arXiv subjects

Ayan Biswas

Publications and source records attributed to Ayan Biswas.

53 records · Page 3Linked to original sources

arXiv2022

IDLat: An Importance-Driven Latent Generation Method for Scientific Data

Deep learning based latent representations have been widely used for numerous scientific visualization applications such as isosurface similarity analysis, volume rendering, flow field synthesis, and data reduction, just to name a few. However, existing latent representations are mostly generated from raw data in an unsupervised manner, which makes it difficult to incorporate domain interest to control the size of the latent representations and the quality of the reconstructed data. In this paper, we present a novel importance-driven latent representation to facilitate domain-interest-guided scientific data visualization and analysis. We utilize spatial importance maps to represent various scientific interests and take them as the input to a feature transformation network to guide latent generation. We further reduced the latent size by a lossless entropy encoding algorithm trained together with the autoencoder, improving the storage and memory efficiency. We qualitatively and quantitatively evaluate the effectiveness and efficiency of latent representations generated by our method with data from multiple scientific visualization applications.

cs.LG↗

arXiv2022

Anisotropic Spatial Entanglement

The photon-pairs generated through spontaneous parametric down-conversion (SPDC) possess strong correlations in their transverse position and momentum degrees of freedom. For such photon pairs, the transverse position correlation length depends on the crystal thickness and the pump wavelength, whereas the transverse momentum correlation length depends on the beam waist and spatial coherence length of the pump. By controlling these parameters, it is possible to engineer the spatial entanglement. Here, we utilize the circular asymmetry of the pump by using an elliptical-Gaussian beam to change the degree of entanglement in transverse directions, which we call anisotropic entanglement. We show the interrelation between the degree of anisotropic entanglement and the asymmetry in the beam width. In addition, we also show that for a highly asymmetric pump beam, the entanglement along the y-direction completely vanishes, whereas the entanglement in the x-direction remains intact.

quant-ph↗

arXiv2022

Robust absolute solar flux density calibration for the Murchison Widefield Array

Sensitive radio instruments are optimized for observing faint astronomical sources, and usually need to attenuate the received signal when observing the Sun. There are only a handful of flux density calibrators which can comfortably be observed with the same attenuation setup as the Sun. Additionally, for wide field-of-view (FoV) instruments like the Murchison Widefield Array (MWA) calibrator observations are generally done when the Sun is below the horizon to avoid the contamination from solar emissions. These considerations imply that the usual radio interferometric approach to flux density calibration is not applicable for solar imaging. A novel technique, relying on a good sky model and detailed characterization of the MWA hardware, was developed for solar flux density calibration for MWA. Though successful, this technique is not general enough to be extended to the data from the extended configuration of the MWA Phase II. Here, we present a robust flux density calibration method for solar observations with MWA independent of the array configuration. We use different approaches -- the serendipitous presence of strong sources; detection of numerous background sources using high dynamic range images in the FoV along with the Sun and observations of strong flux density calibrators with and without the additional attenuation used for solar observations; to obtain the flux scaling parameters required for the flux density calibration. Using the present method, we have achieved an absolute flux density uncertainty $\sim10\%$ for solar observations even in the absence of dedicated calibrator observations.

astro-ph.SR↗

arXiv2022

BBM92 quantum key distribution over a free space dusty channel of 200 meters

Free space quantum communication assumes importance as it is a precursor for satellite-based quantum communication needed for secure key distribution over longer distances. Prepare and measure protocols like BB84 consider the satellite as a trusted device, which is fraught with security threat looking at the current trend for satellite-based optical communication. Therefore, entanglement-based protocols must be preferred, so that one can consider the satellite as an untrusted device too. The current work reports the implementation of BBM92 protocol, an entanglement-based QKD protocol over 200 m distance using an indigenous facility developed at Physical Research Laboratory (PRL), Ahmedabad, India. Our results show the effect of atmospheric aerosols on sift key rate, and eventually, secure key rate. Such experiments are important to validate the models to account for the atmospheric effects on the key rates achieved through satellite-based QKD.

