Papers

Accurately predicting directional radio maps is essential for wireless applications, yet prior approaches primarily focus on omnidirectional signals and typically treat transmitter localization and signal map reconstruction as separate tasks. In omnidirectional settings, predicting the maximum signal location often coincides with the transmitter position, which limits the need for explicit joint modeling. However, in directional propagation where angular effects, reflections, and building occlusions play critical roles, this assumption no longer holds. To address this gap, we propose SymNet, a unified framework that jointly predicts directional radio maps and transmitter locations from sparse signal measurements. SymNet incorporates a prediction head for transmitter localization alongside radio map reconstruction, enabling simultaneous learning of both tasks. This joint formulation leverages their complementary information and leads to consistent improvements over treating them separately. Experiments on challenging directional scenarios demonstrate that SymNet outperforms state-of-the-art baselines, achieving superior accuracy in both radio map reconstruction and transmitter localization.

Model predictive path integral (MPPI) control computes its inputs as cost-weighted averages over simulated trajectories. However, it typically requires many simulations, especially when low-cost trajectories are rare. We therefore formulate this computation as a quantum query problem. By replacing Gaussian input perturbations with random signs, we express the MPPI update as a ratio of averages over a finite set of perturbation sequences. We then construct quantum circuits that encode these averages as measurement probabilities, which allows quantum amplitude estimation to compute the update with fewer queries than classical sampling. Numerical examples confirm the predicted query scalings and show that a quantum advantage requires many perturbation sequences.

Artificial intelligence workloads are driving the rapid expansion of datacenter infrastructure, which imposes substantial stress on the US energy system. While high peak electricity prices are an anticipated outcome, measurable under peak hour simulations, the high off peak prices are a significantly underestimated threat. We simulate different ramping conditions on a congestible 5000-bus system, based on a modified IEEE 118-bus grid, to show that, in the presence of fast ramping loads and slow ramping generation, datacenters can aggravate latent load pockets. This results in unexpectedly high system costs during periods outside of datacenter peak. We test the datacenter effects using two distinct load conditions. We find that in the system coincident peak, coupled simulations result in up to 100% loading of slow expensive units, with an average marginal cost increase of 8%. These findings are of extreme importance as they reveal the hidden costs of preventively ramping slow generation in anticipation of datacenter load changes.

Ensuring operation of aerospace systems within prescribed flight envelope limits is a fundamental requirement for modern flight control architectures. Flight envelope protection aims to prevent violations of aerodynamic and structural constraints, thereby mitigating risks such as stall and excessive load factors. Control barrier functions (CBFs) have emerged as a principled tool for enforcing safety by ensuring that the system state remains within a prescribed safe set. In most existing approaches, safety constraints are imposed at the control input-level based on an open-loop model of the system. While this open-loop model-based CBF formulation enables modular design, it may alter the closed-loop system dynamics, potentially compromising robustness guarantees and complicating integration into existing flight control architectures. This paper proposes a closed-loop model-based control barrier function (CLM-CBF) framework for flight envelope protection. The key idea is to enforce safety at the reference-level using an explicit model of the closed-loop system, thereby preserving the stability and robustness properties of the underlying controller. This formulation enables safety filtering without modifying the control law, facilitating modular integration and retrofitting into existing systems.

The rapid expansion of AI-driven information infrastructure, particularly data centers, is placing unprecedented pressure on power systems and accelerating the pace at which new assets must interconnect with the grid. As bulk transmission expansion rolls out slowly, new loads and generation are increasingly deployed within existing network constraints. Agentic AI is urgently needed to automate the numerous and repetitive connection processes, but its maturity has not been systematically validated on complex tasks and large-scale systems. We replicate the current state of the art in agentic AI for power systems planning and evaluate it against a structured suite of nodal planning problems spanning six levels of task complexity and four grid scales. We find that only the two lowest complexity levels are solvable on some of the test grid sizes, and identify the specific capability upgrades required to close this gap. Adopting stricter testing protocols and reproducible evaluation benchmarks is essential for assessing both genuine progress and the operational readiness of agentic AI.

