Imperial College London

ProfessorDenizGunduz

Faculty of EngineeringDepartment of Electrical and Electronic Engineering

Professor in Information Processing
 
 
 
//

Contact

 

+44 (0)20 7594 6218d.gunduz Website

 
 
//

Assistant

 

Ms Joan O'Brien +44 (0)20 7594 6316

 
//

Location

 

1016Electrical EngineeringSouth Kensington Campus

//

Summary

 

Publications

Publication Type
Year
to

364 results found

Shiri I, Razeghi B, Vafaei Sadr A, Amini M, Salimi Y, Ferdowsi S, Boor P, Gündüz D, Voloshynovskiy S, Zaidi Het al., 2023, Multi-institutional PET/CT image segmentation using federated deep transformer learning., Comput Methods Programs Biomed, Vol: 240

BACKGROUND AND OBJECTIVE: Generalizable and trustworthy deep learning models for PET/CT image segmentation necessitates large diverse multi-institutional datasets. However, legal, ethical, and patient privacy issues challenge sharing of datasets between different centers. To overcome these challenges, we developed a federated learning (FL) framework for multi-institutional PET/CT image segmentation. METHODS: A dataset consisting of 328 FL (HN) cancer patients who underwent clinical PET/CT examinations gathered from six different centers was enrolled. A pure transformer network was implemented as fully core segmentation algorithms using dual channel PET/CT images. We evaluated different frameworks (single center-based, centralized baseline, as well as seven different FL algorithms) using 68 PET/CT images (20% of each center data). In particular, the implemented FL algorithms include clipping with the quantile estimator (ClQu), zeroing with the quantile estimator (ZeQu), federated averaging (FedAvg), lossy compression (LoCo), robust aggregation (RoAg), secure aggregation (SeAg), and Gaussian differentially private FedAvg with adaptive quantile clipping (GDP-AQuCl). RESULTS: The Dice coefficient was 0.80±0.11 for both centralized and SeAg FL algorithms. All FL approaches achieved centralized learning model performance with no statistically significant differences. Among the FL algorithms, SeAg and GDP-AQuCl performed better than the other techniques. However, there was no statistically significant difference. All algorithms, except the center-based approach, resulted in relative errors less than 5% for SUVmax and SUVmean for all FL and centralized methods. Centralized and FL algorithms significantly outperformed the single center-based baseline. CONCLUSIONS: The developed FL-based (with centralized method performance) algorithms exhibited promising performance for HN tumor segmentation from PET/CT images.

Journal article

Xia J-Y, Li S, Huang J-J, Yang Z, Jaimoukha IM, Gunduz Det al., 2023, Metalearning-based alternating minimization algorithm for nonconvex optimization, IEEE Transactions on Neural Networks and Learning Systems, Vol: 34, Pages: 5366-5380, ISSN: 1045-9227

In this article, we propose a novel solution for nonconvex problems of multiple variables, especially for those typically solved by an alternating minimization (AM) strategy that splits the original optimization problem into a set of subproblems corresponding to each variable and then iteratively optimizes each subproblem using a fixed updating rule. However, due to the intrinsic nonconvexity of the original optimization problem, the optimization can be trapped into a spurious local minimum even when each subproblem can be optimally solved at each iteration. Meanwhile, learning-based approaches, such as deep unfolding algorithms, have gained popularity for nonconvex optimization; however, they are highly limited by the availability of labeled data and insufficient explainability. To tackle these issues, we propose a meta-learning based alternating minimization (MLAM) method that aims to minimize a part of the global losses over iterations instead of carrying minimization on each subproblem, and it tends to learn an adaptive strategy to replace the handcrafted counterpart resulting in advance on superior performance. The proposed MLAM maintains the original algorithmic principle, providing certain interpretability. We evaluate the proposed method on two representative problems, namely, bilinear inverse problem: matrix completion and nonlinear problem: Gaussian mixture models. The experimental results validate the proposed approach outperforms AM-based methods.

