arXiv subjects

Takumi Shinohara

Publications and source records attributed to Takumi Shinohara.

10 recordsLinked to original sources

arXiv2026

Data-Driven Unknown Input Reconstruction for MIMO Systems with Convergence Guarantees

In this paper, we consider data-driven reconstruction of unknown inputs for deterministic, linear time-invariant (LTI) multiple-input multiple-output (MIMO) systems. We propose a novel autoregressive estimator based on a constrained least-squares formulation over Hankel matrices, splitting the problem into an output-consistency constraint and an input-history-matching objective. Our method relies on previously recorded input--output data to represent the system, but does not require knowledge of the true input to initialize the algorithm. We show that the proposed estimator is strictly stable if and only if all the invariant zeros of the trajectory-generating system lie strictly inside the unit circle, which can be verified purely from input and output data. This mirrors existing results from model-based input reconstruction and closes the gap between model-based and data-driven settings. Lastly, we provide numerical examples to demonstrate the theoretical results.

eess.SY↗

arXiv2026

Security Index from Input/Output Data: Theory and Computation

The concept of a security index quantifies the minimum number of components that must be compromised to carry out a stealth attack. This metric enables system operators to assess the security risk of each component and implement countermeasures accordingly. In this paper, we introduce a data-driven security index that can be computed solely from input/output data when the system model is unknown. We show a sufficient condition under which the data-driven security index coincides with the model-based security index, which implies that the exact risk level of each component can be identified solely from data. We also provide an algorithm for computing the data-driven security index.

eess.SY↗

arXiv2026

Data-Driven Control from Poisoned Data: Fundamental Limitations and Secure DeePC

We study a data-driven control problem in the presence of arbitrary data poisoning attacks. We assume that a subset of offline output data is stored in unprotected locations and may be poisoned by an adversary. We first establish fundamental limitations for data-driven control arising from such poisoned data: poisoning attacks are not detected/identified from the dataset alone; unprotected data are non-informative for controller design with worst-case guarantees; and hard constraints on unprotected outputs are not certifiable. Motivated by these limitations and the data-enabled predictive control (DeePC) technique, we propose Secure DeePC, a data-driven control algorithm that is resilient against poisoning attacks. It first runs output-truncated DeePC using only the protected dataset until the online input becomes persistently exciting. It then uses online measurements to reconstruct the partial offline dataset, and finally returns to full-output DeePC. Secure DeePC achieves MPC-equivalent performance in finite time almost surely under certain conditions. Simulation results illustrate the efficacy of the proposed framework against poisoning attacks.

eess.SY↗

arXiv2026

Distributed Resilient State Estimation and Control with Strategically Implemented Security Measures

This paper addresses the problem of distributed resilient state estimation and control for linear time-invariant systems in the presence of malicious false data injection sensor attacks and bounded noise. We consider a system operator (defender) capable of deploying cybersecurity measures to counteract the sensor compromises. Although such measures enhance resilience against adversarial attacks, they may incur substantial costs; hence, it is crucial to select countermeasures to balance resilience gains and cost efficiency strategically. We first demonstrate that the system's resilience against attacks is maximized through the appropriate implementation of security measures, implying that no attacker can execute undetectable sensor attacks. Building on this analysis, we propose an algorithm that identifies the optimal security measure. While determining this measure is NP-hard in general, we also derive sufficient conditions under which efficient computation is feasible. Furthermore, we develop a distributed resilient state estimation and control scheme informed by the optimal security measure and establish conditions that guarantee bounded estimation and control errors. Finally, we validate the efficacy of our approach via numerical simulations of a vehicle platooning scenario.

eess.SY↗

arXiv2026

Data-Driven Resilience Assessment against Sparse Sensor Attacks

We develop a data-driven framework for assessing the resilience of linear time-invariant systems against malicious false-data-injection sensor attacks. Leveraging sparse observability, we propose data-driven resilience metrics and derive necessary and sufficient conditions for two data-availability scenarios. For attack-free data, we show that when a rank condition holds, the resilience level can be computed exactly from the data alone, without prior knowledge of the system parameters. We then extend the analysis to the case where only poisoned data are available and show that the resulting assessment is necessarily conservative. For both scenarios, we provide algorithms for computing the proposed metrics and show that they can be computed in polynomial time under an additional spectral condition. A numerical example illustrates the efficacy and limitations of the proposed framework.