This paper investigates resilient control design for leader–follower vehicle platoons with mode-dependent powertrain dynamics subject to False Data Injection (FDI) attacks on vehicle-to-vehicle (V2V) communication channels. The longitudinal motion of each vehicle is described by a switched third-order model that captures changes in the powertrain dynamics across different operating modes. To estimate the attack signals that are injected into the communication channels, each vehicle is equipped with an auxiliary system, and a dedicated observer is implemented for each communication link. The resulting attack estimates are then used to mitigate the effects of the attacks through our proposed resilient controller. For attacks with bounded rates but potentially unbounded amplitudes, the closed-loop platoon is shown to be uniformly ultimately bounded. For a predecessor-following topology, string stability is established for the nominal switched platoon, while the effect of nonzero attack-estimation errors on acceleration propagation is shown to be bounded. Numerical case studies demonstrate the effectiveness of the proposed approach.

Recent advances in generative artificial intelligence have made it easier to fabricate statements and amplify political disinformation during elections. We introduce ParlaSpoof-BR, an audio deepfake dataset derived from recordings of the Brazilian Chamber of Deputies and expanded with synthetic utterances from diverse text-to-speech and voice conversion models. Using ParlaSpoof-BR, we benchmark state-of-the-art audio deepfake detectors, examine their ability to generalize to Brazilian Portuguese political speech, and investigate potential biases in their predictions. Our analysis reveals that current systems struggle to provide consistent decisions across the diversity represented in the dataset, with methodological factors (synthesis model choice, manipulation extent) dominating over demographic disparities. ParlaSpoof-BR provides a domain-specific benchmark for studying audio deepfake detection in a socially consequential and underrepresented setting, supporting the development of more robust detection systems for electoral integrity in Brazil.

As inverter-based resources (IBRs) contribute larger shares of generation in electrical power grids, quantifying system strength becomes increasingly important for identifying stability issues introduced by these devices. While admittance model based system strength metrics have been proposed to identify control interactions in systems with high levels of IBRs, these methods depend on repeated evaluations across many operating points to understand how the state of the system impacts system strength. To address this challenge, we consider the sensitivity of system stability to perturbations in the steady-state operating point, and propose system strength metrics based on sensitivities to power injections, voltages, and line admittances. Using these sensitivities we can identify changes in the system's state (e.g. a line tripping off or a generator increasing its power output) that trigger instability mechanisms. We show that these metrics provide critical insights into system stability and can be computed much faster than repeated eigenvalue calculations. We demonstrate our approach on 14-bus and 118-bus test systems to show how the metrics can be used to find remedial actions for small-signal stability issues.

Small-island developing states face accelerating sea-level rise, intensifying cyclones, and storm surge, while suffering the sparse ground instrumentation that makes timely hazard perception difficult. This paper argues for a shift from passive cellular connectivity to the Network as a Sensor, realized through a Sovereign Cognitive Digital Twin (S-CDT): a federated national digital twin whose perceptual substrate is the 6G radio interface itself. Instead of disjoint sensing systems, a nation can reuse the Integrated Sensing and Communication (ISAC) waveforms of its own network as a distributed radar mesh. We specify a six-layer S-CDT stack in which ISAC collapses the boundary between the dynamic-data and communication layers; we map the physical layer to the ETSI GR ISC 001 and 3GPP Release 19 sensing frameworks; and we formulate a belief-state control loop, an Extended Kalman Filter feeding a Proximal Policy Optimization agent, designed to absorb O-RAN telemetry delay. That loop is specified but not evaluated here. Beyond the reference architecture, we implement a reproducible, CPU-only geodata-to-ray-tracing pipeline over a 2 km study area at the Barbados Heritage District, Newton Plantation: 576 LiDAR-height buildings, a 70x70 terrain grid, and the government tower register become a Sionna RT scene. On identical geometry, median best-server path gain falls from -106 dB at 1.8 GHz to -122 dB at 10 GHz; concrete-only 28/60 GHz runs yield -130/-136 dB, with coverage contracting to line-of-sight lobes. These uncalibrated, 1x1 V-polarized simulations are framed as a towards-6G site model of the background channel, not an operational ISAC deployment. We treat data sovereignty and physical-layer zero-trust security as first-order design constraints. Barbados (166 km2, ~280k population) is the reference deployment, with the Philippines as an archipelagic generalization.

