arXiv subjects

Vaibhav Kumar

Publications and source records attributed to Vaibhav Kumar.

At least 55 records · Page 3Linked to original sources

arXiv2022

On Energy Efficiency and Fairness Maximization in RIS-Assisted MU-MISO mmWave Communications

Reconfigurable intelligent surfaces (RISs) are considered to be a promising solution to overcome the blockage issue in the millimeter-wave (mmWave) band. Energy efficiency is an important performance metric in mmWave systems with a large number of antennas. However, due to the severe path loss in mmWave systems, resource allocation algorithms tend to allocate most of the resources for the benefit of the users with higher channel gains. In this paper, we propose a lexicographic-based approach to find the optimal power allocation, RIS phase shift matrix, and analog precoders that maximize both energy efficiency and user fairness. We solve the corresponding multi-objective optimization problem in two stages. In the first stage, we maximize the energy efficiency, and in the second stage we maximize the fairness subject to a minimum energy efficiency constraint. We propose an alternating optimization procedure to solve the optimization problem in each stage. The optimal power allocation is found using Dinkelbach's method in the first stage and using convex optimization techniques in the second stage, the RIS phase shift matrix is found using a gradient ascent algorithm, and the analog precoder is determined using beam alignment. Numerical results show that the proposed algorithm can achieve an excellent trade-off between the energy efficiency and fairness by boosting the minimum weighted rate with a minor and controllable reduction in the energy efficiency.

eess.SP↗

arXiv2022

On the Energy-Efficiency Maximization for IRS-Assisted MIMOME Wiretap Channels

Security and energy efficiency have become crucial features in the modern-era wireless communication. In this paper, we consider an energy-efficient design for intelligent reflecting surface (IRS)-assisted multiple-input multiple-output multiple-eavesdropper (MIMOME) wiretap channels (WTC). Our objective is to jointly optimize the transmit covariance matrix and the IRS phase-shifts to maximize the secrecy energy efficiency (SEE) of the considered system subject to a secrecy rate constraint at the legitimate receiver. To tackle this challenging non-convex problem in which the design variables are coupled in the objective and the constraint, we propose a penalty dual decomposition based alternating gradient projection (PDDAPG) method to obtain an efficient solution. We also show that the computational complexity of the proposed algorithm grows only linearly with the number of reflecting elements at the IRS, as well as with the number of antennas at transmitter/receivers' nodes. Our results confirm that using an IRS is helpful to improve the SEE of MIMOME WTC compared to its no-IRS counterpart only when the power consumption at IRS is small. In particular, and a large-sized IRS is not always beneficial for the SEE of a MIMOME WTC.

eess.SP↗

arXiv2022

On the Achievable Rate of IRS-Assisted Multigroup Multicast Systems

Intelligent reflecting surfaces (IRSs) have shown huge advantages in many potential use cases and thus have been considered a promising candidate for next-generation wireless systems. In this paper, we consider an IRS-assisted multigroup multicast (IRS-MGMC) system in a multiple-input single-output (MISO) scenario, for which the related existing literature is rather limited. In particular, we aim to jointly design the transmit beamformers and IRS phase shifts to maximize the sum rate of the system under consideration. In order to obtain a numerically efficient solution to the formulated non-convex optimization problem, we propose an alternating projected gradient (APG) method where each iteration admits a closed-form and is shown to be superior to a known solution that is derived from the majorization-minimization (MM) method in terms of both achievable sum rate and required complexity, i.e., run time. In particular, we show that the complexity of the proposed APG method grows linearly with the number of IRS tiles, while that of the known solution in comparison grows with the third power of the number of IRS tiles. The numerical results reported in this paper extend our understanding on the achievable rates of large-scale IRS-assisted multigroup multicast systems.

cs.IT↗

arXiv2022

JARVix at SemEval-2022 Task 2: It Takes One to Know One? Idiomaticity Detection using Zero and One-Shot Learning

Large Language Models have been successful in a wide variety of Natural Language Processing tasks by capturing the compositionality of the text representations. In spite of their great success, these vector representations fail to capture meaning of idiomatic multi-word expressions (MWEs). In this paper, we focus on the detection of idiomatic expressions by using binary classification. We use a dataset consisting of the literal and idiomatic usage of MWEs in English and Portuguese. Thereafter, we perform the classification in two different settings: zero shot and one shot, to determine if a given sentence contains an idiom or not. N shot classification for this task is defined by N number of common idioms between the training and testing sets. In this paper, we train multiple Large Language Models in both the settings and achieve an F1 score (macro) of 0.73 for the zero shot setting and an F1 score (macro) of 0.85 for the one shot setting. An implementation of our work can be found at https://github.com/ashwinpathak20/Idiomaticity_Detection_Using_Few_Shot_Learning.

