Dynamic Causal Attack Graph based Cyber-security Risk Assessment Framework for CTCS System

Zikai Zhang · 2025

Read Dynamic Causal Attack Graph based Cyber-security Risk Assessment Framework for CTCS System by Zikai Zhang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-25786.txt.

Introduction

Zikai Zhang builds a framework to judge cyber risk in the control system that keeps high-speed trains safe. Attacks on such systems unfold in stages, with each step enabling the next, and the pattern shifts over time. His 2025 paper models those linked steps as a dynamic causal attack graph.

The graph is built from a standard attack graph, then enriched with timing and cause links between stages. Two forms of inference, one based on probability and one on logic gates, support the risk sums. A case study on the CTCS-3 system tests the full approach. The framework can rank which attacks matter most, which helps planners set defense priorities.

Catalog record

CREATORZikai Zhang
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE47 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-25786

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Dynamic Causal Attack Graph based Cyber-security Risk Assessment Framework for CTCS System Zikai Zhang Abstract Protecting the security of the train control system is a critical issue to ensure the safe and reliable operation of high-speed trains. Scientific modeling and analysis for the security risk is a promising way to guarantee system security. However, the representation and assessment of the multi-staged, causally related, and temporal-dynamic changed attack dependencies are difficult in the train control system. To solve the above challenges, a security assessment framework based on the Dynamical Causality Attack Graph (DCAG) model is introduced in this paper. Firstly, the DCAG model is generated based on the attack graph with consideration of temporal attack propagation and multi-stage attack event causality propagation. Then, the DCAG model is analyzed based on Bayesian inference and logic gateway-based inference. Through the case analysis of the CTCS-3 system, the security assessment framework is validated. With the DCAG-based security assessment framework, we can not only perform appropriate security risk quantification calculations, but also explore the importance of different attacks on system security risks, which is helpful in adjusting the cyber security defense policy. I Introduction Railways have been developing rapidly in the last few decades. Take China as an example, as of 2023, the high-speed railway has a total length of 45 thousand kilometers, longer than a full circle around the equator of the earth. As the vital equipment and technology of high-speed railway, the train control system is provided to control the spacing and speed of the train, to guarantee the safety, reliability, and efficiency of train operation [ 1 ] . The Chinese train control system (CTCS) is a large-scaled, multi-layered, and complicatedly structured train control system. To improve the automation and informatization level of CTCS systems, information technologies have been widely applied, which also bring great security risks that cause unnecessary braking, degraded running, emergency braking, and even collision accidents of trains [ 2 ] . Generally, a CTCS system can be viewed as a typical cyber-physical system, where the network and software work in the cyber domain and trains operate in the physical domain. So, the goal for our cyber security is to prevent the CTCS system from every malware attack and network attack [ 3 ] . To ensure cyber security, firewalls, anti-virus software, anomaly detection/intrusion detection technologies, and other cyber security protection methods are utilized on the CTCS system. However, due to the delay-sensitive character of the CTCS system, the security protection performances must be constrained by system functional requirements to ensure availability and efficiency [ 4 ] . Moreover, with the improvement of attack technology and the discovery of new vulnerabilities, it is more difficult to protect the CTCS system. So, it is important to assess cyber security risks and provide answers about ”Which kind of attack is more dangerous?” or ”What is the weak point of the current security protection configuration?”. Traditionally, attack graph models [ 5 ] are widely used to describe attack scenarios with isolated vulnerabilities and attack paths. And the risk levels of the system are assessed utilizing Common Vulnerabilities and Exposures (CVE) [ 6 ] or Common Vulnerability Scoring System (CVSS) [ 7 ] . However, directly assessing the security risk of CTCS system with the attack graph model is challenging for the following reasons. The first challenge is about the selection of the security risk metric to quantitatively measure the influence of a security-relevant attribute in the attack scenarios of the system. Just considering the vulnerability-related metric is not suitable, since some viable solutions may be deployed in the system to deal with the vulnerabilities. The second challenge is about the modeling of complex attack scenarios. Simply using an attack graph can not properly describe the multi-stage attacks (focus on risk spread probability), the causal-related attack propagations (focus on risk event trigger probability), and the temporal accumulation of security risks. The third challenge is about dealing with dynamic features in attack-defend scenarios. In the attack-defend process, vulnerabilities are discovered and patches are applied, novel attacks are designed, new protection techniques are deployed, and some attacks succeed and propagate in the system over time. To address the above challenges, a Dynamic Causal Attack Graph (DCAG) based cyber security risk assessment framework is introduced for CTCS systems. Firstly, the DCAG model is generated based on the attack graph, with consideration of attack propagation and multi-stage attacks. And, during the generation process, the root node of the graph is converted from vulnerabilities to attack scenarios, the nodes from the previous time slice are added to deal with temporal security risk accumulation, and event causality relations are added on the edges to represent attack propagation dependencies. Then, combining Bayesian inference and logic gateway-based inference, a risk assessment method is proposed to deal with attack propagation and attack event causality propagation iteratively, which also provides the technical basis for the dynamic security risk assessment framework. The contributions of this paper are as follows: 1. A DCAG modeling method is proposed to deal with complex attack scenarios considering temporal security risk accumulation, attack propagation, and multi-stage attack event causality propagation. 2. Leveraging our DCAG modeling approach, a cyber security risk assessment process is provided to measure the influence of attack scenarios, combining Bayesian inference and logic gateway-based inference. 3. A dynamic security risk assessment framework is proposed based on the DCAG model to evaluate the risk level of crucial components in the system. A case study based on CTCS-3 is illustrated to elucidate the validity of the proposed framework. The remainder of this paper is organized as follows: The preliminaries of this paper are introduced in Section II. The security risk assessment framework and the DCAG model are introduced in Section III. Section IV provides the case study, and Section V concludes this paper. II Preliminary II-A Overview of CTCS-3 system CTCS systems are safety-critical, and the fail-safe mechanisms are applied to achieve the demanded performance, including Reliability, Availability, Maintainability, and Safety (RAMS). As an example, the CTCS-3 train control system satisfies the requirements of a safety train operating with a speed range of 300-350 km/h and minimal tracking intervals of 3min. The structure of CTCS-3 is shown in Fig. 1 , including the on-board subsystem, the trackside subsystem, and the central subsystem. Fig. 1: CTCS-3 train control system The on-board system is connected with the Base Transceiver Station (BTS) in the trackside subsystem based on Global System for Mobile Communication-Railway (GSM-R) wireless communication. On the trackside subsystem, the train control center (TCC) receives real-time train speed and position reports and track circuit (TC) occupancy information. The TC provides continuous train occupancy checking. Lineside Electronic Unit (LEU) and Balise systems are responsible for achieving the dynamic orientation of trains. The computer-based interlocking (CBI) subsystem is used as the basic signal equipment on the spot, which mainly controls the turnout and the annunciator in the station. On the central subsystem, the Centralized Traffic Control (CTC) system is the railway signaling technology equipment that conducts dispatchings and commands of the trains, and achieves centralized control by signal equipments. According to the occupation of the train within the Radio Block Center(RBC)’s jurisdiction and the state of the incoming route, the Movement Authority

Read the complete text