HH-NIDS: Heterogeneous Hardware-Based Network Intrusion Detection Framework for IoT Security

Abstract

This study proposes a heterogeneous hardware-based framework for network intrusion detection using lightweight artificial neural network models. With the increase in the volume of exchanged data, IoT networks’ security has become a crucial issue. Anomaly-based intrusion detection systems (IDS) using machine learning have recently gained increased popularity due to their generation’s ability to detect unseen attacks. However, the deployment of anomaly-based AI-assisted IDS for IoT devices is computationally expensive. A high-performance and ultra-low power consumption anomaly-based IDS framework is proposed and evaluated in this paper. The framework has achieved the highest accuracy of 98.57% and 99.66% on the UNSW-NB15 and IoT-23 datasets, respectively. The inference engine on the MAX78000EVKIT AI-microcontroller is 11.3 times faster than the Intel Core i7-9750H 2.6 GHz and 21.3 times faster than NVIDIA GeForce GTX 1650 graphics cards, when the power drawn was 18mW. In addition, the pipelined design on the PYNQ-Z2 SoC FPGA board with the Xilinx Zynq xc7z020-1clg400c device is optimised to run at the on-chip frequency (100 MHz), which shows a speedup of 53.5 times compared to the MAX78000EVKIT.

Hardware/Software Codesign of Real-Time Intrusion Detection System for Internet of Things Devices

“…Here, a hierarchical decisionmaking approach is taken for real-time IoT network intrusion detection. 3) NN2023 [36]: An NN accelerator working a clock frequency of 100 MHz. This accelerator was extended from NN2021 for a heterogeneous hardware-based network intrusion detection framework.…”

“…Lastly, we evaluated the inference time of the proposed IDS when deployed on the PYNQ-Z2. Here, to facilitate a fair comparison, we performed the inference task on 22,544 records (similar to previous studies [9], [10], [36]). In this evaluation, in addition to the above four accelerators, we also compared a few software implementations of IDSs on an embedded microprocessor (i.e., an ARM Cortex-A9, which is deployed on the PYNQ-Z2) and a desktop PC (with an Intel Core i7 CPU).…”

Dragon_Pi: IoT Side-Channel Power Data Intrusion Detection Dataset and Unsupervised Convolutional Autoencoder for Intrusion Detection

“…While this makes sense, since cybersecurity has been dominated by network-based security for decades, it is imperative to consider security from as many points of view as possible in the modern cyber threat topography. IoT 23 [31] Network Packets 2020 [32][33][34] N/A Derived from the 1998 Defence Advanced Research Project Agency (DARPA) dataset [35], the Knowledge Discovery and Data Mining (KDD) 1999 dataset was created to be used in the development of security-focused ML [10,11]. In essence, the KDD dataset transforms the DARPA dataset's network traces into a collection of features found in the connection between two hosts, with each connection constituting its own record in the dataset.…”

AIfES: A Next-Generation Edge AI Framework

“…Even though training FCNNs and CNNs on MCUs is now feasible, complex arithmetic operations (e.g., matrix multiplication) are still performed that are challenging for MCUs. With customized hardware accelerators, complex calculations are performed faster which saves resources and energy [25], [26], [27]. However, none of the existing developments enable a modular structure to insert custom hardware accelerators in their framework.…”