arXiv subjects

Sundeep Prabhakar Chepuri

Publications and source records attributed to Sundeep Prabhakar Chepuri.

At least 37 records · Page 2Linked to original sources

arXiv2022

Spectrum Surveying: Active Radio Map Estimation with Autonomous UAVs

Radio maps find numerous applications in wireless communications and mobile robotics tasks, including resource allocation, interference coordination, and mission planning. Although numerous techniques have been proposed to construct radio maps from spatially distributed measurements, the locations of such measurements are assumed predetermined beforehand. In contrast, this paper proposes spectrum surveying, where a mobile robot such as an unmanned aerial vehicle (UAV) collects measurements at a set of locations that are actively selected to obtain high-quality map estimates in a short surveying time. This is performed in two steps. First, two novel algorithms, a model-based online Bayesian estimator and a data-driven deep learning algorithm, are devised for updating a map estimate and an uncertainty metric that indicates the informativeness of measurements at each possible location. These algorithms offer complementary benefits and feature constant complexity per measurement. Second, the uncertainty metric is used to plan the trajectory of the UAV to gather measurements at the most informative locations. To overcome the combinatorial complexity of this problem, a dynamic programming approach is proposed to obtain lists of waypoints through areas of large uncertainty in linear time. Numerical experiments conducted on a realistic dataset confirm that the proposed scheme constructs accurate radio maps quickly.

eess.SP↗

arXiv2021

Graph Neural Networks with Parallel Neighborhood Aggregations for Graph Classification

We focus on graph classification using a graph neural network (GNN) model that precomputes the node features using a bank of neighborhood aggregation graph operators arranged in parallel. These GNN models have a natural advantage of reduced training and inference time due to the precomputations but are also fundamentally different from popular GNN variants that update node features through a sequential neighborhood aggregation procedure during training. We provide theoretical conditions under which a generic GNN model with parallel neighborhood aggregations (PA-GNNs, in short) are provably as powerful as the well-known Weisfeiler-Lehman (WL) graph isomorphism test in discriminating non-isomorphic graphs. Although PA-GNN models do not have an apparent relationship with the WL test, we show that the graph embeddings obtained from these two methods are injectively related. We then propose a specialized PA-GNN model, called SPIN, which obeys the developed conditions. We demonstrate via numerical experiments that the developed model achieves state-of-the-art performance on many diverse real-world datasets while maintaining the discriminative power of the WL test and the computational advantage of preprocessing graphs before the training process.

cs.LG↗

arXiv2021

Faster Rates for the Frank-Wolfe Algorithm Using Jacobi Polynomials

The Frank Wolfe algorithm (FW) is a popular projection-free alternative for solving large-scale constrained optimization problems. However, the FW algorithm suffers from a sublinear convergence rate when minimizing a smooth convex function over a compact convex set. Thus, exploring techniques that yield a faster convergence rate becomes crucial. A classic approach to obtain faster rates is to combine previous iterates to obtain the next iterate. In this work, we extend this approach to the FW setting and show that the optimal way to combine the past iterates is using a set of orthogonal Jacobi polynomials. We also a polynomial-based acceleration technique, referred to as Jacobi polynomial accelerated FW, which combines the current iterate with the past iterate using combing weights related to the Jacobi recursion. By carefully choosing parameters of the Jacobi polynomials, we obtain a faster sublinear convergence rate. We provide numerical experiments on real datasets to demonstrate the efficacy of the proposed algorithm.

math.OC↗

arXiv2021

Learning Sparse Graphs with a Core-periphery Structure

In this paper, we focus on learning sparse graphs with a core-periphery structure. We propose a generative model for data associated with core-periphery structured networks to model the dependence of node attributes on core scores of the nodes of a graph through a latent graph structure. Using the proposed model, we jointly infer a sparse graph and nodal core scores that induce dense (sparse) connections in core (respectively, peripheral) parts of the network. Numerical experiments on a variety of real-world data indicate that the proposed method learns a core-periphery structured graph from node attributes alone, while simultaneously learning core score assignments that agree well with existing works that estimate core scores using graph as input and ignoring commonly available node attributes.

