arXiv subjects

James Lam

Publications and source records attributed to James Lam.

11 recordsLinked to original sources

arXiv2025

Distributed Affine Formation Control of Linear Multi-agent Systems with Adaptive Event-triggering

Concerning general multi-agent systems with limited communication, this paper proposes distributed formation control protocols under adaptive event-triggered schemes to operate affine transformations of nominal formations. To accommodate more practical system mechanics, we develop an event-triggered controller that drives the leader to a desired state by bringing in the compensation term. Based on triggering instants' state information, an affine formation control method with adaptive event-triggering is designed for each follower, making the whole protocol effective in refraining from successive communication while not relying on predefined global information. In particular, mitigating the effect of partial state availability, an output-based control solution is presented to expand the protocol's serviceable range. Finally, we perform numerical simulations on the formation and its affine transformations to verify the effectiveness of the control protocol and the feasibility of the event-triggered mechanism.

eess.SY↗

arXiv2025

An Online Cross-layered Defense Strategy with Bandwidth Allocation for Multi-channel Systems under DoS Attacks

This paper proposes an online cross-layered defense strategy for multi-channel systems with switched dynamics under DoS attacks. The enabling condition of a channel under attacks is formulated with respect to attack flow and channel bandwidth, then a new networked control system model bridging the gap between system dynamics and network deployment is built. Based on this, the cross-layered defense strategy is proposed. It jointly optimizes the controller gain and bandwidth allocation of channels according to the real-time attack flow and system dynamics, by solving a mixed-integer semidefinite programming online. A smart enumeration algorithm for non-convex bi-level optimization is proposed to analyze the stability under the strategy. Numerical examples are given to illustrate the high resilience from the cross-layered feature.

eess.SY↗

arXiv2025

Sparse Spectrahedral Shadows for State Estimation and Reachability Analysis: Set Operations, Validations and Order Reductions

Set representations are the foundation of various set-based approaches in state estimation, reachability analysis and fault diagnosis. In this paper, we investigate spectrahedral shadows, a class of nonlinear geometric objects previously studied in semidefinite programming and real algebraic geometry. We demonstrate spectrahedral shadows generalize traditional and emerging set representations like ellipsoids, zonotopes, constrained zonotopes and ellipsotopes. Analytical forms of set operations are provided including linear map, linear inverse map, Minkowski sum, intersection, Cartesian product, Minkowski-Firey Lp sum, convex hull, conic hull and polytopic map, all of which are implemented without approximation in polynomial time. In addition, we develop set validation and order reduction techniques for spectrahedral shadows, thereby establishing spectrahedral shadows as a set representation applicable to a range of set-based tasks.

eess.SY↗

arXiv2024

Global prescribed-time control of a class of uncertain nonholonomic systems by smooth time-varying feedback

This paper investigates the prescribed-time smooth control problem for a class of uncertain nonholonomic systems. With a novel smooth time-varying state transformation, the uncertain chained nonholonomic system is reformulated as an uncertain linear time-varying system. By fully utilizing the properties of a class of parametric Lyapunov equations and constructing time-varying Lyapunov-like functions, smooth time-varying high-gain state and output feedback controllers are designed. The states and controllers are proven to converge to zero at any prescribed time. The proposed smooth time-varying method combines the advantage of a time-varying high-gain function, which enhances control performance, and a smooth time-varying function that can drive the states to zero at the prescribed time. The effectiveness of the proposed methods is verified by a numerical example.

math.OC↗

arXiv2023

Semi-supervised Domain Adaptation on Graphs with Contrastive Learning and Minimax Entropy

Label scarcity in a graph is frequently encountered in real-world applications due to the high cost of data labeling. To this end, semi-supervised domain adaptation (SSDA) on graphs aims to leverage the knowledge of a labeled source graph to aid in node classification on a target graph with limited labels. SSDA tasks need to overcome the domain gap between the source and target graphs. However, to date, this challenging research problem has yet to be formally considered by the existing approaches designed for cross-graph node classification. This paper proposes a novel method called SemiGCL to tackle the graph \textbf{Semi}-supervised domain adaptation with \textbf{G}raph \textbf{C}ontrastive \textbf{L}earning and minimax entropy training. SemiGCL generates informative node representations by contrasting the representations learned from a graph's local and global views. Additionally, SemiGCL is adversarially optimized with the entropy loss of unlabeled target nodes to reduce domain divergence. Experimental results on benchmark datasets demonstrate that SemiGCL outperforms the state-of-the-art baselines on the SSDA tasks. The source codes of SemiGCL are publicly available at https://github.com/ JiarenX/SemiGCL.

cs.LG↗

arXiv2022

Differential Private Discrete Noise Adding Mechanism: Conditions, Properties and Optimization

Differential privacy is a standard framework to quantify the privacy loss in the data anonymization process. To preserve differential privacy, a random noise adding mechanism is widely adopted, where the trade-off between data privacy level and data utility is of great concern. The privacy and utility properties for the continuous noise adding mechanism have been well studied. However, the related works are insufficient for the discrete random mechanism on discretely distributed data, e.g., traffic data, health records. This paper focuses on the discrete random noise adding mechanisms. We study the basic differential privacy conditions and properties for the general discrete random mechanisms, as well as the trade-off between data privacy and data utility. Specifically, we derive a sufficient and necessary condition for discrete epsilon-differential privacy and a sufficient condition for discrete (epsilon, delta)-differential privacy, with the numerical estimation of differential privacy parameters. These conditions can be applied to analyze the differential privacy properties for the discrete noise adding mechanisms with various kinds of noises. Then, with the differential privacy guarantees, we propose an optimal discrete epsilon-differential private noise adding mechanism under the utility-maximization framework, where the utility is characterized by the similarity of the statistical properties between the mechanism's input and output. For this setup, we find that the class of the discrete noise probability distributions in the optimal mechanism is Staircase-shaped.