Journal article

Shao Y, Ozfatura E, Perotti AG, Popović BM, Gündüz Det al., 2023, AttentionCode: Ultra-Reliable Feedback Codes for Short-Packet Communications, IEEE Transactions on Communications, Vol: 71, Pages: 4437-4452, ISSN: 0090-6778

— Ultra-reliable short-packet communication is a major challenge in future wireless networks with critical applications. To achieve ultra-reliable communications beyond 99.999%, this paper envisions a new interaction-based communication paradigm that exploits feedback from the receiver. We present AttentionCode, a new class of feedback codes leveraging deep learning (DL) technologies. The underpinnings of AttentionCode are three architectural innovations: AttentionNet, input restructuring, and adaptation to fading channels, accompanied by several training methods, including large-batch training, distributed learning, look-ahead optimizer, training-test signal-to-noise ratio (SNR) mismatch, and curriculum learning. The training methods can potentially be generalized to other wireless communication applications with machine learning. Numerical experiments verify that AttentionCode establishes a new state of the art among all DL-based feedback codes in both additive white Gaussian noise (AWGN) channels and fading channels. In AWGN channels with noiseless feedback, for example, AttentionCode achieves a block error rate (BLER) of 10−7 when the forward channel SNR is 0 dB for a block size of 50 bits, demonstrating the potential of AttentionCode to provide ultra-reliable short-packet communications.

Journal article

Erdemir E, Tung T-Y, Dragotti PL, Gündüz Det al., 2023, Generative joint source-channel coding for semantic image transmission, IEEE Journal on Selected Areas in Communications, Vol: 41, Pages: 2645-2657, ISSN: 0733-8716

Recent works have shown that joint source-channel coding (JSCC) schemes using deep neural networks (DNNs), called DeepJSCC, provide promising results in wireless image transmission. However, these methods mostly focus on the distortion of the reconstructed signals with respect to the input image, rather than their perception by humans. However, focusing on traditional distortion metrics alone does not necessarily result in high perceptual quality, especially in extreme physical conditions, such as very low bandwidth compression ratio (BCR) and low signal-to-noise ratio (SNR) regimes. In this work, we propose two novel JSCC schemes that leverage the perceptual quality of deep generative models (DGMs) for wireless image transmission, namely InverseJSCC and GenerativeJSCC. While the former is an inverse problem approach to DeepJSCC, the latter is an end-to-end optimized JSCC scheme. In both, we optimize a weighted sum of mean squared error (MSE) and learned perceptual image patch similarity (LPIPS) losses, which capture more semantic similarities than other distortion metrics. InverseJSCC performs denoising on the distorted reconstructions of a DeepJSCC model by solving an inverse optimization problem using the pre-trained style-based generative adversarial network (StyleGAN). Our simulation results show that InverseJSCC significantly improves the state-of-the-art DeepJSCC in terms of perceptual quality in edge cases. In GenerativeJSCC, we carry out end-to-end training of an encoder and a StyleGAN-based decoder, and show that GenerativeJSCC significantly outperforms DeepJSCC both in terms of distortion and perceptual quality.

Journal article

Lan Q, Zeng Q, Popovski P, Gunduz D, Huang Ket al., 2023, Progressive feature transmission for split classification at the wireless edge, IEEE Transactions on Wireless Communications, Vol: 22, Pages: 3837-3852, ISSN: 1536-1276

We consider the scenario of inference at the wire-less edge , in which devices are connected to an edge server and ask the server to carry out remote classification, that is, classify data samples available at edge devices. This requires the edge devices to upload high-dimensional features of samples over resource-constrained wireless channels, which creates a communication bottleneck. The conventional feature pruning solution would require the device to have access to the inference model, which is not available in the current split inference scenario. To address this issue, we propose the progressive feature transmission (ProgressFTX) protocol, which minimizes the overhead by progressively transmitting features until a target confidence level is reached. A control policy is proposed to accelerate inference, comprising two key operations: importance-aware feature selection at the server and transmission-termination control . For the former, it is shown that selecting the most important features, characterized by the largest discriminant gains of the corresponding feature dimensions, achieves a sub-optimal performance. For the latter, the proposed policy is shown to exhibit a threshold structure. Specifically, the transmission is stopped when the incremental uncertainty reduction by further feature transmission is outweighed by its communication cost. The indices of the selected features and transmission decision are fed back to the device in each slot. The control policy is first derived for the tractable case of linear classification, and then extended to the more complex case of classification using a convolutional neural network . Both Gaussian and fading channels are considered. Experimental results are obtained for both a statistical data model and a real dataset. It is shown that ProgressFTX can substantially reduce the communication latency compared to conventional feature pruning and random feature transmission strategies.