We present a novel technique for predicting the imaging quality of anomalies such as cancer cells located inside organic tissues. This technique is useful for evaluating and designing RF tomography sensors.

Optical flow characterizes motion in the visual field and is fundamental to motion perception and tracking in biological and artificial vision systems. Biological retinas extract motion efficiently through local ON/OFF pathways and parallel processing, while conventional frame-based optical flow relies on dense sampling and global computation, resulting in high latency and power consumption. To overcome these limitations, we present a pixel-level Optical Flow Sensor (OFS) integrated circuit. The design combines Dynamic Vision Sensor (DVS) ON/OFF event comparison with time-difference measurement to enable fully parallel optical flow computation on-chip. An optical-flow-specific Address-Event Representation (OF-AER) interface supports low-power, high-throughput readout. \rev{Based on the CMOS-based OFS, we further propose optical memristor-based OFS to reduce sensor power consumption and area overhead.} Experimental results show that the proposed OFS achieves a 303$\times$ reduction in power consumption compared with FPGA-accelerated DVS systems while maintaining microsecond-level latency. Moreover, by directly outputting optical flow vectors, the OFS reduces output data size by approximately 3.3$\times$, demonstrating strong potential for ultra-high-speed, low-power vision sensing applications.

This work presents a mixer-first receiver front-end (RXFE) optimized for sub-6 GHz applications with minimized local oscillator (LO) pulse overlap. A design methodology is proposed to mitigate LO overlap-induced degradation in input matching and noise figure (NF), validated using a current-mode logic (CML)-based LO generator. Implemented in TSMC 65-nm CMOS, the RXFE achieves an input return loss (S11) below -10 dB, an NF degradation of 1.08 dB, and consumes no more than 12.19 mA from a 1.2 V supply across the 1-5.6 GHz frequency range. The proposed RXFE occupies an active layout area of 302 um x 209 um.

This paper presents a subspace data-driven predictive control method for linear parameter-varying (LPV) systems. Starting from an affine LPV state-space model in innovation form, we derive a multi-step predictor that separates the effects of past data, future inputs, scheduling trajectories, and innovations. By projecting this representation onto the row span of lifted input-output-scheduling data, we obtain an asymptotically unbiased data-driven predictor that can be embedded directly in a receding-horizon control problem, without explicitly identifying an LPV model. To make the resulting LPV data-driven predictive control (DDPC) formulation tractable, we introduce an LPV extension of $\gamma$-DDPC based on an LQ factorization. This formulation fixes the number of online decision variables independently of the length of the dataset. A reduced-order predictor is then proposed to curb the exponential growth of scheduling-dependent regressors, which also relaxes the persistence-of-excitation condition. Simulation studies, including an unbalanced-disk example, show that the proposed controller achieves good tracking performance and, compared to existing LPV DDPC schemes, achieves better robustness to measurement noise and reduced computational cost, making multi-step LPV DDPC practically deployable, even with longer past horizons.

This paper a universal implementation method for continuous- and discrete-time two-parameter (2-DOF) stabilizing controllers using exclusively stable constituent blocks. The framework is valid for all regular SISO and MIMO plants, particularly those failing to satisfy the parity interlacing property where unstable controllers are unavoidable. Unlike existing algebraic approaches that compromise parametric design freedom to force sub-block stability, the proposed method preserves design precedence. By leveraging the Youla-Kučera parametrization within input-output (I/O) feedback topologies, we decouple synthesis from implementation. This allows physical performance specifications to drive the controller design, which is subsequently realized via a cyclic configuration of stable blocks. This architecture yields key engineering advantages, including localized anti-windup, simplified initialization, and improved numerical robustness.