cs.LG↗

arXiv2021

Multiview Variational Graph Autoencoders for Canonical Correlation Analysis

We present a novel multiview canonical correlation analysis model based on a variational approach. This is the first nonlinear model that takes into account the available graph-based geometric constraints while being scalable for processing large scale datasets with multiple views. It is based on an autoencoder architecture with graph convolutional neural network layers. We experiment with our approach on classification, clustering, and recommendation tasks on real datasets. The algorithm is competitive with state-of-the-art multiview representation learning techniques.

cs.LG↗

arXiv2021

Channel Estimation in MIMO Systems with One-bit Spatial Sigma-delta ADCs

This paper focuses on channel estimation in single-user and multi-user MIMO systems with multi-antenna base stations equipped with 1-bit spatial sigma-delta analog-to-digital converters (ADCs). A careful selection of the quantization voltage level and phase shift used in the feedback loop of 1-bit sigma-delta ADCs is critical to improve its effective resolution. We first develop a quantization noise model for 1-bit spatial sigma-delta ADCs. Using the developed noise model, we then present a two-step channel estimation algorithm to estimate a multipath channel parameterized by the gains, angles of arrival (AoAs), and angles of departure (AoDs). Specifically, in the first step, the AoAs and path gains are estimated using uplink pilots, which excite all the angles uniformly. Next, in the second step, the AoDs are estimated by progressively refining uplink beams through a recursive bisection procedure. For this algorithm, we propose a technique to select the quantization voltage level and phase shift. Through numerical simulations, we demonstrate that with the proposed parametric channel estimation algorithm, MIMO systems with 1-bit spatial sigma-delta ADCs perform significantly better than those with regular 1-bit ADCs and are on par with MIMO systems with high-resolution ADCs.

eess.SP↗

arXiv2021

Joint Communication and Radar Sensing with Reconfigurable Intelligent Surfaces

In this paper, we use a reconfigurable intelligent surface (RIS) to enhance the radar sensing and communication capabilities of a mmWave dual function radar communication system. To simultaneously localize the target and to serve the user, we propose to adaptively partition the RIS by reserving separate RIS elements for sensing and communication. We design a multi-stage hierarchical codebook to localize the target while ensuring a strong communication link to the user. We also present a method to choose the number of times to transmit the same beam in each stage to achieve a desired target localization probability of error. The proposed algorithm typically requires fewer transmissions than an exhaustive search scheme to achieve a desired target localization probability of error. Furthermore, the average spectral efficiency of the user with the proposed algorithm is found to be comparable to that of a RIS-assisted MIMO communication system without sensing capabilities and is much better than that of traditional MIMO systems without RIS.

eess.SP↗

arXiv2021

Self Calibration of Scalar and Vector Sensor Arrays

In this work, we consider the problem of joint calibration and direction-of-arrival (DOA) estimation using sensor arrays. This joint estimation problem is referred to as self calibration. Unlike many previous iterative approaches, we propose geometry independent convex optimization algorithms for jointly estimating the sensor gain and phase errors as well as the source DOAs. We derive these algorithms based on both the conventional element-space data model and the covariance data model. We focus on sparse and regular arrays formed using scalar sensors as well as vector sensors. The developed algorithms are obtained by transforming the underlying bilinear calibration model into a linear model, and subsequently by using standard convex relaxation techniques to estimate the unknown parameters. Prior to the algorithm discussion, we also derive identifiability conditions for the existence of a unique solution to the self calibration problem. To demonstrate the effectiveness of the developed techniques, numerical experiments and comparisons to the state-of-the-art methods are provided. Finally, the results from an experiment that was performed in an anechoic chamber using an acoustic vector sensor array are presented to demonstrate the usefulness of the proposed self calibration techniques.

eess.SP↗

arXiv2021

Channel Estimation for RIS-assisted Millimeter-wave MIMO Systems

Channel estimation in reconfigurable intelligent surface (RIS) assisted multiple input multiple output (MIMO) communication systems is challenging as pilots cannot be decoded at or transmitted from a RIS with only passive elements. We assume an angular model for the line-of-sight channel between the user equipment (UE) and the RIS and between the RIS and the base station (BS) with no direct path between the UE and BS. In this letter, assuming that we can compute the RIS-BS channel (up to a complex path gain) from their known locations, we first discuss the ambiguity involved in estimating the UE-RIS channel. We propose a multiple channel sounding procedure in which we observe the channel through different RIS phase shifts and present an algorithm for channel estimation to resolve this ambiguity. Through simulations, we show that the proposed method is comparable to an oracle estimator, which perfectly knows the BS, UE, and RIS locations.

