arXiv subjects

Ajeet Kumar

Publications and source records attributed to Ajeet Kumar.

At least 19 recordsLinked to original sources

arXiv2026

A Comparative Study of Coherent and Action-Detected 2D Electronic Spectroscopies of a Multichromophore Photosynthetic System

Coherent two-dimensional electronic spectroscopy (C-2DES) is a powerful tool for resolving excitonic structure and dynamics in light-harvesting complexes with dense energy-level manifolds. In contrast, action-detected 2DES has seen limited use on such systems, and its capacity to reveal electronic structure and dynamics remains unclear. Here we compare C-2DES and fluorescence-detected 2DES (F-2DES) on the major plant light-harvesting complex (LHCII) at 78 K. C-2DES resolves excitonic features and their spectral evolution, mapping energy-transfer pathways. F-2DES shows strong cross-peaks, but these are largely time-invariant and dominated by incoherent mixing, obscuring the dynamics. C-2DES is therefore preferred for resolving ultrafast energy transfer in light-harvesting complexes and other large interacting networks. Combined, the two measurements directly quantify the suppression of population dynamics by incoherent mixing, confirming that nearly the entire LHCII trimer network contributes. We also identify pulse-shaper nonlinearities in F-2DES as spurious features, highlighting the need for rigorous artifact suppression.

physics.chem-ph↗

arXiv2026

Finite strain homogenization of periodic rod networks with application to semi-flexible biopolymers

In this work, we adopt a finite strain computational homogenization approach to characterize the response of semi-flexible biopolymer networks modeled as idealized 8- and 14-chain periodic networks. We use the geometrically exact special Cosserat rod theory to model the microscale fibers forming these 8- and 14-chain networks. This allows us to capture arbitrarily large microscale deformations. Both macroscopic strain- and stress-driven homogenization are performed to study the macroscopic uniaxial tension, compression and simple shear responses. Several phenomena unique to biopolymer networks are recovered such as strain-stiffening and volume shrinkage under uniaxial tension, softening under compression and reverse Poynting effect under simple shear. We find that nonlinearity and non-affine deformation at microscale, especially bending and buckling of microscale fibers, plays an important role in these phenomena. We obtain the postbuckled solutions of the homogenization problem using a nonlinear, imperfection-free path following approach and also check for their stability. We further compare our homogenization results with experimental data for uniaxial tension and compression of biofilament networks and find good agreement. When the fibers are replaced by helical rods in the 8-chain unit cell, we are also able to capture the enlarged stretching behaviour as shown in recently fabricated compliant metastructures.

cond-mat.soft↗

arXiv2026

Robust Nonclassical magnon pair generation and Cauchy-Schwarz inequality violation in a hybrid electromagnonic system

Magnonic systems are a promising resource for quantum technologies because of their ability to couple significantly disparate quantum platforms and the generation of nonclassical magnon states through magnon blockade has attracted growing attention. In practice, however magnon blockade relies on a destructive interference that is easily spoiled by fabrication-induced coupling asymmetries. Here we show that the relative phase between two magnon drives acts as an active compensation knob that restores this interference even when the couplings are mismatched. We study two Kittel magnon modes in a hybrid system in which a superconducting qubit coupled to a common cavity mode mediates the inter-mode interaction and we find that tuning the drive phase produces both magnon blockade and a strong violation of the classical Cauchy--Schwarz inequality. We derive the analytic condition for the phase that compensates a given coupling asymmetry and confirm it against exact numerical simulations. The drive phase thereby serves as a control knob that switches the system between classical and quantum regimes. Our results enable phase-controlled magnonic quantum information processing.

quant-ph↗

arXiv2026

Closing the Social-Semantic Gap: SPSD for Edge-Based Prompt Compression in Cloud LLM Inference