Busbar protection is a cornerstone of industrial power system reliability, as failures at switchgear can propagate rapidly and extend restoration times. The Taiwan "303 blackout" in 2022, initiated by a short-circuit fault and aggravated by CT supervision lockout, illustrates the severe consequences of inadequate designs on busbar protection in centralized grids. Despite extensive academic and industrial guidance, practical challenges remain in busbar protection design, including CT placement, dynamic zone selection, CT saturation, and evolving fault scenarios. Standards and guides often fail to reflect operational limitations under diverse system conditions. This paper provides a structured evaluation of busbar protection (BBP) schemes in industrial applications. Leakage-to-frame, high-impedance differential (87Z), interlocking overcurrent (ILOC), electronic (EBBP), and numerical (NBBP) schemes are examined with respect to their operating principles, implementation requirements, and suitability for different bus configurations. Comparative analysis highlights clear trade-offs in different busbar protection schemes. The main contributions are: (i) systematic review of classical and modern busbar protection schemes, (ii) identification of operational limitations with various busbar protection, and (iii) comparison on the integration of CT supervision, check zones, and directional checks in NBBP to mitigate maloperations. The discussion reinforces the role of robust busbar protection design in preventing cascading outages and maintaining system reliability.

Radio-frequency interference (RFI) remains a significant threat to GNSS in safety-critical applications. Since no single mitigation method is effective for all interferers, reliable classification is needed to select appropriate countermeasures during operation. To prevent loss of lock, RFI classification must run in real time, typically on resource-constrained embedded platforms, necessitating lightweight algorithms. While prior works realize this with simple rule-based algorithms for detecting and characterizing certain types of interferers, this approach does not scale to the broad space of all possible RFI techniques, and data-driven learned algorithms are a better fit. To satisfy these constraints, this work considers pre-correlation RFI classification with an emphasis on compact algorithms that still provide high accuracy. We first introduce a new set of lightweight input features derived from the instantaneous frequency predictions of a virtual adaptive notch filter (ANF). We observe improved classification accuracy for both narrowband and broadband non-stationary RFI by combining these new features with other spectral features from prior literature. Next, we benchmark compact learned classifiers such as gradient-boosted decision trees for accurate prediction under tight compute and memory budgets. The evaluation spans a broad set of simulated and recorded RFI events, including publicly available datasets from recent studies. Finally, we measure resource utilization for our models and for representative methods from the literature, on a compact CRPA platform (EDGE Microwave HEDGE8008).

Personal privacy devices (PPDs) typically emit strong tones or swept narrowband signals to jam nearby GNSS receivers and deliberately cause loss of lock. Excision methods deployed on receivers mitigate such interferers by tracking their instantaneous frequency and removing those components in either the time domain (e.g., notch filtering) or a transform domain (e.g., Fourier-domain excision). For effective mitigation without degrading the GNSS signal, the excision location must be precise; misplacement can sometimes harm performance even more than leaving the interferer unmitigated. Trackers are therefore evaluated for both dynamic tracking performance, especially against fast-sweeping jammers, as well as steady-state estimation jitter, reflecting a fundamental tracking speed versus variance trade-off. We propose a new frequency tracking algorithm that provides a better trade-off than existing methods using first and second moments of the gradient for a complex first-order IIR notch filter under a standard output power minimization objective. We first describe the algorithm, which is inspired by the Adam optimization method widely used for optimizing neural network parameters, and its relation to existing adaptive notch filter (ANF) update rules. Next, we analyze performance over a wide range of simulated and recorded interference events, including publicly available datasets, and benchmark against state-of-the-art methods including frequency-locked-loop (FLL) based designs. Finally, we analyze resource utilization and latency, and quantify the effects of quantization and pipeline delays on the proposed method and on representative state-of-the-art baselines. Our proposed Adam-based method shows marginally higher resource usage than a vanilla ANF while providing better suppression under most scenarios.

Independent training jobs sharing a storage system write their checkpoints through the same finite bandwidth, and the resulting bursts of correlated I/O are commonly described as a self-reinforcing "checkpoint storm". We formalise the self-reinforcement as phase locking in a population of integrate-and-fire oscillators coupled through a shared resource, and show that within that model it fails. Call a resource anonymous if the rate it delivers to an active user depends on how many users are active and not on which. For identical jobs whose write is shorter than their compute interval, an anonymous resource produces no pairwise coupling at all: the two-job return map of the phase gap is the identity, under storage contention, under a shared power cap and under both, so the two-body interaction on which the Kuramoto and Mirollo-Strogatz frameworks are built is not weak here but absent. Anonymity also freezes the firing order, for any fleet size and any cap, so no trajectory reaches the synchronous state from outside it. What survives is a third-order effect: where all $N$ write windows overlap and the cap does not bind, the map is diagonal in the intervals between consecutive write starts, $a_j \mapsto ((N-j)/j)a_j$, with reciprocal spectrum and unit determinant, making synchrony a fixed point with $\lceil N/2\rceil-1$ expanding directions rather than an attractor. That determinant follows from anonymity and not from fairness: for any anonymous throughput $f$ with $f(n)\le n$ the spectrum becomes $(N-j)f(j)/(j f(N-j))$, whose product is still one. Numerically, a fleet launched at random neither locks nor clusters, and absence of locking is not absence of bursts: the upper tail of the number of concurrent writers stays above its independent-phase value. Heterogeneous jobs behind a binding cap do acquire a genuine pairwise coupling, which is where the statement stops generalising.