cs.CL↗

arXiv2022

Achievable Rate Maximization for Underlay Spectrum Sharing MIMO System with Intelligent Reflecting Surface

In this letter, the achievable rate maximization problem is considered for intelligent reflecting surface (IRS) assisted multiple-input multiple-output (MIMO) systems in an underlay spectrum sharing scenario, subject to interference power constraints at the primary users. The formulated non-convex optimization problem is challenging to solve due to its non-convexity as well as coupling design variables in the constraints. Different from existing works that are mostly based on alternating optimization (AO), we propose a penalty dual decomposition based gradient projection (PDDGP) algorithm to solve this problem. We also provide a convergence proof and a complexity analysis for the proposed algorithm. We benchmark the proposed algorithm against two known solutions, namely a minimum mean-square error based AO algorithm and an inner approximation method with block coordinate descent. Specifically, the complexity of the proposed algorithm grows linearly with respect to the number of reflecting elements at the IRS, while that of the two benchmark methods grows with the third power of the number of IRS elements. Moreover, numerical results show that the proposed PDDGP algorithm yields considerably higher achievable rate than the benchmark solutions.

cs.IT↗

arXiv2021

Identifying and Mitigating Gender Bias in Hyperbolic Word Embeddings

Euclidean word embedding models such as GloVe and Word2Vec have been shown to reflect human-like gender biases. In this paper, we extend the study of gender bias to the recently popularized hyperbolic word embeddings. We propose gyrocosine bias, a novel measure for quantifying gender bias in hyperbolic word representations and observe a significant presence of gender bias. To address this problem, we propose Poincaré Gender Debias (PGD), a novel debiasing procedure for hyperbolic word representations. Experiments on a suit of evaluation tests show that PGD effectively reduces bias while adding a minimal semantic offset.

cs.CL↗

arXiv2021

On the Effective Rate of NOMA in Underlay Spectrum Sharing

In this paper, we present the delay-constrained performance analysis of a multi-antenna-assisted multiuser non-orthogonal multiple access (NOMA) based spectrum sharing system over Rayleigh fading channels. We derive analytical expressions for the sum effective rate (ER) for the downlink NOMA system under a peak interference constraint. In particular, we show the effect of the availability of different levels of channel state information (instantaneous and statistical) on the system performance. We also show the effect of different parameters of interest, including the peak tolerable interference power, the delay exponent, the number of antennas and the number of users, on the sum ER of the system under consideration. An excellent agreement between simulation and theoretical results confirms the accuracy of the analysis.

cs.IT↗

arXiv2021

Secrecy Rate Maximization for Intelligent Reflecting Surface Assisted MIMOME Wiretap Channels

Intelligent reflecting surface (IRS) has gained tremendous attention recently as a disruptive technology for beyond 5G networks. In this paper, we consider the problem of secrecy rate maximization for an IRS-assisted Gaussian multiple-input multiple-output multi-antenna-eavesdropper (MIMOME) wiretap channel (WTC). In this context, we aim to jointly optimize the input covariance matrix and the IRS phase shifts to maximize the achievable secrecy rate of the considered system. To solve the formulated problem which is non-convex, we propose an iterative method based on the block successive maximization (BSM), where each iteration is done in closed form. More specifically, we maximize a lower bound on the achievable secrecy rate to update the input covariance matrix for fixed phase shifts, and then maximize the (exact) achievable secrecy rate to update phase shifts for a given input covariance.We consider the total free space path loss (FSPL) in this system to emphasize the first-order measure of the applicability of the IRS in the considered communication system. We present a convergence proof and the associated complexity analysis of the proposed algorithm. Numerical results are provided to demonstrate the superiority of the proposed method compared to a known solution, and also to show the effect of different parameters of interest on the achievable secrecy rate of the IRS-assisted MIMOME WTC.

cs.IT↗

arXiv2021

On the Secrecy Rate under Statistical QoS Provisioning for RIS-Assisted MISO Wiretap Channel