The prefill stage of Large Language Model (LLM) inference is a growing contributor to cloud-scale energy cost. Many consumer-support and conversational prompts contain social scaffolding: politeness markers, apologetic preamble, repetition, and rapport-building language that is important for human communication but carries low marginal information for machine reasoning. We call this discrepancy the Social-Semantic Gap. We present SPSD (Sentiment Preserving Semantic Distillation), an edge-based pipeline that compresses user prompts using a 4-bit quantised Small Language Model before transmission to a cloud-deployed LLM. Evaluation on a 248-prompt corpus using Gemma-2-2B-Instruct (Q4_K_M) as the SLM and Llama-3.1-8B-Instruct as the cloud evaluation model yields a mean input token saving of 99.9 tokens per distilled call, with all 146 distilled calls yielding positive savings. Response quality, assessed by blind LLM-as-judge scoring across 121 pairs, is non-inferior to the raw path within a pre-specified 1-point margin on a 15-point rubric; the judge awarded 43 percent ties, 28 percent distilled wins, and 29 percent raw wins. Cosine similarity is mixed: mean 0.682, median 0.712, with 54.1 percent of pairs above the 0.70 reference threshold. Safety-critical domains are conservatively routed to passthrough via rule-based gates. Per-call net energy saving is estimated at 70-270 uWh under stated assumptions. SPSD shows that on-device prompt distillation can reduce cloud LLM input-token cost while preserving response quality within a practical non-inferiority margin.

cs.LG↗

arXiv2026

Transformer-Based Heartbeat Monitoring with FMCW Radar Under Random Body Motion

Millimeter-wave Frequency Modulated Continuous Wave (FMCW) radar enables contactless cardiac monitoring, but heartbeat estimation becomes challenging when respiration and random body motion (RBM) distort the radar signal. In this paper, we propose a hybrid framework for 77 GHz FMCW radar that combines model-based signal processing with a Convolutional Neural Network (CNN)-Transformer network. The first block extracts chest displacement and constructs meaningful high-level motion features from raw radar data, while the second block reconstructs a photoplethysmography (PPG)-like signal from the extracted features. In this study, a synchronized PPG signal is used as the ground truth for heartbeat monitoring in supervised training. The method is evaluated following the IEEE AESS Radar Challenge Problem I protocol using the official datasets and figures of merit across three motion scenarios: stationary, deep breathing, and RBM. Results show that the proposed architecture reliably reconstructs the PPG signal in all scenarios, achieving high fidelity in controlled conditions and maintaining robust performance under motion. This enables reliable average heart rate (AHR) and heart rate variability (HRV) estimation even where benchmark methods fail, and leads to the highest total score among the compared approaches.

eess.SP↗

arXiv2026

Homogenization of one-dimensional periodic rod networks as special Cosserat rods

One-dimensional periodic rod networks are the structures that are obtained by periodically assembling its microstructural unit, a network of rods itself, in just one direction. In this work, we present a scheme for obtaining the nonlinear constitutive response of such structures when homogenized macroscopically as a continuum rod. To accurately capture arbitrary and large deformations, the geometrically exact special Cosserat rod theory is used for modeling the rod at both micro- and macroscales. By assuming the periodic structure to be strained uniformly, at macroscale, along its arc length, the full structure problem is reduced to just that of its microstructural unit but subjected to helically periodic boundary condition. The microscale problem, consisting of a network of rods and formulated in a variational setting, is solved in the presence of rod-joint constraints and helically periodic boundary conditions. The expressions for the macroscale rod's stress resultants, i.e. internal contact force and moment, and stiffnesses are then obtained. Finally, several numerical examples having different microstructural units are presented to demonstrate our method. First, the results corresponding to simpler square and cross microstructural units are presented and validated with the existing literature. Square microstructural units having helical constituent rods are then taken up which have application as artificial muscle fiber material. Eventually, homogenization of auxetic tubular metamaterials is performed. It is shown how various design parameters of these microstructural units can be tuned to obtain the desired macroscopic response.

cond-mat.mtrl-sci↗

arXiv2026

Neural network model for mathematical programming problems with complementary constraints

In this paper, we propose a a gradient-based neural network model to solve the mathematical programming problems with complementary constraints (MPCC). In order to facilitate tractable optimization, the problem MPCC is transformed via a regularized approach into a relaxed nonlinear optimization problem NLP($\beta$). After that employing the penalty function and neural network model an estimate of the optimal solution of the problem NLP($\beta$) is obtained. On the basis of Lyapunov stability theory and LaSalle invariance principle, the equilibrium point of proposed neural network is theoretically proven to be asymptotically stable and capable to generate optimal solution of the problem MPCC. Further, we demonstrate the performance and dynamic behavior of the proposed neural network through various illustrative examples and its effectiveness via theoretical and numerical experiments.