quant-ph↗

arXiv2021

Quantum key distribution with multiphoton pulses: An advantage

In this article, we introduce a quantum key distribution protocol for the line of sight channels based on coincidence measurements. We present a proof-of-concept implementation of our protocol. We show that using coincidence measurements to monitor multi-photon pulses results in a higher secure key rate over longer distances for such channels. This key rate is higher than popular implementations of quantum key distribution protocol based on BB84, for example, the GLLP analysis [Quant. Info. Comput. \textbf{4}, 325 (2004)]. In the experiment, we could generate around $74 \%$ more key bits per signal pulse as compared to the GLLP analysis of BB84 protocol with similar parameters and equal value of mean photon number.

quant-ph↗

arXiv2021

Information transmission in a two-step cascade: Interplay of activation and repression

We present an information-theoretic formalism to study signal transduction in four architectural variants of a model two-step cascade with increasing input population. Our results categorize these four types into two classes depending upon the effect played out by activation and repression on mutual information, net synergy, and signal-to-noise ratio. Within the Gaussian framework and using the linear noise approximation, we derive the analytic expressions for these metrics to establish their underlying relationships in terms of the biochemical parameters. We also verify our approximations through stochastic simulations.

q-bio.MN↗

arXiv2021

Experimental Side Channel Analysis of BB84 QKD Source

A typical implementation of BB84 protocol for quantum communication uses four laser diodes for transmitting weak coherent pulses, which may not have the same characteristics. We have characterized these lasers for mismatch in various parameters such as spectral width, pulse width, spatial mode, peak wavelength, polarization and their arrival times at the receiver. This information is utilized to calculate possible information leakage through side channel attacks by evaluating mutual information between source and eavesdropper. Based on our experimental observations of cross correlation between parameter values for different laser diodes, we suggest ways to reduce information leakage to Eve.

quant-ph↗

arXiv2020

Relationship-aware Multivariate Sampling Strategy for Scientific Simulation Data

With the increasing computational power of current supercomputers, the size of data produced by scientific simulations is rapidly growing. To reduce the storage footprint and facilitate scalable post-hoc analyses of such scientific data sets, various data reduction/summarization methods have been proposed over the years. Different flavors of sampling algorithms exist to sample the high-resolution scientific data, while preserving important data properties required for subsequent analyses. However, most of these sampling algorithms are designed for univariate data and cater to post-hoc analyses of single variables. In this work, we propose a multivariate sampling strategy which preserves the original variable relationships and enables different multivariate analyses directly on the sampled data. Our proposed strategy utilizes principal component analysis to capture the variance of multivariate data and can be built on top of any existing state-of-the-art sampling algorithms for single variables. In addition, we also propose variants of different data partitioning schemes (regular and irregular) to efficiently model the local multivariate relationships. Using two real-world multivariate data sets, we demonstrate the efficacy of our proposed multivariate sampling strategy with respect to its data reduction capabilities as well as the ease of performing efficient post-hoc multivariate analyses.

cs.LG↗

arXiv2020

The coherent feed-forward loop acts as an efficient information transmitting motif

We present a theoretical formalism to study steady state information transmission in type 1 coherent feed-forward loop motif with an additive signal integration mechanism. Our construct allows a two-step cascade to be slowly transformed into a bifurcation network via a feed-forward loop which is a prominent network motif. Utilizing a Gaussian framework, we show that the feed-forward loop motif harnesses the maximum amount of Shannon mutual information fractions constructed between the input signaling species and the final gene product and also between the two downstream gene products. Our findings may explain why a feed-forward loop is naturally selected by the force of evolution as a motif in bacterial transcription control network while the two-step cascade and bifurcation network are not.

q-bio.MN↗

arXiv2019

Deep Learning-Based Feature-Aware Data Modeling for Complex Physics Simulations

Data modeling and reduction for in situ is important. Feature-driven methods for in situ data analysis and reduction are a priority for future exascale machines as there are currently very few such methods. We investigate a deep-learning based workflow that targets in situ data processing using autoencoders. We propose a Residual Autoencoder integrated Residual in Residual Dense Block (RRDB) to obtain better performance. Our proposed framework compressed our test data into 66 KB from 2.1 MB per 3D volume timestep.

