DeepFed: Federated Deep Learning for Intrusion Detection in Industrial Cyber-Physical Systems

In the present computing world, network intrusion detection systems are playing a vital part in detecting malicious activities, and enormous attention has been given to deep learning from several years. During the past few years, cyber-physical systems (CPSs) have become ubiquitous in modern critical infrastructure and industrial applications. Safety is therefore a primary concern. Because of the success of deep learning (DL) in several domains, DL-based CPS security applications have been developed in the last few years. However, despite the wide range of efforts to use DL to ensure safety for CPSs. The major challenges in front of the research community are developing an efficient and reliable ID that is capable of handling a large amount of data, in analyzing the changing behavioral patterns of attacks in real-time. The work presented in this manuscript reviews the various deep learning generative methodologies and their performance in detecting anomalies in CPSs. The metrics accuracy, precision, recall, and F1-score are used to measure the performance.

Download Full-text

Security of Cyber-Physical Systems: A Generalized Algorithm for Intrusion Detection and Determining Security Robustness of Cyber Physical Systems using Logical Truth Tables

Vanderbilt Undergraduate Research Journal ◽

10.15695/vurj.v9i0.3765 ◽

2013 ◽

Vol 9 ◽

Cited By ~ 4

Author(s):

Curtis G. Northcutt

Keyword(s):

Intrusion Detection ◽

Cyber Security ◽

Logical Truth ◽

Cyber Physical Systems ◽

Security Model ◽

Security Measures ◽

Physical Systems ◽

Multiple Signal ◽

Security Attack ◽

Truth Tables

The recent proliferation of embedded cyber components in modern physical systems [1] has generated a variety of new security risks which threaten not only cyberspace, but our physical environment as well. Whereas earlier security threats resided primarily in cyberspace, the increasing marriage of digital technology with mechanical systems in cyber-physical systems (CPS), suggests the need for more advanced generalized CPS security measures. To address this problem, in this paper we consider the first step toward an improved security model: detecting the security attack. Using logical truth tables, we have developed a generalized algorithm for intrusion detection in CPS for systems which can be defined over discrete set of valued states. Additionally, a robustness algorithm is given which determines the level of security of a discrete-valued CPS against varying combinations of multiple signal alterations. These algorithms, when coupled with encryption keys which disallow multiple signal alteration, provide for a generalized security methodology for both cyber-security and cyber-physical systems.

Download Full-text

Embedded Deep Learning Prototyping Approach for Cyber-Physical Systems: Smart LIDAR Case Study

Journal of Sensor and Actuator Networks ◽

10.3390/jsan10010018 ◽

2021 ◽

Vol 10 (1) ◽

pp. 18

Author(s):

Quentin Cabanes ◽

Benaoumeur Senouci ◽

Amar Ramdane-Cherif

Keyword(s):

Neural Network ◽

Deep Learning ◽

Computation Time ◽

Physical World ◽

Cyber Physical Systems ◽

Physical Systems ◽

Research Technology ◽

Standard Design ◽

Hardware Architectures ◽

Voxel Grid

Cyber-Physical Systems (CPSs) are a mature research technology topic that deals with Artificial Intelligence (AI) and Embedded Systems (ES). They interact with the physical world via sensors/actuators to solve problems in several applications (robotics, transportation, health, etc.). These CPSs deal with data analysis, which need powerful algorithms combined with robust hardware architectures. On one hand, Deep Learning (DL) is proposed as the main solution algorithm. On the other hand, the standard design and prototyping methodologies for ES are not adapted to modern DL-based CPS. In this paper, we investigate AI design for CPS around embedded DL. The main contribution of this work is threefold: (1) We define an embedded DL methodology based on a Multi-CPU/FPGA platform. (2) We propose a new hardware design architecture of a Neural Network Processor (NNP) for DL algorithms. The computation time of a feed forward sequence is estimated to 23 ns for each parameter. (3) We validate the proposed methodology and the DL-based NNP using a smart LIDAR application use-case. The input of our NNP is a voxel grid hardware computed from 3D point cloud. Finally, the results show that our NNP is able to process Dense Neural Network (DNN) architecture without bias.

Download Full-text