Reconfigurable intelligent surface (RIS) assisted radio is considered as an enabling technology with great potential for the sixth-generation (6G) wireless communications standard. The achievable secrecy rate (ASR) is one of the most fundamental metrics to evaluate the capability of facilitating secure communication for RIS-assisted systems. However, the definition of ASR is based on Shannon's information theory, which generally requires long codewords and thus fails to quantify the secrecy of emerging delay-critical services. Motivated by this, in this paper we investigate the problem of maximizing the secrecy rate under a delay-limited quality-of-service (QoS) constraint, termed as the effective secrecy rate (ESR), for an RIS-assisted multiple-input single-output (MISO) wiretap channel subject to a transmit power constraint. We propose an iterative method to find a stationary solution to the formulated non-convex optimization problem using a block coordinate ascent method (BCAM), where both the beamforming vector at the transmitter as well as the phase shifts at the RIS are obtained in closed forms in each iteration. We also present a convergence proof, an efficient implementation, and the associated complexity analysis for the proposed method. Our numerical results demonstrate that the proposed optimization algorithm converges significantly faster that an existing solution. The simulation results also confirm that the secrecy rate performance of the system with stringent delay requirements reduce significantly compared to the system without any delay constraints, and that this reduction can be significantly mitigated by an appropriately placed large-size RIS.

cs.IT↗

arXiv2021

On the Optimality of the Stationary Solution of Secrecy Rate Maximization for MIMO Wiretap Channel

To achieve perfect secrecy in a multiple-input multiple-output (MIMO) Gaussian wiretap channel (WTC), we need to find its secrecy capacity and optimal signaling, which involves solving a difference of convex functions program known to be non-convex for the non-degraded case. To deal with this, a class of existing solutions have been developed but only local optimality is guaranteed by standard convergence analysis. Interestingly, our extensive numerical experiments have shown that these local optimization methods indeed achieve global optimality. In this paper, we provide an analytical proof for this observation. To achieve this, we show that the Karush-Kuhn-Tucker (KKT) conditions of the secrecy rate maximization problem admit a unique solution for both degraded and non-degraded cases. Motivated by this, we also propose a low-complexity algorithm to find a stationary point. Numerical results are presented to verify the theoretical analysis.

cs.IT↗

arXiv2021

On the Secrecy Rate of Downlink NOMA in Underlay Spectrum Sharing with Imperfect CSI

In this paper, we present the ergodic sum secrecy rate (ESSR) analysis of an underlay spectrum sharing non-orthogonal multiple access (NOMA) system. We consider the scenario where the power transmitted by the secondary transmitter (ST) is constrained by the peak tolerable interference at multiple primary receivers (PRs) as well as the maximum transmit power of the ST. The effect of channel estimation error is also taken into account in our analysis. We derive exact and asymptotic closed-form expressions for the ESSR of the downlink NOMA system, and show that the performance can be classified into two distinct regimes, i.e., it is dictated either by the interference constraint or by the power constraint. Our results confirm the superiority of the NOMA-based system over its orthogonal multiple access (OMA) based counterpart. More interestingly, our results show that NOMA helps in maintaining the secrecy rate of the strong user while significantly enhancing the secrecy performance of the weak user as compared to OMA. The correctness of the proposed investigation is corroborated through Monte Carlo simulation.

cs.IT↗

arXiv2020

Fair Embedding Engine: A Library for Analyzing and Mitigating Gender Bias in Word Embeddings

Non-contextual word embedding models have been shown to inherit human-like stereotypical biases of gender, race and religion from the training corpora. To counter this issue, a large body of research has emerged which aims to mitigate these biases while keeping the syntactic and semantic utility of embeddings intact. This paper describes Fair Embedding Engine (FEE), a library for analysing and mitigating gender bias in word embeddings. FEE combines various state of the art techniques for quantifying, visualising and mitigating gender bias in word embeddings under a standard abstraction. FEE will aid practitioners in fast track analysis of existing debiasing methods on their embedding models. Further, it will allow rapid prototyping of new methods by evaluating their performance on a suite of standard metrics.

cs.CL↗

arXiv2020

Ranking Clarification Questions via Natural Language Inference

Given a natural language query, teaching machines to ask clarifying questions is of immense utility in practical natural language processing systems. Such interactions could help in filling information gaps for better machine comprehension of the query. For the task of ranking clarification questions, we hypothesize that determining whether a clarification question pertains to a missing entry in a given post (on QA forums such as StackExchange) could be considered as a special case of Natural Language Inference (NLI), where both the post and the most relevant clarification question point to a shared latent piece of information or context. We validate this hypothesis by incorporating representations from a Siamese BERT model fine-tuned on NLI and Multi-NLI datasets into our models and demonstrate that our best performing model obtains a relative performance improvement of 40 percent and 60 percent respectively (on the key metric of Precision@1), over the state-of-the-art baseline(s) on the two evaluation sets of the StackExchange dataset, thereby, significantly surpassing the state-of-the-art.