In this paper, a computationally lightweight approximate robust nonlinear model predictive control (NMPC) law is proposed based on a pair of input-to-state control Lyapunov function and robust control barrier function. The result builds upon and augments a recently introduced nominal infinitesimal- horizon NMPC scheme which permits small-sized quadratic programs to compute the feedback law for nonlinear constraint systems on embedded hardware in real time. Numerical experiments for nonlinear constrained spacecraft control and comparison to other robust NMPC schemes from the literature demonstrate the effectiveness of the proposed scheme.

Latent heat thermal energy storage (LHTES) using phase change materials (PCMs) is a promising solution to shifting heat supply and reducing peak demand in district heating (DH). However, the combined impacts of PCM thermophysical properties and practical control strategies on DH system-level operational and economic performance remain insufficiently understood. To bridge the research gap, this study investigates a PCM-integrated DH system with heat pump assisted waste heat recovery under an adaptive demand-driven (ADD) control strategy to enhance operational flexibility. A dynamic simulation model was developed and the system performance was evaluated against a baseline case and a rule-based control (RBC) approach based on peak-load reduction, operational cost, heat pump performance, and indoor thermal comfort. Furthermore, sensitivity analyses were conducted to examine the influence of PCM thermophysical properties on system performance. The results showed that the RBC can shift peak demand but tends to generate secondary peaks during charging periods. In contrast, the ADD strategy effectively smoothed the heat demand profile and achieved up to 5.3% peak-load reduction while maintaining thermal comfort. Sensitivity analysis revealed that a phase-change temperature of 80$^\circ$C and thermal conductivity above 2 W/(m$\cdot$K) achieved a higher peak-load reduction and improved economic performance. Despite the enhanced peak-shaving capability achieved by the proposed control strategy, the system exhibited a payback period of 25.3 years, indicating that further cost reductions and supportive market incentives are required. Nevertheless, the proposed approach provides significant potential for enhancing DH flexibility and supporting the transition toward future low-carbon energy systems.

Semantic communication is an emerging paradigm that can preserve the meaning of data during transmission. However, human users are often interested in specific semantic content based on their intent, and users' intent is often not considered in current semantic coding design. Moreover, most of the existing semantic models are fine-tuned using specific datasets, which limits their generalization capability. Furthermore, how to prioritize semantically important features in large-scale multiple-input multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM) systems remains largely unexplored. To address the aforementioned challenges, in this paper, we propose a generalized query-oriented image semantic coding (QO-ISC) framework. In the proposed framework, the transmitter extracts features which are relevant to the user's query and the receiver reconstructs an image based on those features. We use a pretrained large artificial intelligence (AI) model (LAM) to enhance general feature representations. We develop a semantic-aware hybrid beamforming (SA-HBF) algorithm to prioritize semantically important features for large-scale MIMO-OFDM system. When evaluated on unseen object categories within the dataset, simulation results show that our proposed generalized QO-ISC framework achieves better performance than the traditional codec and two state-of-the-art semantic coding schemes.

