FPGA-Oriented Real-Time EMD-Based Breath Signal Processing System on ARM11 MPCore Platform

Abstract

In this paper, a complete FPGA-oriented real-time empirical mode decomposition (EMD)-based breath signal processing system on ARM11 MPCore platform is developed. The system consists of three subsystems including biomedical signal acquisition, FPGA-oriented EMD accelerator and EMD processing result display. In this work, the previous published multiple stopping criteria and high IMFs Application Specific Integrated Circuit (ASIC)-oriented EMD accelerator is deployed to FPGA of ARM 11 MPCore platform such that the computation time of the overall system can he reduced. An external SRAM architecture in ARM 11 MPCore platform is adopted to process a large amount of data. The three subsystems are connected by a high speed internet such that real time data transportation is achieved. From the evaluation results, for 10K data size, the FPGA-oriented EMD accelerator can speed up by 29.21 times, 20.20 times, and 15.47 times compared with ARM11 processor running single core, dual cores, and four cores, respectively.