Vibration data compression in wireless sensors network

Wireless sensors networks are acquiring more attention in the last years due to their application in many fields such as health monitoring and machinery fault diagnosis. However, they suffer from limited bandwidth as compared to wired networks or cables which affect their usage in data streaming tasks especially when high sampling rates are required. Therefore, data compression becomes relevant solution not only to increase sampling rates but also to save valuable power. In this work, a Modified Discrete Cosine Transform (MDCT) based compression scheme is proposed to compress vibration signals. Some enhancements to the algorithm of implementation of the MDCT are proposed to make it more suitable to the limited resources of wireless nodes. Three wireless nodes, made by the authors, are employed to host the compression scheme. The accuracy and efficiency of the proposed scheme are investigated by conducting several experimental tests on actual vibration signals generated by a machinery vibration simulator (MVS-1).