Journal article

Buyukates B, Ozfatura E, Ulukus S, Gunduz Det al., 2023, Gradient coding with dynamic clustering for straggler-tolerant distributed learning, IEEE Transactions on Communications, Vol: 71, Pages: 3317-3332, ISSN: 0090-6778

Distributed implementations are crucial in speeding up large scale machine learning applications. Distributed gradient descent (GD) is widely employed to parallelize the learning task by distributing the dataset across multiple workers. A significant performance bottleneck for the per-iteration completion time in distributed synchronous GD is straggling workers. Coded distributed computation techniques have been introduced recently to mitigate stragglers and to speed up GD iterations by assigning redundant computations to workers. In this paper, we introduce a novel paradigm of dynamic coded computation, which assigns redundant data to workers to acquire the flexibility to dynamically choose from among a set of possible codes depending on the past straggling behavior. In particular, we propose gradient coding (GC) with dynamic clustering, called GC-DC, and regulate the number of stragglers in each cluster by dynamically forming the clusters at each iteration. With time-correlated straggling behavior, GC-DC adapts to the straggling behavior over time; in particular, at each iteration, GC-DC aims at distributing the stragglers across clusters as uniformly as possible based on the past straggler behavior. For both homogeneous and heterogeneous worker models, we numerically show that GC-DC provides significant improvements in the average per-iteration completion time without an increase in the communication load compared to the original GC scheme.

Journal article

Razeghi B, Calmon FP, Gunduz D, Voloshynovskiy Set al., 2023, Bottlenecks CLUB: Unifying information-theoretic trade-offs among complexity, leakage, and utility, IEEE Transactions on Information Forensics and Security, Vol: 18, Pages: 2060-2075, ISSN: 1556-6013

Bottleneck problems are an important class of optimization problems that have recently gained increasing attention in the domain of machine learning and information theory. They are widely used in generative models, fair machine learning algorithms, design of privacy-assuring mechanisms, and appear as information-theoretic performance bounds in various multi-user communication problems. In this work, we propose a general family of optimization problems, termed as complexity-leakage-utility bottleneck (CLUB) model, which (i) provides a unified theoretical framework that generalizes most of the state-of-the-art literature for the information-theoretic privacy models, (ii) establishes a new interpretation of the popular generative and discriminative models, (iii) constructs new insights for the generative compression models, and (iv) can be used to obtain fair generative models. We first formulate the CLUB model as a complexity-constrained privacy-utility optimization problem. We then connect it with the closely related bottleneck problems, namely information bottleneck (IB), privacy funnel (PF), deterministic IB (DIB), conditional entropy bottleneck (CEB), and conditional PF (CPF). We show that the CLUB model generalizes all these problems as well as most other information-theoretic privacy models. Then, we construct the deep variational CLUB (DVCLUB) models by employing neural networks to parameterize variational approximations of the associated information quantities. Building upon these information quantities, we present unified objectives of the supervised and unsupervised DVCLUB models. Leveraging the DVCLUB model in an unsupervised setup, we then connect it with state-of-the-art generative models, such as variational auto-encoders (VAEs), generative adversarial networks (GANs), as well as the Wasserstein GAN (WGAN), Wasserstein auto-encoder (WAE), and adversarial auto-encoder (AAE) models through the optimal transport (OT) problem. We then show that the DVCLUB m

Journal article

Shao Y, Gunduz D, Liew SC, 2023, Bayesian over-the-air computation, IEEE Journal on Selected Areas in Communications, Vol: 41, Pages: 589-606, ISSN: 0733-8716

As an important piece of the multi-tier computing architecture for future wireless networks, over-the-air computation (OAC) enables efficient function computation in multiple-access edge computing, where a fusion center aims to compute a function of the data distributed at edge devices. Existing OAC relies exclusively on the maximum likelihood (ML) estimation at the fusion center to recover the arithmetic sum of the transmitted signals from different devices. ML estimation, however, is much susceptible to noise. In particular, in the misaligned OAC where there are channel misalignments among received signals, ML estimation suffers from severe error propagation and noise enhancement. To address these challenges, this paper puts forth a Bayesian approach by letting each edge device transmit two pieces of statistical information to the fusion center such that Bayesian estimators can be devised to tackle the misalignments. Numerical and simulation results verify that, 1) For the aligned and synchronous OAC, our linear minimum mean squared error (LMMSE) estimator significantly outperforms the ML estimator. In the low signal-to-noise ratio (SNR) regime, the LMMSE estimator reduces the mean squared error (MSE) by at least 6 dB; in the high SNR regime, the LMMSE estimator lowers the error floor of MSE by 86.4%; 2) For the asynchronous OAC, our LMMSE and sum-product maximum a posteriori (SP-MAP) estimators are on an equal footing in terms of the MSE performance, and are significantly better than the ML estimator. Moreover, the SP-MAP estimator is computationally efficient, the complexity of which grows linearly with the packet length.