cs.LG↗

arXiv2020

ClarQ: A large-scale and diverse dataset for Clarification Question Generation

Question answering and conversational systems are often baffled and need help clarifying certain ambiguities. However, limitations of existing datasets hinder the development of large-scale models capable of generating and utilising clarification questions. In order to overcome these limitations, we devise a novel bootstrapping framework (based on self-supervision) that assists in the creation of a diverse, large-scale dataset of clarification questions based on post-comment tuples extracted from stackexchange. The framework utilises a neural network based architecture for classifying clarification questions. It is a two-step method where the first aims to increase the precision of the classifier and second aims to increase its recall. We quantitatively demonstrate the utility of the newly created dataset by applying it to the downstream task of question-answering. The final dataset, ClarQ, consists of ~2M examples distributed across 173 domains of stackexchange. We release this dataset in order to foster research into the field of clarification question generation with the larger goal of enhancing dialog and question answering systems.

cs.CL↗

arXiv2020

Nurse is Closer to Woman than Surgeon? Mitigating Gender-Biased Proximities in Word Embeddings

Word embeddings are the standard model for semantic and syntactic representations of words. Unfortunately, these models have been shown to exhibit undesirable word associations resulting from gender, racial, and religious biases. Existing post-processing methods for debiasing word embeddings are unable to mitigate gender bias hidden in the spatial arrangement of word vectors. In this paper, we propose RAN-Debias, a novel gender debiasing methodology which not only eliminates the bias present in a word vector but also alters the spatial distribution of its neighbouring vectors, achieving a bias-free setting while maintaining minimal semantic offset. We also propose a new bias evaluation metric - Gender-based Illicit Proximity Estimate (GIPE), which measures the extent of undue proximity in word vectors resulting from the presence of gender-based predilections. Experiments based on a suite of evaluation metrics show that RAN-Debias significantly outperforms the state-of-the-art in reducing proximity bias (GIPE) by at least 42.02%. It also reduces direct bias, adding minimal semantic disturbance, and achieves the best performance in a downstream application task (coreference resolution).

cs.CL↗

arXiv2020

On Dimensional Linguistic Properties of the Word Embedding Space

Word embeddings have become a staple of several natural language processing tasks, yet much remains to be understood about their properties. In this work, we analyze word embeddings in terms of their principal components and arrive at a number of novel and counterintuitive observations. In particular, we characterize the utility of variance explained by the principal components as a proxy for downstream performance. Furthermore, through syntactic probing of the principal embedding space, we show that the syntactic information captured by a principal component does not correlate with the amount of variance it explains. Consequently, we investigate the limitations of variance based embedding post-processing and demonstrate that such post-processing is counter-productive in sentence classification and machine translation tasks. Finally, we offer a few precautionary guidelines on applying variance based embedding post-processing and explain why non-isotropic geometry might be integral to word embedding performance.

cs.CL↗

arXiv2019

On Compositionality in Neural Machine Translation

We investigate two specific manifestations of compositionality in Neural Machine Translation (NMT) : (1) Productivity - the ability of the model to extend its predictions beyond the observed length in training data and (2) Systematicity - the ability of the model to systematically recombine known parts and rules. We evaluate a standard Sequence to Sequence model on tests designed to assess these two properties in NMT. We quantitatively demonstrate that inadequate temporal processing, in the form of poor encoder representations is a bottleneck for both Productivity and Systematicity. We propose a simple pre-training mechanism which alleviates model performance on the two properties and leads to a significant improvement in BLEU scores.

cs.CL↗

arXiv2019

Link-Layer Capacity of Downlink NOMA with Generalized Selection Combining Receivers

Non-orthogonal multiple access (NOMA) has drawn tremendous attention, being a potential candidate for the spectrum access technology for the fifth-generation (5G) and beyond 5G (B5G) wireless communications standards. Most research related to NOMA focuses on the system performance from Shannon's capacity perspective, which, although a critical system design criterion, fails to quantify the effect of delay constraints imposed by future wireless applications. In this paper, we analyze the performance of a single-input multiple-output (SIMO) two-user downlink NOMA system, in terms of the link-layer achievable rate, known as effective capacity (EC), which captures the performance of the system under a delay-limited quality-of-service (QoS) constraint. For signal combining at the receiver side, we use generalized selection combining (GSC), which bridges the performance gap between the two conventional diversity combining schemes, namely selection combining (SC) and maximal-ratio combining (MRC). We also derive two approximate expressions for the EC of NOMA-GSC which are accurate at low-SNR and at high-SNR, respectively. The analysis reveals a tradeoff between the number of implemented receiver radio-frequency (RF) chains and the achieved performance, and can be used to determine the appropriate number of paths to combine in a practical receiver design.

cs.IT↗