Application specific processor design implementation to monitor seismic activity

In this paper, we build an application specific processor based on Artificial Neural Networks (ANN) to monitor seismic activity. The training of the ANN is carried out using a software framework to evaluate the initial weights of proposed architecture. Once the layers and number of neurons of ANN are estimated using the neural heuristic, the proposed architecture is implemented on Xilinx Virtex-6 FPGA to showcase the applicability in monitoring seismic activity. During the implementation, a hardware framework is built on FPGA and the accuracy of hardware framework is examined by comparing the estimated outcome with the software framework.