Journal article

Shao Y, Gunduz D, 2023, Semantic communications with discrete-time analog transmission: a PAPR perspective, IEEE Wireless Communications Letters, Vol: 12, Pages: 510-514, ISSN: 2162-2337

Recent progress in deep learning (DL)-based joint source-channel coding (DeepJSCC) has led to a new paradigm of semantic communications. Two salient features of DeepJSCC-based semantic communications are the exploitation of semantic-aware features directly from the source signal, and the discrete-time analog transmission (DTAT) of these features. Compared with traditional digital communications, semantic communications with DeepJSCC provide superior reconstruction performance at the receiver and graceful degradation with diminishing channel quality, but also exhibit a large peak-to-average power ratio (PAPR) in the transmitted signal. An open question has been whether the gains of DeepJSCC come at the expense of high-PAPR continuous-amplitude signal, which can limit its adoption in practice. In this letter, we first show that conventional DeepJSCC does suffer from high PAPR. Then, we explore three PAPR reduction techniques and confirm that the superior image reconstruction performance of DeepJSCC can be retained while the PAPR is suppressed to an acceptable level. This is an important step towards the implementation of DeepJSCC in practical semantic communication systems.

Journal article

Sreekumar S, Gündüz D, 2023, Distributed hypothesis testing over a noisy channel: error-exponents trade-off, Entropy: international and interdisciplinary journal of entropy and information studies, Vol: 25, Pages: 1-33, ISSN: 1099-4300

A two-terminal distributed binary hypothesis testing problem over a noisy channel is studied. The two terminals, called the observer and the decision maker, each has access to n independent and identically distributed samples, denoted by U and V, respectively. The observer communicates to the decision maker over a discrete memoryless channel, and the decision maker performs a binary hypothesis test on the joint probability distribution of (U,V) based on V and the noisy information received from the observer. The trade-off between the exponents of the type I and type II error probabilities is investigated. Two inner bounds are obtained, one using a separation-based scheme that involves type-based compression and unequal error-protection channel coding, and the other using a joint scheme that incorporates type-based hybrid coding. The separation-based scheme is shown to recover the inner bound obtained by Han and Kobayashi for the special case of a rate-limited noiseless channel, and also the one obtained by the authors previously for a corner point of the trade-off. Finally, we show via an example that the joint scheme achieves a strictly tighter bound than the separation-based scheme for some points of the error-exponents trade-off.

Journal article

Mital N, Ozyilkan E, Garjani A, Gunduz Det al., 2023, Neural distributed image compression with cross-attention feature alignment, 2023 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Publisher: IEEE, Pages: 2497-2506

We consider the problem of compressing an information source when a correlated one is available as side information only at the decoder side, which is a special case of the distributed source coding problem in information theory. In particular, we consider a pair of stereo images, which have overlapping fields of view, and are captured by a synchronized and calibrated pair of cameras as correlated image sources. In previously proposed methods, the encoder transforms the input image to a latent representation using a deep neural network, and compresses the quantized latent representation losslessly using entropy coding. The decoder decodes the entropy-coded quantized latent representation, and reconstructs the input image using this representation and the available side information. In the proposed method, the decoder employs a cross-attention module to align the feature maps obtained from the received latent representation of the input image and a latent representation of the side information. We argue that aligning the correlated patches in the feature maps allows better utilization of the side information. We empirically demonstrate the competitiveness of the proposed algorithm on KITTI and Cityscape datasets of stereo image pairs. Our experimental results show that the proposed architecture is able to exploit the decoder-only side information in a more efficient manner compared to previous works.

Conference paper

Isik B, Pase F, Gunduz D, Weissman T, Zorzi Met al., 2023, Sparse random networks for communication-efficient federated learning, International Conference on Learning Representations

One main challenge in federated learning is the large communication cost of ex-changing weight updates from clients to the server at each round. While prior work has made great progress in compressing the weight updates through gradient compression methods, we propose a radically different approach that does not update the weights at all. Instead, our method freezes the weights at their initial random values and learns how to sparsify the random network for the best performance. To this end, the clients collaborate in training a stochastic binary maskto find the optimal sparse random network within the original one. At the end of the training, the final model is a sparse network with random weights – or a sub-network inside the dense random network. We show improvements in accuracy, communication (less than 1 bit per parameter (bpp)), convergence speed, and final model size (less than 1 bpp) over relevant baselines on MNIST, EMNIST, CIFAR-10, and CIFAR-100 datasets, in the low bitrate regime.