math.OC↗

arXiv2025

Measures to characterise Approximate Mutually Unbiased Bases

Mutually Unbiased bases has various application in quantum information procession and coding theory. There can be maximum d + 1 MUBs in C^d and d/2 +1 MUBs in R^d. But , over R^d MUBs are known to be non existent when d is odd and for most of the other even d there are mostly 3 Real MUBs. In case of C^d the construction for complete set of MUBs are known for only Prime Power dimension. Thus in general large set of MUBs are not known, particularly for composite dimensions which are not of the form of prime powers. Because of this, there are many constructions of Approximate version of MUBs. In this paper we make an attempt to define certain measures to characterise the AMUBs. Our construction of measures derives its inspiration from the applications of MUBs, and based on them, we define certain quantifiable measures, which are can be computed and gives estimates of how close the Approximate MUBs are to the MUBs. We use geometric interpretation, projective design features of MUBs and applications like Optimal State determination and Entropic Uncertainty of MUBs. We show generic relationship between these measures and show that it can be evaluated for APMUBs without known the exact construction details, thereby showing that definition of APMUB is sufficient completely characterise it. We also evaluate these measure for an interesting class of AMUBs called Weak MUBs and certain AMUBs constructed using RBDs.

quant-ph↗

arXiv2025

Strong lead-free bioinspired piezoceramics for durable energy transducers

Durable, high-performance and eco-friendly lead-free piezoceramics are essential for next-generation sustainable energy transducers and electromechanical systems. While significant performance enhancements have been made, through chemical composition, texture, or crystal defects, piezoceramics are intrinsically weak mechanically, which negatively impact their working conditions and durability. What's more, improving comprehensive mechanical durability without sacrificing piezoelectric performance remains a key challenge. Here, we design bioinspired Bi0.5Na0.5TiO3 (BNT) ceramics using a scalable colloidal process that enables multiscale control over the microstructure. The design comprises plate-like monocrystalline BNT bricks stacked to induce a crystallographic texture along the poling direction, bonded together by a silica-based mortar, forming the brick-and-mortar phase. This deliberate microstructure design yields 2- to 3-fold increase in flexural strength, and 1.6- to 2-fold increase in fracture toughness compared with a BNT synthesized conventionally, comparable to common structural ceramics, without sacrificing the piezoelectric performance. In addition, the bioinspired BNT exhibit dramatically enhanced ferroelectric fatigue resistance, with a 10- to 15-folds improvement in the number of field-induced electromechanical cycles before failure. These gains originate from anisotropic residual stress fields, revealed by Raman spectroscopy and XRD, which delay crack initiation events. Furthermore, we demonstrated enhanced transducing capability and electromechanical fatigue resistance using a cantilever beam-based piezoelectric transducer under bending mode. Given its non-chemical-compositional origin, this bioinspired strategy could be broadly applicable to other piezoelectric material systems for applications where both functional and structural performance are critical.

cond-mat.mtrl-sci↗

arXiv2025

On Construction of Approximate Real Mutually Unbiased Bases for an infinite class of dimensions $d \not\equiv 0 \bmod 4$

It is known that real Mutually Unbiased Bases (MUBs) do not exist for any dimension $d > 2$ which is not divisible by 4. Thus, the next combinatorial question is how one can construct Approximate Real MUBs (ARMUBs) in this direction with encouraging parameters. In this paper, for the first time, we show that it is possible to construct $> \lceil \sqrt{d} \rceil$ many ARMUBs for certain odd dimensions $d$ of the form $d = (4n-t)s$, $t = 1, 2, 3$, where $n$ is a natural number and $s$ is an odd prime power. Our method exploits any available $4n \times 4n$ real Hadamard matrix $H_{4n}$ (conjectured to be true) and uses this to construct an orthogonal matrix ${Y}_{4n-t}$ of size $(4n - t) \times (4n - t)$, such that the absolute value of each entry varies a little from $\frac{1}{\sqrt{4n-t}}$. In our construction, the absolute value of the inner product between any pair of basis vectors from two different ARMUBs will be $\leq \frac{1}{\sqrt{d}}(1 + O(d^{-\frac{1}{4}})) < 2$, for proper choices of parameters, the class of dimensions $d$ being infinitely large.

cs.DM↗

arXiv2025

On Obtaining New MUBs by Finding Points on Complete Intersection Varieties over $\mathbb{R}$