cs.LG↗

arXiv2019

Multivariate Pointwise Information-Driven Data Sampling and Visualization

With increasing computing capabilities of modern supercomputers, the size of the data generated from the scientific simulations is growing rapidly. As a result, application scientists need effective data summarization techniques that can reduce large-scale multivariate spatiotemporal data sets while preserving the important data properties so that the reduced data can answer domain-specific queries involving multiple variables with sufficient accuracy. While analyzing complex scientific events, domain experts often analyze and visualize two or more variables together to obtain a better understanding of the characteristics of the data features. Therefore, data summarization techniques are required to analyze multi-variable relationships in detail and then perform data reduction such that the important features involving multiple variables are preserved in the reduced data. To achieve this, in this work, we propose a data sub-sampling algorithm for performing statistical data summarization that leverages pointwise information theoretic measures to quantify the statistical association of data points considering multiple variables and generates a sub-sampled data that preserves the statistical association among multi-variables. Using such reduced sampled data, we show that multivariate feature query and analysis can be done effectively. The efficacy of the proposed multivariate association driven sampling algorithm is presented by applying it on several scientific data sets.

cs.HC↗

arXiv2019

Single photon sources with different spatial modes

We study the correlation properties of single photons carrying orbital angular momentum (OAM) in a Hanbury Brown and Twiss (HBT) type experiment. We have characterized single photon sources obtained by pumping a nonlinear crystal with a laser beam carrying different OAM under same experimental conditions. For heralded twisted single photons carrying OAM, we calculate $g^{(2)}(0)$, a measurable parameter characterizing the quality of a single photon source, and observe an increment with the OAM of the single photon.

quant-ph↗

arXiv2018

Information processing in a simple one-step cascade

Using the formalism of information theory, we analyze the mechanism of information transduction in a simple one-step signaling cascade S$\rightarrow$X representing the gene regulatory network. Approximating the signaling channel to be Gaussian, we describe the dynamics using Langevin equations. Upon discretization, we calculate the associated second moments for linear and nonlinear regulation of the output by the input, which follows the birth-death process. While mutual information between the input and the output characterizes the channel capacity, the Fano factor of the output gives a clear idea of how internal and external fluctuations assemble at the output level. To quantify the contribution of the present state of the input to predict the future output, transfer entropy is computed. We find that higher amount of transfer entropy is accompanied by the greater magnitude of external fluctuations (quantified by the Fano factor of the output) propagation from the input to the output. We notice that low input population characterized by the number of signaling molecules S, which fluctuates in a relatively slower fashion compared to its downstream (target) species X, is maximally able to predict (as quantified by transfer entropy) the future state of the output. Our computations also reveal that with increased linear nature of the input-output interaction, all three metrics of mutual information, Fano factor and, transfer entropy achieve relatively larger magnitudes.

q-bio.MN↗

arXiv2018

Energy-efficient switching of nanomagnets for computing: Straintronics and other methodologies

The need for increasingly powerful computing hardware has spawned many ideas stipulating, primarily, the replacement of traditional transistors with alternate "switches" that dissipate miniscule amounts of energy when they switch and provide additional functionality that are beneficial for information processing. An interesting idea that has emerged recently is the notion of using two-phase (piezoelectric/magnetostrictive) multiferroic nanomagnets with bistable (or multi-stable) magnetization states to encode digital information (bits), and switching the magnetization between these states with small voltages (that strain the nanomagnets) to carry out digital information processing. The switching delay is ~1 ns and the energy dissipated in the switching operation can be few to tens of aJ, which is comparable to, or smaller than, the energy dissipated in switching a modern-day transistor. Unlike a transistor, a nanomagnet is "non-volatile", so a nanomagnetic processing unit can store the result of a computation locally without refresh cycles, thereby allowing it to double as both logic and memory. These dual-role elements promise new, robust, energy-efficient, high-speed computing and signal processing architectures (usually non-Boolean and often non-von-Neumann) that can be more powerful, architecturally superior (fewer circuit elements needed to implement a given function) and sometimes faster than their traditional transistor-based counterparts. This topical review covers the important advances in computing and information processing with nanomagnets with emphasis on strain-switched multiferroic nanomagnets acting as non-volatile and energy-efficient switches - a field known as "straintronics". It also outlines key challenges in straintronics.