Conference paper

Gunduz D, 2023, Sparse Random Networks for Communication-Efficient Federated Learning, Eleventh International Conference on Learning Representations (ICLR)

Conference paper

Song J, Gunduz D, Choi W, 2023, Optimal Scheduling Policy for Minimizing Age of Information With a Relay, IEEE Internet of Things Journal

We investigate age of information in an Internet-of-things (IoT) sensor network where a single relay terminal connects multiple IoT sensors to their corresponding destination nodes. In order to minimize average weighted sum AoI, joint optimization of sampling and updating policy of a relay is studied. For error-free and symmetric case where weights are identical, the necessary and sufficient condition for optimal policy is figured out. We also obtain the minimum average sum AoI in a closed-form expression which can be interpreted as the fundamental limit of sum AoI in a single relay network. Moreover, we prove that the greedy policy is optimal for minimizing the average sum AoI at the destination nodes in the error-prone symmetric network. For general case where weights are arbitrarily given, we propose a scheduling policy obtained via deep reinforcement learning.

Journal article

Shao Y, Liew SC, Gunduz D, 2023, Denoising Noisy Neural Networks: A Bayesian Approach With Compensation, IEEE Transactions on Signal Processing, Vol: 71, Pages: 2460-2474, ISSN: 1053-587X

Deep neural networks (DNNs) with noisy weights, which we refer to as noisy neural networks (NoisyNNs), arise from the training and inference of DNNs in the presence of noise. NoisyNNs emerge in many new applications, including the wireless transmission of DNNs, the efficient deployment or storage of DNNs in analog devices, and the truncation or quantization of DNN weights. This article studies a fundamental problem of NoisyNNs: how to reconstruct the DNN weights from their noisy manifestations. While prior works relied exclusively on the maximum likelihood (ML) estimation, this article puts forth a denoising approach to reconstruct DNNs with the aim of maximizing the inference accuracy of the reconstructed models. The superiority of our denoiser is rigorously proven in two small-scale problems, wherein we consider a quadratic neural network function and a shallow feedforward neural network, respectively. When applied to advanced learning tasks with modern DNN architectures, our denoiser exhibits significantly better performance than the ML estimator. Consider the average test accuracy of the denoised DNN model versus the weight variance to noise power ratio (WNR) performance. When denoising a noisy ResNet34 model arising from noisy inference, our denoiser outperforms ML estimation by up to 4.1 dB to achieve a test accuracy of 60%. When denoising a noisy ResNet18 model arising from noisy training, our denoiser outperforms ML estimation by 13.4 dB and 8.3 dB to achieve test accuracies of 60% and 80%, respectively.

Journal article

Gunduz D, Qin Z, Estella Aguerri I, Dhillon HS, Yang Z, Yener A, Wong KK, Chae C-Bet al., 2023, Guest Editorial Special Issue on Beyond Transmitting Bits: Context, Semantics, and Task-Oriented Communications, IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, Vol: 41, Pages: 1-4, ISSN: 0733-8716

Journal article

Gunduz D, Qin Z, Aguerri IE, Dhillon HS, Yang Z, Yener A, Wong KK, Chae C-Bet al., 2023, Beyond transmitting bits: context, semantics, and task-oriented communications, IEEE Journal on Selected Areas in Communications, Vol: 41, Pages: 5-41, ISSN: 0733-8716

Communication systems to date primarily aim at reliably communicating bit sequences. Such an approach provides efficient engineering designs that are agnostic to the meanings of the messages or to the goal that the message exchange aims to achieve. Next generation systems, however, can be potentially enriched by folding message semantics and goals of communication into their design. Further, these systems can be made cognizant of the context in which communication exchange takes place, thereby providing avenues for novel design insights. This tutorial summarizes the efforts to date, starting from its early adaptations, semantic-aware and task-oriented communications, covering the foundations, algorithms and potential implementations. The focus is on approaches that utilize information theory to provide the foundations, as well as the significant role of learning in semantics and task-aware communications.