eess.SY↗

arXiv2026

Detection and Identification of Sensor Attacks Using Partially Attack-Free Data

In this paper, we investigate data-driven attack detection and identification in a model-free setting. We consider a practically motivated scenario in which the available dataset may be compromised by malicious sensor attacks, but contains an unknown, contiguous, partially attack-free interval. The control input is assumed to include a small stochastic watermarking signal. Under these assumptions, we establish sufficient conditions for attack detection and identification from partially attack-free data. We also develop data-driven detection and identification procedures and characterize their computational complexity. Notably, the proposed framework does not impose a limit on the number of compromised sensors; thus, it can detect and identify attacks even when all sensor outputs are compromised outside the attack-free interval, provided that the attack-free interval is sufficiently long. Finally, we demonstrate the effectiveness of the proposed framework via numerical simulations.

eess.SY↗

arXiv2023

Probing warm and mixed dark matter models using lensing shift power spectrum

We argue that the lensing power spectrum of astrometric shift (lensing shift power spectrum) is a powerful tool of the clustering property of dark matter on subgalactic scales. First we give the formalism to probe the nature of dark matter by using the lensing shift power spectrum. Then, leveraging recent measurements of the lensing shift power spectrum on an angular scale of approximately $1~$arcsec towards the gravitationally lensed quasar MG$\,$J0414+0534 at the redshift of $z_S=2.639$, we place constraints on the mass of warm dark matter (WDM) particles $m_{\rm WDM}$ and their fraction in a mixed dark matter (MDM) model $r_{\rm WDM}$, in which WDM and cold dark matter coexist. Although the constraint derived from the above single lensing system is not as strong as the existing constraints, as we show in this paper, the lensing shift power spectrum has a great potential to obtain much tighter constraints on WDM and MDM models through future observations, highlighting the importance of well-controlled systematic error considerations for achieving enhanced precision.

astro-ph.CO↗

arXiv2023

Generalized early dark energy and its cosmological consequences

We investigate cosmological consequences of a generalized early dark energy (EDE) model where a scalar field behaves as dark energy at various cosmological epochs for a broad range of parameters such as the energy scale and the initial field value. We consider power-law and axion-type potentials for such an EDE field and study how it affects the cosmological evolution. We show that gravitational wave background can be significantly enhanced to be detected in future observations such as LISA and DECIGO in some parameter space. Implications of the EDE model are also discussed for a scenario where a blue-tilted inflationary tensor power spectrum can explain the recent NANOGrav 15-year signal. We argue that the bounds on the reheating temperature can be relaxed compared to the case of the standard thermal history.

astro-ph.CO↗

arXiv2023

Supermassive primordial black holes: a view from clustering of quasars at $z \sim 6$

We investigate a scenario where primordial black holes (PBHs) can be the progenitors of supermassive black holes (SMBHs) observed at $z\sim6$. To this end, we carried out clustering analysis using a sample of 81 quasars at $5.88 <z<6.49$, which is constructed in Subaru High-$z$ Exploration of Low-Luminosity Quasars (SHELLQs) project, and 11 quasars in the same redshift range selected from the literature. The resulting angular auto-correlation function (ACF) can be fitted to a power-law form of $ω_θ= 0.045^{+0.114}_{-0.106}~θ^{-0.8}$ over a scale of $0.2\!-\!10$ degrees. We compare the ACF of the quasars to that predicted for the PBH model at $z\sim 6$ and found that such a scenario is excluded for a broad range of parameter space, from which we can conclude that a scenario with PBHs as SMBHs is not viable. We also discuss a model in which SMBHs at $z \sim 6$ originate from the direct collapse of PBH clumps and argue that the observed ACF excludes such a scenario in the context of our PBH model.

astro-ph.CO↗

arXiv2021

Angular correlation as a novel probe of supermassive primordial black holes

We investigate the clustering property of primordial black holes (PBHs) in a scenario where PBHs can explain the existence of supermassive black holes (SMBHs) at high redshifts. We analyze the angular correlation function of PBHs originating from fluctuations of a spectator field which can be regarded as a representative model to explain SMBHs without conflicting with the constraint from the spectral distortion of cosmic microwave background. We argue that the clustering property of PBHs can give a critical test for models with PBHs as the origin of SMBHs and indeed show that the spatial distribution of PBHs in such a scenario is highly clustered, which suggests that those models may be disfavored from observations of SMBHs although a careful comparison with observational data would be necessary.

astro-ph.CO↗