Research Article
High-Throughput Machine Learning Approaches for Network Attacks Detection on FPGA
@INPROCEEDINGS{10.1007/978-3-030-34365-1_5, author={Duc-Minh Ngo and Binh Tran-Thanh and Truong Dang and Tuan Tran and Tran Thinh and Cuong Pham-Quoc}, title={High-Throughput Machine Learning Approaches for Network Attacks Detection on FPGA}, proceedings={Context-Aware Systems and Applications, and Nature of Computation and Communication. 8th EAI International Conference, ICCASA 2019, and 5th EAI International Conference, ICTCC 2019, My Tho City, Vietnam, November 28-29, 2019, Proceedings}, proceedings_a={ICCASA \& ICTCC}, year={2019}, month={12}, keywords={Machine learning FPGA platform Network attacks}, doi={10.1007/978-3-030-34365-1_5} }- Duc-Minh Ngo
Binh Tran-Thanh
Truong Dang
Tuan Tran
Tran Thinh
Cuong Pham-Quoc
Year: 2019
High-Throughput Machine Learning Approaches for Network Attacks Detection on FPGA
ICCASA & ICTCC
Springer
DOI: 10.1007/978-3-030-34365-1_5
Abstract
The popularity of applying Artificial Intelligence (AI) to perform prediction and automation tasks has become one of the most conspicuous trends in computer science. However, AI systems usually require heavy computational tasks and result in violating applications that need real-time interactions. In this work, we propose a system which is a combination of FPGA platform and AI to achieve a high-throughput network attacks detection. Our architecture consists of 2 well-known and powerful classification techniques, which are the Decision Tree and Neural Network. To prove the feasibility of the proposed approach, we implement a prototype on NetFPGA-10G board using Verilog-HDL. Moreover, the prototype is trained and tested with NSL-KDD dataset, the most popular dataset for network attack detection system. Our experimental results show that the Neural network core can detect attacks with speed at up to 9.86 Gbps for all packet sizes from 64B to 1500B, which is thoroughly 11x and 83x times faster than Geforce GTX 850M GPU and i5 8th generation CPU, respectively. The Neural Network classifier system can function at 104.091 MHz and achieve the accuracy at 87.3.