eess.SP↗

arXiv2021

Learning Multi-layer Graphs and a Common Representation for Clustering

In this paper, we focus on graph learning from multi-view data of shared entities for spectral clustering. We can explain interactions between the entities in multi-view data using a multi-layer graph with a common vertex set, which represents the shared entities. The edges of different layers capture the relationships of the entities. Assuming a smoothness data model, we jointly estimate the graph Laplacian matrices of the individual graph layers and low-dimensional embedding of the common vertex set. We constrain the rank of the graph Laplacian matrices to obtain multi-component graph layers for clustering. The low-dimensional node embeddings, common to all the views, assimilate the complementary information present in the views. We propose an efficient solver based on alternating minimization to solve the proposed multi-layer multi-component graph learning problem. Numerical experiments on synthetic and real datasets demonstrate that the proposed algorithm outperforms state-of-the-art multi-view clustering techniques.

cs.LG↗

arXiv2020

Product Graph Learning from Multi-domain Data with Sparsity and Rank Constraints

In this paper, we focus on learning product graphs from multi-domain data. We assume that the product graph is formed by the Cartesian product of two smaller graphs, which we refer to as graph factors. We pose the product graph learning problem as the problem of estimating the graph factor Laplacian matrices. To capture local interactions in data, we seek sparse graph factors and assume a smoothness model for data. We propose an efficient iterative solver for learning sparse product graphs from data. We then extend this solver to infer multi-component graph factors with applications to product graph clustering by imposing rank constraints on the graph Laplacian matrices. Although working with smaller graph factors is computationally more attractive, not all graphs may readily admit an exact Cartesian product factorization. To this end, we propose efficient algorithms to approximate a graph by a nearest Cartesian product of two smaller graphs. The efficacy of the developed framework is demonstrated using several numerical experiments on synthetic data and real data.

cs.LG↗

arXiv2020

Dr-COVID: Graph Neural Networks for SARS-CoV-2 Drug Repurposing

The 2019 novel coronavirus (SARS-CoV-2) pandemic has resulted in more than a million deaths, high morbidities, and economic distress worldwide. There is an urgent need to identify medications that would treat and prevent novel diseases like the 2019 coronavirus disease (COVID-19). Drug repurposing is a promising strategy to discover new medical indications of the existing approved drugs due to several advantages in terms of the costs, safety factors, and quick results compared to new drug design and discovery. In this work, we explore computational data-driven methods for drug repurposing and propose a dedicated graph neural network (GNN) based drug repurposing model, called Dr-COVID. Although we analyze the predicted drugs in detail for COVID-19, the model is generic and can be used for any novel diseases. We construct a four-layered heterogeneous graph to model the complex interactions between drugs, diseases, genes, and anatomies. We pose drug repurposing as a link prediction problem. Specifically, we design an encoder based on the scalable inceptive graph neural network (SIGN) to generate embeddings for all the nodes in the four-layered graph and propose a quadratic norm scorer as a decoder to predict treatment for a disease. We provide a detailed analysis of the 150 potential drugs (such as Dexamethasone, Ivermectin) predicted by Dr-COVID for COVID-19 from different pharmacological classes (e.g., corticosteroids, antivirals, antiparasitic). Out of these 150 drugs, 46 drugs are currently in clinical trials. Dr-COVID is evaluated in terms of its prediction performance and its ability to rank the known treatment drugs for diseases as high as possible. For a majority of the diseases, Dr-COVID ranks the actual treatment drug in the top 15.

cs.LG↗

arXiv2020

Millimeter Wave MIMO Channel Estimation with 1-bit Spatial Sigma-delta Analog-to-Digital Converters

This paper focuses on channel estimation for mmWave MIMO systems with 1-bit spatial sigma-delta analog-to-digital converters (ADCs). The channel estimation performance with 1-bit spatial sigma-delta ADCs depends on the quantization noise modeling. Therefore, we present a new method for modeling the quantization noise by leveraging the deterministic input-output relation of the 1-bit spatial sigma-delta ADC. Using this new noise model, we propose an algorithm for channel estimation for a narrowband single-user mmWave line-of-sight MIMO system by determining the unknown angles and path attenuation that characterize the flat fading channel. Through simulations, we demonstrate that the performance of the developed method is comparable to the traditional analog systems and significantly better than the conventional 1-bit quantized systems.

