arXiv subjects

Arash Behboodi

Publications and source records attributed to Arash Behboodi.

At least 37 records · Page 2Linked to original sources

arXiv2024

Differentiable and Learnable Wireless Simulation with Geometric Transformers

Modelling the propagation of electromagnetic wireless signals is critical for designing modern communication systems. Wireless ray tracing simulators model signal propagation based on the 3D geometry and other scene parameters, but their accuracy is fundamentally limited by underlying modelling assumptions and correctness of parameters. In this work, we introduce Wi-GATr, a fully-learnable neural simulation surrogate designed to predict the channel observations based on scene primitives (e.g., surface mesh, antenna position and orientation). Recognizing the inherently geometric nature of these primitives, Wi-GATr leverages an equivariant Geometric Algebra Transformer that operates on a tokenizer specifically tailored for wireless simulation. We evaluate our approach on a range of tasks (i.e., signal strength and delay spread prediction, receiver localization, and geometry reconstruction) and find that Wi-GATr is accurate, fast, sample-efficient, and robust to symmetry-induced transformations. Remarkably, we find our results also translate well to the real world: Wi-GATr demonstrates more than 35% lower error than hybrid techniques, and 70% lower error than a calibrated wireless tracer.

cs.LG↗

arXiv2024

Reinforcement Learning of Adaptive Acquisition Policies for Inverse Problems

A promising way to mitigate the expensive process of obtaining a high-dimensional signal is to acquire a limited number of low-dimensional measurements and solve an under-determined inverse problem by utilizing the structural prior about the signal. In this paper, we focus on adaptive acquisition schemes to save further the number of measurements. To this end, we propose a reinforcement learning-based approach that sequentially collects measurements to better recover the underlying signal by acquiring fewer measurements. Our approach applies to general inverse problems with continuous action spaces and jointly learns the recovery algorithm. Using insights obtained from theoretical analysis, we also provide a probabilistic design for our methods using variational formulation. We evaluate our approach on multiple datasets and with two measurement spaces (Gaussian, Radon). Our results confirm the benefits of adaptive strategies in low-acquisition horizon settings.

cs.LG↗

arXiv2024

Variational Learning ISTA

Compressed sensing combines the power of convex optimization techniques with a sparsity-inducing prior on the signal space to solve an underdetermined system of equations. For many problems, the sparsifying dictionary is not directly given, nor its existence can be assumed. Besides, the sensing matrix can change across different scenarios. Addressing these issues requires solving a sparse representation learning problem, namely dictionary learning, taking into account the epistemic uncertainty of the learned dictionaries and, finally, jointly learning sparse representations and reconstructions under varying sensing matrix conditions. We address both concerns by proposing a variant of the LISTA architecture. First, we introduce Augmented Dictionary Learning ISTA (A-DLISTA), which incorporates an augmentation module to adapt parameters to the current measurement setup. Then, we propose to learn a distribution over dictionaries via a variational approach, dubbed Variational Learning ISTA (VLISTA). VLISTA exploits A-DLISTA as the likelihood model and approximates a posterior distribution over the dictionaries as part of an unfolded LISTA-based recovery algorithm. As a result, VLISTA provides a probabilistic way to jointly learn the dictionary distribution and the reconstruction algorithm with varying sensing matrices. We provide theoretical and experimental support for our architecture and show that our model learns calibrated uncertainties.

cs.LG↗

arXiv2024

Simulating, Fast and Slow: Learning Policies for Black-Box Optimization

In recent years, solving optimization problems involving black-box simulators has become a point of focus for the machine learning community due to their ubiquity in science and engineering. The simulators describe a forward process $f_{\mathrm{sim}}: (ψ, x) \rightarrow y$ from simulation parameters $ψ$ and input data $x$ to observations $y$, and the goal of the optimization problem is to find parameters $ψ$ that minimize a desired loss function. Sophisticated optimization algorithms typically require gradient information regarding the forward process, $f_{\mathrm{sim}}$, with respect to the parameters $ψ$. However, obtaining gradients from black-box simulators can often be prohibitively expensive or, in some cases, impossible. Furthermore, in many applications, practitioners aim to solve a set of related problems. Thus, starting the optimization ``ab initio", i.e. from scratch, each time might be inefficient if the forward model is expensive to evaluate. To address those challenges, this paper introduces a novel method for solving classes of similar black-box optimization problems by learning an active learning policy that guides a differentiable surrogate's training and uses the surrogate's gradients to optimize the simulation parameters with gradient descent. After training the policy, downstream optimization of problems involving black-box simulators requires up to $\sim$90\% fewer expensive simulator calls compared to baselines such as local surrogate-based approaches, numerical optimization, and Bayesian methods.

cs.LG↗

arXiv2024

Pruning vs Quantization: Which is Better?