The gradual substitution of conventional synchronous generators by converter-interfaced renewable energy sources raises concerns about the reduction of conventional inertia in electric power systems and the ensuing threat to their stability. In this regard, grid-forming voltage source converters have been proposed as a key solution to address this challenge. Although a growing body of literature addresses DC-link voltage regulation in grid-forming converters, most existing approaches implicitly assume an ideal and unconstrained DC power source. As a result, the dynamic response and operational limits of the primary energy source, which can critically shape the available DC-side power during transients, are rarely modeled or explicitly accounted for in the design of the DC-link control. This paper demonstrates that incorporating these aspects at the design stage reduces the risk of converter disconnection from the power grid under sudden power imbalances, while enhancing system resilience. A systematic methodology based on a Genetic Algorithm is proposed to tune the control parameters. The performance of the proposed control is validated by simulation using a detailed electromagnetic transient model.

Despite advances in policy pretraining, embodied AI systems can plateau during task-specific fine-tuning as uniform scenario collection encounters fewer of the remaining failures. We address this problem with a policy-level recursive self-improvement loop: execution outcomes train a criticality world model, whose risk scores guide scenario collection for specialist training. To account for the resulting shift in scenario frequencies, weighted resampling approximately corrects the collection bias. Because specialized updates can weaken nominal behavior, a risk-based gate selects between the specialist and a frozen nominal policy at inference. We extend this procedure to reinforcement learning and behavior cloning. A sampling analysis characterizes the correction, while a controlled study examines the trade-off between critical and nominal coverage. Across five embodied domains, the composed systems reduce failure rates by 59–67% for locomotion and manipulation and 8–25% for VLA benchmarks relative to their respective baselines.

This paper presents a Cramér-Rao lower bound (CRLB)-driven beamforming (BF) and power allocation (PA) framework for cooperative integrated sensing and communication (ISAC) networks, where a set of multi-antenna base stations (BSs) jointly serve multiple users and simultaneously perform multi-static target estimation. In our design, we investigate how the position error bound (PEB) and velocity error bound (VEB) can be exploited and incorporated into BF and PA optimization. The design leveraging PEB and VEB as metrics directly characterizes sensing accuracy. First, we propose a semidefinite programming (SDP)-based BF, where the PEB and VEB constraints are handled by double Schur complements, and the rank-1 constraints of the BF covariance matrices are relaxed by semidefinite relaxation (SDR), whose tightness is proved using the Karush-Kuhn-Tucker (KKT) conditions. Addressing the complexity, we further develop a two-stage PA algorithm, where the communication and sensing beams are formed by the regularized zero-forcing (RZF) and null-space projection (NSP) methods, respectively, and the communication and sensing PAs are solved sequentially by second-order cone programming (SOCP). Although the PA algorithm sacrifices the degrees of freedom, its performance shows a slight gap compared to BF in the simulation, while the execution time of around 25 ms demonstrates its applicability in dynamic environments.

Near-field covariance matrices are central to local- ization, covariance-aware estimation, and linear MMSE filtering in large-aperture arrays, but Gaussian position uncertainty requires costly numerical averaging of nonlinear spherical-wave steering vectors. This letter proposes a low-complexity framework for two-dimensional near-field Gaussian covariance approxima- tion. By writing the quadrature covariance as RQ = HHH , the dominant covariance spectrum is obtained from a reduced observation representation, avoiding full MxM eigendecom- position. A self-calibrated non-reference spectral-error estimator is further introduced using only grid-to-grid dominant-spectrum differences. Numerical results show accurate convergence track- ing and substantial complexity reduction.

The inverse synthesis problem of selecting a minimal leader set to guarantee strong structural controllability (SSC) in matrix-weighted networks remains an unresolved NP-hard challenge. This paper proposes a rigorous mathematical framework to solve this. We prove that structural uncontrollability stems exclusively from dimension-specific reachability isolation and topological symmetry equivalence. To overcome these bottlenecks, we formulate a two-phase synthesis: a reachability prerequisite to identify structural roots, followed by three distinct symmetry-breaking algorithms (Greedy Weisfeiler-Lehman Selection, Submodular Bound Maximization, and Partition Entropy Maximization). Mathematical proofs guarantee immunity to invariant subspaces and structural dilation, validated by extensive numerical evaluations across diverse topologies.

