论文信息 - Key Issues of FPGA Implementation of Neural Networks

Key Issues of FPGA Implementation of Neural Networks

Recent years, the artificial neural networks were noticeable on development. As the bridge between the theoretical research and application research, the hardware implementation technologies have developed rapidly, particularly in configurable FPGA implementation technologies. However we have found the shortcomings of the existing methods which need to be improved. This paper has taken research on the key issues of the FPGA implementation of neural networks, discussing on the following issues: data representation, inner-products computation, and implementation of activation function, storage and update of weights, nature of learning algorithm and design constraints. It also introduces some relatively mature and new methods and pointed out their deficiencies and future works.

[1] Wan Yong. Neural networks hardware implementation based on FPGA , 2007 .

[2] Jagath C. Rajapakse,et al. FPGA Implementations of Neural Networks , 2006 .

[3] Jihan Zhu,et al. FPGA Implementations of Neural Networks - A Survey of a Decade of Progress , 2003, FPL.

[4] J. M. Tarela,et al. Approximation of sigmoid function and the derivative for hardware implementation of artificial neurons , 2004 .

[5] Wayne Luk,et al. Parameterized High Throughput Function Evaluation for FPGAs , 2004, J. VLSI Signal Process..

[6] Stefano Ramat,et al. A Parallel Neural Processor for Real-Time Applications , 2002, IEEE Micro.

[7] Wang Zhensong. Realization of BP Algorithm in FPGA Based on Systolic Array Architecture , 2006 .

[8] Stamatis Vassiliadis,et al. Elementary function generators for neural-network emulators , 2000, IEEE Trans. Neural Networks Learn. Syst..

[9] Zhang Xiu. Research on the Key Problems in FPGA-based Neural Network Hardware Implementation , 2005 .