Neural network pruning and quantization techniques are almost as old as neural networks themselves. However, to date only ad-hoc comparisons between the two have been published. In this paper, we set out to answer the question on which is better: neural network quantization or pruning? By answering this question, we hope to inform design decisions made on neural network hardware going forward. We provide an extensive comparison between the two techniques for compressing deep neural networks. First, we give an analytical comparison of expected quantization and pruning error for general data distributions. Then, we provide lower bounds for the per-layer pruning and quantization error in trained networks, and compare these to empirical error after optimization. Finally, we provide an extensive experimental comparison for training 8 large-scale models on 3 tasks. Our results show that in most cases quantization outperforms pruning. Only in some scenarios with very high compression ratio, pruning might be beneficial from an accuracy standpoint.

cs.LG↗

arXiv2024

Vision-Assisted Digital Twin Creation for mmWave Beam Management

In the context of communication networks, digital twin technology provides a means to replicate the radio frequency (RF) propagation environment as well as the system behaviour, allowing for a way to optimize the performance of a deployed system based on simulations. One of the key challenges in the application of Digital Twin technology to mmWave systems is the prevalent channel simulators' stringent requirements on the accuracy of the 3D Digital Twin, reducing the feasibility of the technology in real applications. We propose a practical Digital Twin creation pipeline and a channel simulator, that relies only on a single mounted camera and position information. We demonstrate the performance benefits compared to methods that do not explicitly model the 3D environment, on downstream sub-tasks in beam acquisition, using the real-world dataset of the DeepSense6G challenge

eess.SP↗

arXiv2023

Algebraic Topological Networks via the Persistent Local Homology Sheaf

In this work, we introduce a novel approach based on algebraic topology to enhance graph convolution and attention modules by incorporating local topological properties of the data. To do so, we consider the framework of sheaf neural networks, which has been previously leveraged to incorporate additional structure into graph neural networks' features and construct more expressive, non-isotropic messages. Specifically, given an input simplicial complex (e.g. generated by the cliques of a graph or the neighbors in a point cloud), we construct its local homology sheaf, which assigns to each node the vector space of its local homology. The intermediate features of our networks live in these vector spaces and we leverage the associated sheaf Laplacian to construct more complex linear messages between them. Moreover, we extend this approach by considering the persistent version of local homology associated with a weighted simplicial complex (e.g., built from pairwise distances of nodes embeddings). This i) solves the problem of the lack of a natural choice of basis for the local homology vector spaces and ii) makes the sheaf itself differentiable, which enables our models to directly optimize the topology of their intermediate features.

cs.LG↗

arXiv2023

Transformer-Based Neural Surrogate for Link-Level Path Loss Prediction from Variable-Sized Maps

Estimating path loss for a transmitter-receiver location is key to many use-cases including network planning and handover. Machine learning has become a popular tool to predict wireless channel properties based on map data. In this work, we present a transformer-based neural network architecture that enables predicting link-level properties from maps of various dimensions and from sparse measurements. The map contains information about buildings and foliage. The transformer model attends to the regions that are relevant for path loss prediction and, therefore, scales efficiently to maps of different size. Further, our approach works with continuous transmitter and receiver coordinates without relying on discretization. In experiments, we show that the proposed model is able to efficiently learn dominant path losses from sparse training data and generalizes well when tested on novel maps.

cs.LG↗

arXiv2022

Spatially Sparse Precoding in Wideband Hybrid Terahertz Massive MIMO Systems

In terahertz (THz) massive multiple-input multiple-output (MIMO) systems, the combination of huge bandwidth and massive antennas results in severe beam split, thus making the conventional phase-shifter based hybrid precoding architecture ineffective. With the incorporation of true-time-delay (TTD) lines in the hardware implementation of the analog precoders, delay-phase precoding (DPP) emerges as a promising architecture to effectively overcome beam split. However, existing DPP approaches suffer from poor performance, high complexity, and weak robustness in practical THz channels. In this paper, we propose a novel DPP approach in wideband THz massive MIMO systems. First, the optimization problem is converted into a compressive sensing (CS) form, which can be solved by the extended spatially sparse precoding (SSP) algorithm. To compensate for beam split, frequency-dependent measurement matrices are introduced, which can be approximately realized by feasible phase and delay codebooks. Then, several efficient atom selection techniques are developed to further reduce the complexity of extended SSP. In simulation, the proposed DPP approach achieves superior performance, complexity, and robustness by using it alone or in combination with existing DPP approaches.