Mutually Unbiased Bases (MUBs) are closely connected with quantum physics, and the structure has a rich mathematical background. We provide equivalent criteria for extending a set of MUBs for $C^n$ by studying real points of a certain affine algebraic variety. This variety comes from the relations that determine the extendability of a system of MUBs. Finally, we show that some part of this variety gives rise to complete intersection domains. Further, we show that there is a one-to-one correspondence between MUBs and the maximal commuting classes (bases) of orthogonal normal matrices in $\mathcal M_n({\mathbb{C}})$. It means that for $m$ MUBs in $C^n$, there are $m$ commuting classes, each consisting of $n$ commuting orthogonal normal matrices and the existence of maximal commuting basis for $\mathcal M_n({\mathbb{C}})$ ensures the complete set of MUBs in $\mathcal M_n({\mathbb{C}})$.

cs.DM↗

arXiv2024

Almost Perfect Mutually Unbiased Bases that are Sparse

In dimension $d$, Mutually Unbiased Bases (MUBs) are a collection of orthonormal bases over $\mathbb{C}^d$ such that for any two vectors $v_1, v_2$ belonging to different bases, the scalar product $|\braket{v_1|v_2}| = \frac{1}{\sqrt{d}}$. The upper bound on the number of such bases is $d+1$. Constructions to achieve this bound are known when $d$ is some power of prime. The situation is more restrictive in other cases and also when we consider the results over real rather than complex. Thus, certain relaxations of this model are considered in literature and consequently Approximate MUBs (AMUB) are studied. This enables one to construct potentially large number of such objects for $\mathbb{C}^d$ as well as in $\mathbb{R}^d$. In this regard, we propose the concept of Almost Perfect MUBs (APMUB), where we restrict the absolute value of inner product $|\braket{v_1|v_2}|$ to be two-valued, one being 0 and the other $ \leq \frac{1+\mathcal{O}(d^{-\lambda})}{\sqrt{d}}$, such that $\lambda > 0$ and the numerator $1 + \mathcal{O}(d^{-\lambda}) \leq 2$. Each such vector constructed, has an important feature that large number of its components are zero and the non-zero components are of equal magnitude. Our techniques are based on combinatorial structures related to RBDs. We show that for several composite dimensions $d$, one can construct $\mathcal{O}(\sqrt{d})$ many APMUBs, in which cases the number of MUBs are significantly small. To be specific, this result works for $d$ of the form $(q-e)(q+f), \ q, e, f \in \mathbb{N}$, with the conditions $0 \leq f \leq e$ for constant $e, f$ and $q$ some power of prime. We also show that such APMUBs provide sets of Bi-angular vectors which are $\mathcal{O}(d^{\frac{3}{2}})$ in numbers, having high angular distances among them. Finally, as the MUBs are equivalent to a set of Hadamard matrices, we show that the APMUBs are so with the set of Weighing matrices.

cs.DM↗

arXiv2024

Further Constructions of AMUBs for Non-prime power Composite Dimensions

Construction of a large class of Mutually Unbiased Bases (MUBs) for non-prime power composite dimensions ($d = k\times s$) is a long standing open problem, which leads to different construction methods for the class Approximate MUBs (AMUBs) by relaxing the criterion that the absolute value of the dot product between two vectors chosen from different bases should be $\leq \frac{\beta}{\sqrt{d}}$. In this chapter, we consider a more general class of AMUBs (ARMUBs, considering the real ones too), compared to our earlier work in [Cryptography and Communications, 14(3): 527--549, 2022]. We note that the quality of AMUBs (ARMUBs) constructed using RBD$(X,A)$ with $|X|= d$, critically depends on the parameters, $|s-k|$, $\mu$ (maximum number of elements common between any pair of blocks), and the set of block sizes. We present the construction of $\mathcal{O}(\sqrt{d})$ many $\beta$-AMUBs for composite $d$ when $|s-k|< \sqrt{d}$, using RBDs having block sizes approximately $\sqrt{d}$, such that $|\braket{\psi^l_i|\psi^m_j}| \leq \frac{\beta}{\sqrt{d}}$ where $\beta = 1 + \frac{|s-k|}{2\sqrt{d}}+ \mathcal{O}(d^{-1}) \leq 2$. Moreover, if real Hadamard matrix of order $k$ or $s$ exists, then one can construct at least $N(k)+1$ (or $N(s)+1$) many $\beta$-ARMUBs for dimension $d$, with $\beta \leq 2 - \frac{|s-k|}{2\sqrt{d}}+ \mathcal{O}(d^{-1})< 2$, where $N(w)$ is the number of MOLS$(w)$. This improves and generalizes some of our previous results for ARMUBs from two points, viz., the real cases are now extended to complex ones too. The earlier efforts use some existing RBDs, whereas here we consider new instances of RBDs that provide better results. Similar to the earlier cases, the AMUBs (ARMUBs) constructed using RBDs are in general very sparse, where the sparsity $(\epsilon)$ is $1 - \mathcal{O}(d^{-\frac{1}{2}})$.