eess.SP↗

arXiv2020

Federated Deep Unfolding for Sparse Recovery

This paper proposes a federated learning technique for deep algorithm unfolding with applications to sparse signal recovery and compressed sensing. We refer to this architecture as Fed-CS. Specifically, we unfold and learn the iterative shrinkage thresholding algorithm for sparse signal recovery without transporting to a central location, the training data distributed across many clients. We propose a layer-wise federated learning technique, in which each client uses local data to train a common model. Then we transmit only the model parameters of that layer from all the clients to the server, which aggregates these local models to arrive at a consensus model. The proposed layer-wise federated learning for sparse recovery is communication efficient and preserves data privacy. Through numerical experiments on synthetic and real datasets, we demonstrate Fed-CS's efficacy and present various trade-offs in terms of the number of participating clients and communications involved compared to a centralized approach of deep unfolding.

eess.SP↗

arXiv2020

Hybrid Beamforming for Active Sensing using Sparse Arrays

This paper studies hybrid beamforming for active sensing applications, such as millimeter-wave or ultrasound imaging. Hybrid beamforming can substantially lower the cost and power consumption of fully digital sensor arrays by reducing the number of active front ends. Sparse arrays can be used to further reduce hardware costs. We consider phased arrays and employ linear beamforming with possibly sparse array configurations at both the transmitter and receiver. The quality of the acquired images is improved by adding together several component images corresponding to different transmissions and receptions. In order to limit the acquisition time of an image, we formulate an optimization problem for minimizing the number of component images subject to achieving a desired point spread function. Since this problem is not convex, we propose algorithms for finding approximate solutions in the fully digital beamforming case, as well as in the more challenging hybrid and analog beamforming cases that employ quantized phase shifters. We also determine upper bounds on the number of component images needed for achieving the fully digital solution using fully analog and hybrid architectures, and derive closed-form expressions for the beamforming weights in these cases. Simulations demonstrate that a hybrid sparse array with very few elements, and even fewer front ends, can achieve the resolution of a fully digital uniform array at the expense of a longer image acquisition time.

eess.SP↗

arXiv2020

Aerial Spectrum Surveying: Radio Map Estimation with Autonomous UAVs

Radio maps are emerging as a popular means to endow next-generation wireless communications with situational awareness. In particular, radio maps are expected to play a central role in unmanned aerial vehicle (UAV) communications since they can be used to determine interference or channel gain at a spatial location where a UAV has not been before. Existing methods for radio map estimation utilize measurements collected by sensors whose locations cannot be controlled. In contrast, this paper proposes a scheme in which a UAV collects measurements along a trajectory. This trajectory is designed to obtain accurate estimates of the target radio map in a short time operation. The route planning algorithm relies on a map uncertainty metric to collect measurements at those locations where they are more informative. An online Bayesian learning algorithm is developed to update the map estimate and uncertainty metric every time a new measurement is collected, which enables real-time operation.

eess.SP↗

arXiv2020

Fast Graph Convolutional Recurrent Neural Networks

This paper proposes a Fast Graph Convolutional Neural Network (FGRNN) architecture to predict sequences with an underlying graph structure. The proposed architecture addresses the limitations of the standard recurrent neural network (RNN), namely, vanishing and exploding gradients, causing numerical instabilities during training. State-of-the-art architectures that combine gated RNN architectures, such as Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) with graph convolutions are known to improve the numerical stability during the training phase, but at the expense of the model size involving a large number of training parameters. FGRNN addresses this problem by adding a weighted residual connection with only two extra training parameters as compared to the standard RNN. Numerical experiments on the real 3D point cloud dataset corroborates the proposed architecture.

eess.SP↗

arXiv2019

Generative Adversarial Networks For Graph Data Imputation From Signed Observations

We study the problem of missing data imputation for graph signals from signed one-bit quantized observations. More precisely, we consider that the true graph data is drawn from a distribution of signals that are smooth or bandlimited on a known graph. However, instead of observing these signals, we observe a signed version of them and only at a subset of the nodes on the graph. Our goal is to estimate the true underlying graph signals from our observations. To achieve this, we propose a generative adversarial network (GAN) where the key is to incorporate graph-aware losses in the associated minimax optimization problem. We illustrate the benefits of the proposed method via numerical experiments on hand-written digits from the MNIST dataset

eess.SP↗