Document Positioning

This document serves as the entry point for the entire knowledge base. Its purpose is to first help readers establish an overall perspective, understand what these two papers are jointly researching, why they can be categorized under the same research direction, what core problems each solves, and what the exact relationship is between the two papers.

To avoid confusion between filenames and paper order, this document uniformly adopts the following convention:

First Paper: second.tex (second author) Second Paper: third.tex (third author)

What is the Main Research Line?

Looking at these two papers together, the main line is very clear. They both revolve around the same core problem:

When a complex, large-scale stochastic system operates in a networked environment and suffers from DoS attacks, can the system's stability continue to be guaranteed?

The term "complex" here has multiple layers of meaning. The system itself consists of multiple coupled nodes that influence each other; the system's evolution process contains stochastic disturbances; the sampling, switching, control updates, or attack moments for different nodes are often not synchronized; meanwhile, state measurement, control input, and feedback transmission rely on network communication. Once a DoS attack occurs, both information and control links can be disrupted, weakening or even completely invalidating the system's original stability mechanisms.

Therefore, the common problem faced by both papers can be summarized as:

In an asynchronous networked attack environment, can a complex large-scale stochastic system still achieve exponential stability; if stability is compromised, what mechanisms should be used to pull the system back to a stable state.

Why is This Problem Important?

Many real-world engineering systems can be abstracted as networks composed of multiple interacting subsystems, such as microgrids, communication networks, neural networks, complex mechanical systems, and networked control systems. Such systems typically share several distinct characteristics.

• The system scale is often large, with coupling between nodes, meaning a local change can propagate globally.
• Systems are often affected by noise, environmental fluctuations, and parameter uncertainties, making their evolution process stochastic.
• The operating rhythms of different nodes are not synchronized; sampling, switching, communication, and control updates often occur asynchronously.
• The system is highly dependent on network communication; once the network is attacked, control effectiveness is immediately impacted.

Therefore, the problem studied by these two papers is essentially:

When the system itself is already sufficiently complex, and the control and communication links become unreliable, can stability still be strictly guaranteed.

First, Clarify Some Fundamental Concepts

To make this entry page more accessible, let's first provide a basic explanation of the core concepts in the papers.

What is a Large-Scale Stochastic System?

Here, a large-scale system refers to an overall system composed of multiple interacting nodes or subsystems. Each node has its own state evolution law while also being influenced by other nodes. "Stochastic" means the system's evolution includes noise, disturbances, or uncertain factors, so the system state does not change in a completely deterministic manner.

You can think of it as: A dynamic system composed of many interconnected nodes operating in an uncertain environment.

What is a DoS Attack?

DoS stands for Denial-of-Service attack. In the context of control systems and networked systems, its direct effect is usually to render communication links ineffective for a period, preventing state information from reaching the controller in time or control commands from reaching the actuator in time. Once this happens, the controller may "not see the system" or "be unable to send control" during certain intervals.

Therefore, a DoS attack here is not a background condition but a key factor that directly alters system stability.

What is Asynchrony?

Asynchrony means that different nodes and events do not share a unified common clock. For example, one node may have updated its control while another has not; one node may have entered an attack interval while another might still be in a normal communication interval; the moment of system switching and the moment the controller receives information may also be inconsistent.

Asynchrony is closer to real-world scenarios than synchrony and also makes stability analysis significantly more difficult.

What is Exponential Stability?

The two papers focus not on simple "the system does not diverge" but on a stronger stability conclusion: exponential stability. Intuitively, exponential stability means the system state decays relatively quickly and returns to the equilibrium point.

In stochastic systems, this further discusses mean-square exponential stability, commonly expressed as $\mathbb{E}\lVert x(t)\rVert^2 \le C e^{-\lambda (t-t_0)}$

Where $x(t)$ is the system state, $C>0$ is a constant, and $\lambda>0$ represents the decay rate. This expression indicates that the expectation of the squared system state decays exponentially over time.

What is Stabilization?

Stabilization means that the system may not be inherently stable under original conditions and requires some intervention mechanism to meet stability requirements. This intervention mechanism can be a controller, an impulsive action, or other designed feedback or regulation methods.

Therefore, what these two papers jointly care about is not just "analyzing how the system behaves" but also "how to actually make the system stable."

Why Do They Belong to the Same Research Direction?

From the research object perspective, both papers deal with large-scale stochastic networked systems. From the external environment perspective, they both incorporate DoS attacks into the model as factors that directly affect stability. From the system characteristics perspective, they both emphasize asynchrony, moving away from idealized fully synchronous assumptions. From the research goal perspective, they both focus on whether the system can achieve mean-square exponential stability. From the technical approach perspective, they both rely on Lyapunov methods, balancing functions, and graph-theoretic methods to complete the analysis.

This shows that the relevance of the two papers is not merely at the title level but is reflected across multiple layers: research object, problem awareness, goal setting, and the main methodological framework.

Summarizing from a higher level, the common research direction they point to can be expressed as:

Stabilization Theory for Complex Large-Scale Stochastic Systems in Asynchronous Networked Attack Environments.

What Core Problem Does the First Paper Solve?

The title of the first paper second.tex is:

Discrete Observation Decentralized Control for Switched Delayed Large-Scale Systems under Asynchronous Denial-of-Service Attacks

This paper primarily studies:

How to achieve exponential stability for switched large-scale stochastic systems with time-varying delays under asynchronous DoS attacks via discrete-observation decentralized control.

The model complexity in this paper is high. Besides large scale, stochastic disturbances, and DoS attacks, it simultaneously considers switching behavior, time-varying delays, mode mismatch, discrete observation, and node-dependent switching. The controller is also not continuously acting; it updates based on discrete sampling information, and attacks further affect the accessibility and effectiveness of control inputs.