cs.CR↗

arXiv2021

Adversarially Regularized Graph Attention Networks for Inductive Learning on Partially Labeled Graphs

The high cost of data labeling often results in node label shortage in real applications. To improve node classification accuracy, graph-based semi-supervised learning leverages the ample unlabeled nodes to train together with the scarce available labeled nodes. However, most existing methods require the information of all nodes, including those to be predicted, during model training, which is not practical for dynamic graphs with newly added nodes. To address this issue, an adversarially regularized graph attention model is proposed to classify newly added nodes in a partially labeled graph. An attention-based aggregator is designed to generate the representation of a node by aggregating information from its neighboring nodes, thus naturally generalizing to previously unseen nodes. In addition, adversarial training is employed to improve the model's robustness and generalization ability by enforcing node representations to match a prior distribution. Experiments on real-world datasets demonstrate the effectiveness of the proposed method in comparison with the state-of-the-art methods. The code is available at https://github.com/JiarenX/AGAIN.

cs.LG↗

arXiv2019

Geometric Programming for Optimal Positive Linear Systems

This paper studies the parameter tuning problem of positive linear systems for optimizing their stability properties. We specifically show that, under certain regularity assumptions on the parametrization, the problem of finding the minimum-cost parameters that achieve a given requirement on a system norm reduces to a \emph{geometric program}, which in turn can be exactly and efficiently solved by convex optimization. The flexibility of geometric programming allows the state, input, and output matrices of the system to simultaneously depend on the parameters to be tuned. The class of system norms under consideration includes the $H^2$ norm, $H^\infty$ norm, Hankel norm, and Schatten $p$-norm. Also, the parameter tuning problem for ensuring the robust stability of the system under structural uncertainties is shown to be solved by geometric programming. The proposed optimization framework is further extended to delayed positive linear systems, where it is shown that the parameter tunning problem jointly constrained by the exponential decay rate, the $\mathcal L^1$-gain, and the $\mathcal L^\infty$-gain can be solved by convex optimization. The assumption on the system parametrization is stated in terms of posynomial functions, which form a broad class of functions and thus allow us to deal with various interesting positive linear systems arising from, for example, dynamical buffer networks and epidemic spreading processes. We present numerical examples to illustrate the effectiveness of the proposed optimization framework.

math.OC↗

arXiv2013

Stability Analysis of Continuous-Time Switched Systems with a Random Switching Signal

This paper is concerned with the stability analysis of continuous-time switched systems with a random switching signal. The switching signal manifests its characteristics with that the dwell time in each subsystem consists of a fixed part and a random part. The stochastic stability of such switched systems is studied using a Lyapunov approach. A necessary and sufficient condition is established in terms of linear matrix inequalities. The effect of the random switching signal on system stability is illustrated by a numerical example and the results coincide with our intuition.

eess.SY↗

arXiv2012

Positive Definite Solutions of the Nonlinear Matrix Equation $X+A^{\mathrm{H}}\bar{X}^{-1}A=I$

This paper is concerned with the positive definite solutions to the matrix equation $X+A^{\mathrm{H}}\bar{X}^{-1}A=I$ where $X$ is the unknown and $A$ is a given complex matrix. By introducing and studying a matrix operator on complex matrices, it is shown that the existence of positive definite solutions of this class of nonlinear matrix equations is equivalent to the existence of positive definite solutions of the nonlinear matrix equation $W+B^{\mathrm{T}}W^{-1}B=I$ which has been extensively studied in the literature, where $B$ is a real matrix and is uniquely determined by $A.$ It is also shown that if the considered nonlinear matrix equation has a positive definite solution, then it has the maximal and minimal solutions. Bounds of the positive definite solutions are also established in terms of matrix $A$. Finally some sufficient conditions and necessary conditions for the existence of positive definite solutions of the equations are also proposed.

math.NA↗

arXiv2012

Toward Solution of Matrix Equation X=Af(X)B+C

This paper studies the solvability, existence of unique solution, closed-form solution and numerical solution of matrix equation $X=Af(X) B+C$ with $f(X) =X^{\mathrm{T}},$ $f(X) =\bar{X}$ and $f(X) =X^{\mathrm{H}},$ where $X$ is the unknown. It is proven that the solvability of these equations is equivalent to the solvability of some auxiliary standard Stein equations in the form of $W=\mathcal{A}W\mathcal{B}+\mathcal{C}$ where the dimensions of the coefficient matrices $\mathcal{A},\mathcal{B}$ and $\mathcal{C}$ are the same as those of the original equation. Closed-form solutions of equation $X=Af(X) B+C$ can then be obtained by utilizing standard results on the standard Stein equation. On the other hand, some generalized Stein iterations and accelerated Stein iterations are proposed to obtain numerical solutions of equation equation $X=Af(X) B+C$. Necessary and sufficient conditions are established to guarantee the convergence of the iterations.

math.NA↗