Journal article

Hu Z, Liu G, Xie Q, Xue J, Meng D, Gündüz Det al., 2023, A learnable optimization and regularization approach to massive MIMO CSI feedback, IEEE Transactions on Wireless Communications, Pages: 1-1, ISSN: 1536-1276

Channel state information (CSI) plays a critical role in achieving the potential benefits of massive multiple input multiple output (MIMO) systems. In frequency division duplex (FDD) massive MIMO systems, the base station (BS) relies on sustained and accurate CSI feedback from users. However, due to the large number of antennas and users being served in massive MIMO systems, feedback overhead can become a bottleneck. In this paper, we propose a model-driven deep learning method for CSI feedback, called learnable optimization and regularization algorithm (LORA). Instead of using l 1 -norm as the regularization term, LORA introduces a learnable regularization module that adapts to characteristics of CSI automatically. The conventional Iterative Shrinkage-Thresholding Algorithm (ISTA) is unfolded into a neural network, which can learn both the optimization process and the regularization term by end-to-end training. We show that LORA improves the CSI feedback accuracy and speed. Besides, a novel learnable quantization method and the corresponding training scheme are proposed, and it is shown that LORA can operate successfully at different bit rates, providing flexibility in terms of the CSI feedback overhead. Various realistic scenarios are considered to demonstrate the effectiveness and robustness of LORA through numerical simulations.

Journal article

Erdemir E, Dragotti PL, Gündüz D, 2023, Active privacy-utility trade-off against inference in time-series data sharing, IEEE Journal on Selected Areas in Information Theory, Vol: 4, Pages: 159-173, ISSN: 2641-8770

Internet of things devices have become highly popular thanks to the services they offer. However, they also raise privacy concerns since they share fine-grained time-series user data with untrusted third parties. We model the users personal information as the secret variable, to be kept private from an honest-but-curious service provider, and the useful variable, to be disclosed for utility. We consider an active learning framework, where one out of a finite set of measurement mechanisms is chosen at each time step, each revealing some information about the underlying secret and useful variables, albeit with different statistics. The measurements are taken such that the correct value of useful variable can be detected quickly, while the confidence on the secret variable remains below a predefined level. For privacy measure, we consider both the probability of correctly detecting the secret variable value and the mutual information between the secret and released data. We formulate both problems as partially observable Markov decision processes, and numerically solve by advantage actor-critic deep reinforcement learning. We evaluate the privacy-utility trade-off of the proposed policies on both the synthetic and real-world time-series datasets.

Journal article

Lo WF, Mital N, Wu H, Gunduz Det al., 2023, Collaborative semantic communication for edge inference, IEEE Wireless Communications Letters, ISSN: 2162-2337

We study the collaborative image retrieval problem at the wireless edge, where multiple edge devices capture images of the same object from different angles and locations, which are then used jointly to retrieve similar images at the edge server over a shared multiple access channel (MAC). We propose two novel deep learning-based joint source and channel coding (JSCC) schemes for the task over both additive white Gaussian noise (AWGN) and Rayleigh slow fading channels, with the aim of maximizing the retrieval accuracy under a total bandwidth constraint. The proposed schemes are evaluated on a wide range of channel signal-to-noise ratios (SNRs), and shown to outperform the single-device JSCC and the separation-based multiple-access benchmarks. We also propose a channel state information-aware JSCC scheme with attention modules to enable our method to adapt to varying channel conditions.

Journal article

Zheng J, Ni W, Tian H, Gündüz D, Quek TQS, Han Zet al., 2023, Semi-federated learning: convergence analysis and optimization of a hybrid learning framework, IEEE Transactions on Wireless Communications, Pages: 1-19, ISSN: 1536-1276