cs.DM↗

arXiv2021

Analysis of Model-Free Reinforcement Learning Control Schemes on self-balancing Wheeled Extendible System

Traditional linear control strategies have been extensively researched and utilized in many robotic and industrial applications and yet they do not respond to the total dynamics of the systems. To avoid tedious calculations for nonlinear control schemes like H-infinity control and predictive control, the application of Reinforcement Learning(RL) can provide alternative solutions. This article presents the implementation of RL control with Deep Deterministic Policy Gradient and Proximal Policy Optimization on a mobile self-balancing Extendable Wheeled Inverted Pendulum (E-WIP) system with provided state history to attain improved control. Such RL models make the task of finding satisfactory control schemes easier and responding to the dynamics effectively while self-tuning the parameters to provide better control. In this article, RL-based controllers are pitted against an MPC controller to evaluate the performance on the basis of state variables and trajectory errors of the E-WIP system while following a specific desired trajectory.

cs.RO↗

arXiv2019

NbOx based memristor as artificial synapse emulating short term plasticity

Memristors can mimic the functions of biological synapse, where it can simultaneously store the synaptic weight and modulate the transmitted signal. Here, we report Nb/Nb2O5/Pt based memristors with bipolar resistive switching, exhibiting synapse like property of gradual and continuously change of conductance with subsequent voltage signals. Mimicking of basic functions of remembering and forgetting processes of biological brain were demonstrated through short term plasticity, spike rate dependent plasticity, paired pulse facilitation and post-titanic potentiation. The device layer interface tuning was shown to affect the device properties shift from digital to analog behaviour. Demonstration of basic synaptic functions in the NbOx based devices makes them suitable for neuromorphic applications.

physics.app-ph↗

arXiv2019

Controlled inter-state switching between quantized conductance states in resistive devices for multilevel memory

A detailed understanding of quantization conductance (QC), their correlation with resistive switching phenomena and controlled manipulation of quantized states is crucial for realizing atomic-scale multilevel memory elements. Here, we demonstrate highly stable and reproducible quantized conductance states (QC-states) in Al/Niobium oxide/Pt resistive switching devices. Three levels of control over the QC-states, required for multilevel quantized state memories, like, switching ON to different quantized states, switching OFF from quantized states, and controlled inter-state switching among one QC states to another has been demonstrated by imposing limiting conditions of stop-voltage and current compliance. The well defined multiple QC-states along with a working principle for switching among various states show promise for implementation of multilevel memory devices.

physics.app-ph↗

arXiv2018

Investigating Unipolar Switching in Niobium Oxide Resistive Switches: Correlating Quantized Conductance and Mechanism

Memory devices based on resistive switching (RS) have not been fully realised due to lack of understanding of the underlying switching mechanisms. Nature of ion transport responsible for switching and growth of conducting filament in transition metal oxide based RS devices is still in debate. Here, we investigated the mechanism in Niobium oxide based RS devices, which shows unipolar switching with high ON/OFF ratio, good endurance cycles and high retention times. We controlled the boundary conditions between low-conductance insulating and a high-conductance metallic state where conducting filament (CF) can form atomic point contact and exhibit quantized conductance behaviour. Based on the statistics generated from quantized steps data, we demonstrated that the CF is growing atom by atom with the applied voltage sweeps. We also observed stable quantized states, which can be utilized in multistate switching.

cond-mat.mtrl-sci↗

arXiv2018

Quantum oscillations of the Stoner susceptibility: theory

Oscillatory effects in magnetic susceptibility of free electrons in a strong magnetic field is well known phenomenon and is well captured by Lifshitz-Kosevich formula. In this paper we point out similar oscillatory effects in Stoner susceptibility which makes the system to oscillate between paramagnetic phase and ferromagnetic phase alternatively as a function of external magnetic field strength. This effect can happen in a material which is tuned near to its magnetic instability. We suggest an experimental set-up to observe this effect. We also suggest that our result can be exploited to control a quantum critical system around its quantum critical point to study its thermodynamical or transport properties.

cond-mat.str-el↗