cs.IT↗

arXiv2022

Beyond Codebook-Based Analog Beamforming at mmWave: Compressed Sensing and Machine Learning Methods

Analog beamforming is the predominant approach for millimeter wave (mmWave) communication given its favorable characteristics for limited-resource devices. In this work, we aim at reducing the spectral efficiency gap between analog and digital beamforming methods. We propose a method for refined beam selection based on the estimated raw channel. The channel estimation, an underdetermined problem, is solved using compressed sensing (CS) methods leveraging angular domain sparsity of the channel. To reduce the complexity of CS methods, we propose dictionary learning iterative soft-thresholding algorithm, which jointly learns the sparsifying dictionary and signal reconstruction. We evaluate the proposed method on a realistic mmWave setup and show considerable performance improvement with respect to code-book based analog beamforming approaches.

eess.SP↗

arXiv2022

A PAC-Bayesian Generalization Bound for Equivariant Networks

Equivariant networks capture the inductive bias about the symmetry of the learning task by building those symmetries into the model. In this paper, we study how equivariance relates to generalization error utilizing PAC Bayesian analysis for equivariant networks, where the transformation laws of feature spaces are determined by group representations. By using perturbation analysis of equivariant networks in Fourier domain for each layer, we derive norm-based PAC-Bayesian generalization bounds. The bound characterizes the impact of group size, and multiplicity and degree of irreducible representations on the generalization error and thereby provide a guideline for selecting them. In general, the bound indicates that using larger group size in the model improves the generalization error substantiated by extensive numerical experiments.

cs.LG↗

arXiv2022

Multi-Modal Beam Prediction Challenge 2022: Towards Generalization

Beam management is a challenging task for millimeter wave (mmWave) and sub-terahertz communication systems, especially in scenarios with highly-mobile users. Leveraging external sensing modalities such as vision, LiDAR, radar, position, or a combination of them, to address this beam management challenge has recently attracted increasing interest from both academia and industry. This is mainly motivated by the dependency of the beam direction decision on the user location and the geometry of the surrounding environment -- information that can be acquired from the sensory data. To realize the promised beam management gains, such as the significant reduction in beam alignment overhead, in practice, however, these solutions need to account for important aspects. For example, these multi-modal sensing aided beam selection approaches should be able to generalize their learning to unseen scenarios and should be able to operate in realistic dense deployments. The "Multi-Modal Beam Prediction Challenge 2022: Towards Generalization" competition is offered to provide a platform for investigating these critical questions. In order to facilitate the generalizability study, the competition offers a large-scale multi-modal dataset with co-existing communication and sensing data collected across multiple real-world locations and different times of the day. In this paper, along with the detailed descriptions of the problem statement and the development dataset, we provide a baseline solution that utilizes the user position data to predict the optimal beam indices. The objective of this challenge is to go beyond a simple feasibility study and enable necessary research in this direction, paving the way towards generalizable multi-modal sensing-aided beam management for real-world future communication systems.

eess.SP↗

arXiv2022

Learning Perturbations for Soft-Output Linear MIMO Demappers

Tree-based demappers for multiple-input multiple-output (MIMO) detection such as the sphere decoder can achieve near-optimal performance but incur high computational cost due to their sequential nature. In this paper, we propose the perturbed linear demapper (PLM), which is a novel data-driven model for computing soft outputs in parallel. To achieve this, the PLM learns a distribution centered on an initial linear estimate and a log-likelihood ratio clipping parameter using end-to-end Bayesian optimization. Furthermore, we show that lattice-reduction can be naturally incorporated into the PLM pipeline, which allows to trade off computational cost against coded block error rate reduction. We find that the optimized PLM can achieve near maximum-likelihood (ML) performance in Rayleigh channels, making it an efficient alternative to tree-based demappers.

cs.IT↗

arXiv2022

Position Aided Beam Prediction in the Real World: How Useful GPS Locations Actually Are?