Under the organization of the base station (BS), wireless federated learning (FL) enables collaborative model training among multiple devices. However, the BS is merely responsible for aggregating local updates during the training process, which incurs a waste of the computational resource at the BS. To tackle this issue, we propose a semi-federated learning (SemiFL) paradigm to leverage the computing capabilities of both the BS and devices for a hybrid implementation of centralized learning (CL) and FL. Specifically, each device sends both local gradients and data samples to the BS for training a shared global model. To improve communication efficiency over the same time-frequency resources, we integrate over-the-air computation for aggregation and non-orthogonal multiple access for transmission by designing a novel transceiver structure. To gain deep insights, we conduct convergence analysis by deriving a closed-form optimality gap for SemiFL and extend the result to two extra cases. In the first case, the BS uses all accumulated data samples to calculate the CL gradient, while a decreasing learning rate is adopted in the second case. Our analytical results capture the destructive effect of wireless communication and show that both FL and CL are special cases of SemiFL. Then, we formulate a non-convex problem to reduce the optimality gap by jointly optimizing the transmit power and receive beamformers. Accordingly, we propose a two-stage algorithm to solve this intractable problem, in which we provide the closed-form solutions to the beamformers. Extensive simulation results on two real-world datasets corroborate our theoretical analysis, and show that the proposed SemiFL outperforms conventional FL and achieves 3.2% accuracy gain on the MNIST dataset compared to state-of-the-art benchmarks.

Journal article

Shiri I, Sadr AV, Akhavan A, Salimi Y, Sanaat A, Amini M, Razeghi B, Saberi A, Arabi H, Ferdowsi S, Voloshynovskiy S, Gunduz D, Rahmim A, Zaidi Het al., 2022, Decentralized collaborative multi-institutional PET attenuation and scatter correction using federated deep learning, European Journal of Nuclear Medicine and Molecular Imaging, Vol: 50, Pages: 1034-1050, ISSN: 0340-6997

Purpose:Attenuation correction and scatter compensation (AC/SC) are two main steps toward quantitative PET imaging, which remain challenging in PET-only and PET/MRI systems. These can be effectively tackled via deep learning (DL) methods. However, trustworthy, and generalizable DL models commonly require well-curated, heterogeneous, and large datasets from multiple clinical centers. At the same time, owing to legal/ethical issues and privacy concerns, forming a large collective, centralized dataset poses significant challenges. In this work, we aimed to develop a DL-based model in a multicenter setting without direct sharing of data using federated learning (FL) for AC/SC of PET images.Methods:Non-attenuation/scatter corrected and CT-based attenuation/scatter corrected (CT-ASC) 18F-FDG PET images of 300 patients were enrolled in this study. The dataset consisted of 6 different centers, each with 50 patients, with scanner, image acquisition, and reconstruction protocols varying across the centers. CT-based ASC PET images served as the standard reference. All images were reviewed to include high-quality and artifact-free PET images. Both corrected and uncorrected PET images were converted to standardized uptake values (SUVs). We used a modified nested U-Net utilizing residual U-block in a U-shape architecture. We evaluated two FL models, namely sequential (FL-SQ) and parallel (FL-PL) and compared their performance with the baseline centralized (CZ) learning model wherein the data were pooled to one server, as well as center-based (CB) models where for each center the model was built and evaluated separately. Data from each center were divided to contribute to training (30 patients), validation (10 patients), and test sets (10 patients). Final evaluations and reports were performed on 60 patients (10 patients from each center).Results:In terms of percent SUV absolute relative error (ARE%), both FL-SQ (CI:12.21–14.81%) and FL-PL (CI:11.82–13.84%) models demo

Journal article

Pase F, Gündüz D, Zorzi M, 2022, Rate-Constrained Remote Contextual Bandits, IEEE Journal on Selected Areas in Information Theory, Vol: 3, Pages: 789-802

Journal article

Tung T-Y, Kurka DB, Jankowski M, Gündüz Det al., 2022, DeepJSCC-Q: Constellation Constrained Deep Joint Source-Channel Coding, IEEE Journal on Selected Areas in Information Theory, Vol: 3, Pages: 720-731

Journal article

Wu H, Shao Y, Mikolajczyk K, Gunduz Det al., 2022, Channel-adaptive wireless image transmission with OFDM, IEEE Wireless Communications Letters, Vol: 11, Pages: 2400-2404, ISSN: 2162-2337

We present a learning-based channel-adaptive joint source and channel coding (CA-JSCC) scheme for wireless image transmission over multipath fading channels. The proposed method is an end-to-end autoencoder architecture with a dual-attention mechanism employing orthogonal frequency division multiplexing (OFDM) transmission. Unlike the previous works, our approach is adaptive to channel-gain and noise-power variations by exploiting the estimated channel state information (CSI). Specifically, with the proposed dual-attention mechanism, our model can learn to map the features and allocate transmission-power resources judiciously to the available subchannels based on the estimated CSI. Extensive numerical experiments verify that CA-JSCC achieves state-of-the-art performance among existing JSCC schemes. In addition, CA-JSCC is robust to varying channel conditions and can better exploit the limited channel resources by transmitting critical features over better subchannels.