cond-mat.mes-hall↗

arXiv2018

Interplay of synergy and redundancy in diamond motif

The formalism of partial information decomposition provides independent or non-overlapping components constituting total information content provided by a set of source variables about the target variable. These components are recognised as unique information, synergistic information and, redundant information. The metric of net synergy, conceived as the difference between synergistic and redundant information, is capable of detecting synergy, redundancy and, information independence among stochastic variables. And it can be quantified, as it is done here, using appropriate combinations of different Shannon mutual information terms. Utilisation of such a metric in network motifs with the nodes representing different biochemical species, involved in information sharing, uncovers rich store for interesting results. In the current study, we make use of this formalism to obtain a comprehensive understanding of the relative information processing mechanism in a diamond motif and two of its sub-motifs namely bifurcation and integration motif embedded within the diamond motif. The emerging patterns of synergy and redundancy and their effective contribution towards ensuring high fidelity information transmission are duly compared in the sub-motifs and independent motifs (bifurcation and integration). In this context, the crucial roles played by various time scales and activation coefficients in the network topologies are especially emphasised. We show that the origin of synergy and redundancy in information transmission can be physically justified by decomposing diamond motif into bifurcation and integration motif.

q-bio.MN↗

arXiv2018

Exploiting Inherent Error-Resiliency of Neuromorphic Computing to achieve Extreme Energy-Efficiency through Mixed-Signal Neurons

Neuromorphic computing, inspired by the brain, promises extreme efficiency for certain classes of learning tasks, such as classification and pattern recognition. The performance and power consumption of neuromorphic computing depends heavily on the choice of the neuron architecture. Digital neurons (Dig-N) are conventionally known to be accurate and efficient at high speed, while suffering from high leakage currents from a large number of transistors in a large design. On the other hand, analog/mixed-signal neurons are prone to noise, variability and mismatch, but can lead to extremely low-power designs. In this work, we will analyze, compare and contrast existing neuron architectures with a proposed mixed-signal neuron (MS-N) in terms of performance, power and noise, thereby demonstrating the applicability of the proposed mixed-signal neuron for achieving extreme energy-efficiency in neuromorphic computing. The proposed MS-N is implemented in 65 nm CMOS technology and exhibits > 100X better energy-efficiency across all frequencies over two traditional digital neurons synthesized in the same technology node. We also demonstrate that the inherent error-resiliency of a fully connected or even convolutional neural network (CNN) can handle the noise as well as the manufacturing non-idealities of the MS-N up to certain degrees. Notably, a system-level implementation on MNIST datasets exhibits a worst-case increase in classification error by 2.1% when the integrated noise power in the bandwidth is ~ 0.1 uV2, along with +-3σ amount of variation and mismatch introduced in the transistor parameters for the proposed neuron with 8-bit precision.

cs.ET↗

arXiv2016

Redundancy in the information transmission in a two-step cascade

We present a stochastic framework to study signal transmission in a generic two-step cascade $\text{S} \rightarrow \text{X} \rightarrow \text{Y}$. Starting from a set of Langevin equations obeying Gaussian noise processes we calculate the variance and covariance while considering both linear and nonlinear production terms for different biochemical species of the cascade. These quantities are then used to calculate the net synergy within the purview of partial information decomposition. We show that redundancy in information transmission is essentially an important consequence of Markovian property of the two-step cascade motif. We also show that redundancy increases fidelity of the signalling pathway.

q-bio.MN↗