameliorate this, we aim to compress the recordings and, in turn, to boost the battery life. We propose a new framework from two aspects: First, an optimization solution is proposed for the measurement matrix, which uses the shrinking singular value decomposition (SSVD) strategy at the compression terminal. Second, an accelerated method based on the non-uniform norm (ANN) is proposed to estimate and reconstruct the received signal. The proposed framework of the measurement matrix optimization and ANN estimator is firstly used for the monitoring of vital parameters such as electrocardiography (ECG). Experiments are conducted to confirm the superiority of the proposed SSVD and ANN methods.