Journal article

Li GY, Saad W, Ozgur A, Kairouz P, Qin Z, Hoydis J, Han Z, Gunduz D, Elmirghani Jet al., 2022, Series Editorial The Sixth Issue of the Series on Machine Learning in Communications and Networks, IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, Vol: 40, Pages: 2507-2509, ISSN: 0733-8716

Journal article

Yang Z, Xia J-Y, Luo J, Zhang S, Gunduz Det al., 2022, A learning-aided flexible gradient descent approach to MISO beamforming, IEEE Wireless Communications Letters, Vol: 11, Pages: 1895-1899, ISSN: 2162-2337

This letter proposes a learning aided gradient descent (LAGD) algorithm to solve the weighted sum rate (WSR) maximization problem for multiple-input single-output (MISO) beamforming. The proposed LAGD algorithm directly optimizes the transmit precoder through implicit gradient descent based iterations, at each of which the optimization strategy is determined by a neural network, and thus, is dynamic and adaptive. At each instance of the problem, this network is initialized randomly, and updated throughout the iterative solution process. Therefore, the LAGD algorithm can be implemented at any signal-to-noise ratio (SNR) and for arbitrary antenna/user numbers, does not require labelled data or training prior to deployment. Numerical results show that the LAGD algorithm can outperform of the well-known WMMSE algorithm as well as other learning-based solutions with a modest computational complexity. Our code is available at https://github.com/XiaGroup/LAGD .

Journal article

Tung T-Y, Gunduz D, 2022, DeepWiVe: deep-learning-aided wireless video transmission, IEEE Journal on Selected Areas in Communications, Vol: 40, Pages: 2570-2583, ISSN: 0733-8716

We present DeepWiVe , the first-ever end-to-end joint source-channel coding (JSCC) video transmission scheme that leverages the power of deep neural networks (DNNs) to directly map video signals to channel symbols, combining video compression, channel coding, and modulation steps into a single neural transform. Our DNN decoder predicts residuals without distortion feedback, which improves the video quality by accounting for occlusion/disocclusion and camera movements. We simultaneously train different bandwidth allocation networks for the frames to allow variable bandwidth transmission. Then, we train a bandwidth allocation network using reinforcement learning (RL) that optimizes the allocation of limited available channel bandwidth among video frames to maximize the overall visual quality. Our results show that DeepWiVe can overcome the cliff-effect , which is prevalent in conventional separation-based digital communication schemes, and achieve graceful degradation with the mismatch between the estimated and actual channel qualities. DeepWiVe outperforms H.264 video compression followed by low-density parity check (LDPC) codes in all channel conditions by up to 0.0485 in terms of the multi-scale structural similarity index measure (MS-SSIM), and H.265+ LDPC by up to 0.0069 on average. We also illustrate the importance of optimizing bandwidth allocation in JSCC video transmission by showing that our optimal bandwidth allocation policy is superior to uniform allocation as well as a heuristic policy benchmark.

Journal article

Ozfatura E, Shao Y, Perotti AG, Popović BM, Gündüz Det al., 2022, All You Need Is Feedback: Communication With Block Attention Feedback Codes, IEEE Journal on Selected Areas in Information Theory, Vol: 3, Pages: 587-602

Journal article

Yilmaz SF, Hasircioglu B, Gunduz D, 2022, Over-the-air ensemble inference with model privacy, 2022 IEEE International Symposium on Information Theory (ISIT), Publisher: IEEE, Pages: 1265-1270

We consider distributed inference at the wireless edge, where multiple clients with an ensemble of models, each trained independently on a local dataset, are queried in parallel to make an accurate decision on a new sample. In addition to maximizing inference accuracy, we also want to maximize the privacy of local models. We exploit the superposition property of the air to implement bandwidth-efficient ensemble inference methods. We introduce different over-the-air ensemble methods and show that these schemes perform significantly better than their orthogonal counterparts, while using less resources and providing privacy guarantees. We also provide experimental results verifying the benefits of the proposed over-the-air inference approach, whose source code is shared publicly on Github.

Conference paper

This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.

Request URL: http://wlsprd.imperial.ac.uk:80/respub/WEB-INF/jsp/search-html.jsp Request URI: /respub/WEB-INF/jsp/search-html.jsp Query String: respub-action=search.html&id=00761710&limit=30&person=true