This paper addresses the distributed multi-sensor fusion state estimation and consensus filtering for Boolean control networks (BCNs). Existing centralized multi-sensor estimation schemes for stochastic BCNs have limitations of high communication costs and single-point failures, and continuous-state consensus algorithms are difficult to extend to discrete logical systems. By integrating probability measure transformation, semi-tensor product and stochastic approximation, a distributed stochastic average consensus filter is proposed. Moreover, the almost sure convergence of the algorithm is proved by martingale convergence theorem and perturbed stochastic Lyapunov functions. The proposed framework realizes global state estimation via local communication, avoiding the defects of centralized architectures.

We present ReGenVC, an end-to-end generative video codec that compresses talking-head video to an ultra-low bitrate and decodes it in real time. The encoder reduces a source clip to a compact bitstream – a neurally compressed first frame, per-frame pose keypoints, and metadata – totaling about 26 kB for a 77-frame sequence. The decoder is a four-step distilled diffusion transformer that reconstructs the video conditioned on the transmitted pose and reference frame. Compared with x264/x265, ReGenVC reduces the bitrate to roughly one tenth of that required by traditional codecs (about 26 kB vs. 250–280 kB for essentially artifact-free reconstruction); at a matched ultra-low bitrate, conventional codecs collapse into blocking artifacts while ReGenVC stays sharp by exploiting a strong generative prior. The central obstacle to deploying such a codec is decoder latency: multi-step sampling with transformer and VAE components is too slow for interactive use. We make the decoder real-time through four-step distillation and three model-preserving system techniques: (i) eight-GPU unified sequence parallelism (Ulysses & Ring), (ii) a spatially-split VAE, and (iii) a three-stage overlapped pipeline; an analytical timing model characterizes the real-time feasibility region. On an 8-GPU node, the system sustains 24 fps output (972 ms per 25-frame window, within the 1000 ms budget), enabling a live browser stream without observed frame underruns. A hybrid CPU-GPU deployment further runs the encoder on the CPU at 24 fps and offloads the decoder-side one-shot conditioning encoders to the CPU, reducing the per-GPU memory peak from 21.1 GB to about 7.7 GB. To our knowledge, ReGenVC is the first end-to-end generative video codec to combine ultra-low-bitrate encoding with real-time decoding on an 8-GPU system.

Actuator dead-zones are a common and troublesome nonlinearity in motion control: a band of commanded effort over which the plant does not respond, leaving a steady-state offset or a limit cycle. This paper proposes a data-driven architecture that compensates such mismatches without a model of the plant and without any parameterization of the dead-zone. The central idea is to identify, alongside the velocity-form predictor used for control, a second absolute subspace predictor. Because the absolute predictor carries no integral action, it behaves as a data-driven steady-state sensor, so a persistent actuator mismatch appears as a proportional prediction residual. Embedding this residual as a proxy in a behavioral Hankel matrix reduces the mismatch estimate to a single fixed orthogonal projection evaluated online, with no dynamic estimator, no injected probing signal, and no run-time prediction-error computation. Integrated into a subspace predictive controller, the framework is shown to be recursively feasible and practically input-to-state stable, and it recovers offset-free tracking once the dead-band traversal settles. The approach is validated in real time on a sixth-order, lightly damped Quanser multi-DOF torsion system, whose complex-conjugate poles give a lightly damped open-loop response, achieving offset-free tracking across a $\pm 0.18$\,V actuator dead-band. A second study on a high-precision power amplifier shows that the same architecture rejects dead-time-induced nonlinearities in fast-switching power electronics.

Indirect data-driven predictive control methods often suffer under process noise and data scarcity. This paper introduces Projection-Regularized Predictive Control (PRPC), retaining the fundamental-lemma weight vector via a regularized projection analytically condensed into an efficient, fixed-dimension covariance update. A rigorous bias–variance analysis proves PRPC strictly reduces prediction error under process noise (errors-in-variables) and structural rank deficiencies compared to unregularized subspace methods. We leverage these properties to develop an adaptive sliding-window controller for linear time-varying (LTV) systems. To guarantee safety despite closed-loop data correlations, we derive a uniform-in-time, finite-sample confidence bound on the empirical predictor using vector-valued martingale concentration inequalities. Embedding this statistical uncertainty radius into a dynamically tightened constraint set rigorously ensures robust recursive feasibility and Input-to-State practical Stability (ISpS) with high probability. Simulations on LTI and LTV benchmarks demonstrate real-time tractability and strict constraint satisfaction.