Therefore, this paper is more concerned with a problem leaning towards control implementation:

Under conditions where the controller can only observe discretely, nodes control independently, and mode switching and attacks may occur asynchronously, how should the control law be designed to guarantee the system achieves exponential stability?

From an overall positioning perspective, the first paper leans more towards:

Control Design and Stabilization Implementation in Complex Scenarios.

What Core Problem Does the Second Paper Solve?

The title of the second paper third.tex is:

Asynchronous Impulsive Stabilization of Large-scale Networks Under Denial-of-Service Attacks

This paper primarily studies: How to achieve exponential stability for large-scale networks with stochastic disturbances under asynchronous DoS attacks via asynchronous impulsive actions.

It deals with a type of hybrid dynamical system. The system evolves according to stochastic differential equations in continuous time and undergoes impulsive jumps at certain discrete moments. More crucially, these impulses are not necessarily all beneficial for stability; stabilizing impulses and destabilizing impulses may coexist; attack intervals and non-attack intervals alternate; and the impulse moments and attack effects on different nodes may also exhibit asynchrony.

Therefore, this paper is more concerned with a problem leaning towards theoretical unification: When the system simultaneously contains stabilizing impulses, destabilizing impulses, DoS attacks, asynchronous impulse moments, and asynchronous attack intervals, can a unified stability analysis framework be established?

From an overall positioning perspective, the second paper leans more towards: Unified Stability Analysis for Asynchronous Hybrid Systems.

What is the Relationship Between the Two Papers?

Looking at the two papers together, their relationship becomes clearer.

They are not two separate, scattered topics, nor do they simply correspond to "control" and "analysis" respectively. More accurately, they represent the unfolding of the same research line at two different levels.

The second paper leans more towards establishing a unified stability perspective at the theoretical level. It is concerned with how system stability should be characterized holistically under the combined influence of multiple favorable and unfavorable factors.

The first paper pushes this line of thinking further into a more complex system environment closer to engineering implementation, focusing on how to actually construct a working stabilization mechanism under constraints like discrete observation, decentralized control, switching, and delays.

Therefore, their relationship can be summarized as: The second paper leans more towards a unified stability analysis framework, while the first paper leans more towards stabilization control design for complex models.

The former provides a more general, unified theoretical perspective; the latter demonstrates how to implement the stabilization idea under more complex constraints.

What Common Technical Approach Do They Use?

Although the two papers deal with different objects, their methodologies show clear continuity.

First, they both identify stabilizing and destabilizing factors in the system. Destabilizing factors may come from the system's own dynamic growth, node coupling, delay effects, stochastic disturbances, control failure due to attacks, or destabilizing impulses. Stabilizing factors may come from negative feedback control, stabilizing impulses, or effective regulation during certain non-attack intervals.

Second, they both construct node-level Lyapunov functions or Lyapunov functionals to characterize the magnitude of each node's state and the trend of the system's "energy" change.

Third, they both introduce balancing functions. The core role of this tool is to uniformly record the influences favorable and unfavorable to stability across different time intervals, thereby characterizing the overall convergence trend.

Finally, because the research object is a networked coupled system, they both need to use graph-theoretic methods to handle inter-node coupling terms, extending local node analysis to global stability conclusions for the entire system.

Therefore, the common technical approach of the two papers can be understood as: Using Lyapunov methods to characterize state evolution, using balancing functions to coordinate stabilizing and destabilizing factors, using graph theory to handle network coupling, and ultimately deriving mean-square exponential stability criteria.

What Exactly is the "Stability" Jointly Studied by the Two Papers?

The stability discussed here is not a vague notion that the system state does not diverge, but the stronger mean-square exponential stability. It is usually written as $\mathbb{E}\lVert x(t)\rVert^2 \le C e^{-\lambda (t-t_0)}$

Where:

• $x(t)$ represents the overall system state
• $C>0$ is a constant
• $\lambda>0$ is the decay rate

This conclusion means that the expectation of the squared system state decreases exponentially over time. In other words, the system gradually returns to the equilibrium point, and this return has a definite convergence speed.

In the context of coexisting unfavorable factors like attacks, asynchrony, stochastic disturbances, and network coupling, obtaining such a result indicates that the papers pursue a strong and rigorous stability guarantee.

How to Summarize the Two Papers in One Sentence?

They can be summarized as: Both papers revolve around the stability of complex large-scale stochastic systems under asynchronous DoS attacks, with the core problem being how to guarantee exponential stability through control or impulsive intervention when node coupling, stochastic disturbances, and asynchronous behavior coexist.

To further distinguish the focus of each paper, it can be stated more specifically: The first paper focuses on stabilization design for switched large-scale stochastic systems with delays under discrete-observation decentralized control, while the second paper focuses on unified stability analysis for large-scale stochastic networks with asynchronous impulsive effects under DoS attacks.

What Understanding Should You Take Away from This Entry Page?

After reading this document, readers should establish the following overall understanding:

• The two papers belong to the same research direction.
• The common research object is large-scale stochastic systems in asynchronous networked attack environments.
• The common goal is to guarantee the system achieves mean-square exponential stability.
• The first paper leans more towards control design in complex scenarios.
• The second paper leans more towards unified stability analysis for asynchronous impulsive systems.
• The two papers have clear methodological continuity.

Therefore, in subsequent reading, these two papers should not be viewed as isolated articles but should be understood within the same main research line. Entering subsequent detailed reading, concept clarification, and method analysis documents with this global perspective will make it easier to see the connections between different parts.