A Performance Study of Random Interleaver Based Data Loss Recovery Technique for Structural Health Monitoring

The extensive use of Wireless Sensor Networks (WSNs) has simplified health monitoring of structures (especially long bridges) using acceleration/vibration data. However, various factors such as wireless medium, hardware malfunction strongly influence the robust data transmission in WSNs. Recently, for Structural Health Monitoring (SHM) applications, Compressive Sensing (CS) based techniques are widely used for efficiently handling data losses. However, CS implements encoding process by the use of random matrices, which requires significant memory and computational power. This paper showcases a CS model in which the raw acceleration data is interleaved and transmitted without performing any encoding process. At the receiver, even after data losses, the original signal is faithfully reconstructed using optimization techniques. The whole model is realized using Raspberry Pi B+ board, MEMS accelerometer MMA8452Q, and Wi-Fi network. This paper also presents the performance analysis of the proposed random interleaver based technique with the existing CS techniques, majorly regarding computational power, processing time and memory.