Millimeter-wave (mmWave) communication systems rely on narrow beams for achieving sufficient receive signal power. Adjusting these beams is typically associated with large training overhead, which becomes particularly critical for highly-mobile applications. Intuitively, since optimal beam selection can benefit from the knowledge of the positions of communication terminals, there has been increasing interest in leveraging position data to reduce the overhead in mmWave beam prediction. Prior work, however, studied this problem using only synthetic data that generally does not accurately represent real-world measurements. In this paper, we investigate position-aided beam prediction using a real-world large-scale dataset to derive insights into precisely how much overhead can be saved in practice. Furthermore, we analyze which machine learning algorithms perform best, what factors degrade inference performance in real data, and which machine learning metrics are more meaningful in capturing the actual communication system performance.

eess.SP↗

arXiv2022

Quantized Sparse Weight Decomposition for Neural Network Compression

In this paper, we introduce a novel method of neural network weight compression. In our method, we store weight tensors as sparse, quantized matrix factors, whose product is computed on the fly during inference to generate the target model's weights. We use projected gradient descent methods to find quantized and sparse factorization of the weight tensors. We show that this approach can be seen as a unification of weight SVD, vector quantization, and sparse PCA. Combined with end-to-end fine-tuning our method exceeds or is on par with previous state-of-the-art methods in terms of the trade-off between accuracy and model size. Our method is applicable to both moderate compression regimes, unlike vector quantization, and extreme compression regimes.

cs.LG↗

arXiv2022

Equivariant Priors for Compressed Sensing with Unknown Orientation

In compressed sensing, the goal is to reconstruct the signal from an underdetermined system of linear measurements. Thus, prior knowledge about the signal of interest and its structure is required. Additionally, in many scenarios, the signal has an unknown orientation prior to measurements. To address such recovery problems, we propose using equivariant generative models as a prior, which encapsulate orientation information in their latent space. Thereby, we show that signals with unknown orientations can be recovered with iterative gradient descent on the latent space of these models and provide additional theoretical recovery guarantees. We construct an equivariant variational autoencoder and use the decoder as generative prior for compressed sensing. We discuss additional potential gains of the proposed approach in terms of convergence and latency.

cs.LG↗

arXiv2022

Deep Learning-based Channel Estimation for Wideband Hybrid MmWave Massive MIMO

Hybrid analog-digital (HAD) architecture is widely adopted in practical millimeter wave (mmWave) massive multiple-input multiple-output (MIMO) systems to reduce hardware cost and energy consumption. However, channel estimation in the context of HAD is challenging due to only limited radio frequency (RF) chains at transceivers. Although various compressive sensing (CS) algorithms have been developed to solve this problem by exploiting inherent channel sparsity and sparsity structures, practical effects, such as power leakage and beam squint, can still make the real channel features deviate from the assumed models and result in performance degradation. Also, the high complexity of CS algorithms caused by a large number of iterations hinders their applications in practice. To tackle these issues, we develop a deep learning (DL)-based channel estimation approach where the sparse Bayesian learning (SBL) algorithm is unfolded into a deep neural network (DNN). In each SBL layer, Gaussian variance parameters of the sparse angular domain channel are updated by a tailored DNN, which is able to effectively capture complicated channel sparsity structures in various domains. Besides, the measurement matrix is jointly optimized for performance improvement. Then, the proposed approach is extended to the multi-block case where channel correlation in time is further exploited to adaptively predict the measurement matrix and facilitate the update of Gaussian variance parameters. Based on simulation results, the proposed approaches significantly outperform existing approaches but with reduced complexity.

cs.IT↗

arXiv2022

The Restricted Isometry Property of Block Diagonal Matrices for Group-Sparse Signal Recovery

Group-sparsity is a common low-complexity signal model with widespread application across various domains of science and engineering. The recovery of such signal ensembles from compressive measurements has been extensively studied in the literature under the assumption that measurement operators are modeled as densely populated random matrices. In this paper, we turn our attention to an acquisition model intended to ease the energy consumption of sensing devices by splitting the measurements up into distinct signal blocks. More precisely, we present uniform guarantees for group-sparse signal recovery in the scenario where a number of sensors obtain independent partial signal observations modeled by block diagonal measurement matrices. We establish a group-sparse variant of the classical restricted isometry property for block diagonal sensing matrices acting on group-sparse vectors, and provide conditions under which subgaussian block diagonal random matrices satisfy this group-RIP with high probability. Two different scenarios are considered in particular. In the first scenario, we assume that each sensor is equipped with an independently drawn measurement matrix. We later lift this requirement by considering measurement matrices with constant block diagonal entries. In other words, every sensor is equipped with a copy of the same prototype matrix. The problem of establishing the group-RIP is cast into a form in which one needs to establish the concentration behavior of the suprema of chaos processes which involves estimating Talagrand's $γ_2$ functional. As a side effect of the proof, we present an extension to Maurey's empirical method to provide new bounds on the covering number of sets consisting of finite convex combinations of possibly infinite sets.

cs.IT↗