arXiv subjects

Deniz Gunduz

Publications and source records attributed to Deniz Gunduz.

At least 19 recordsLinked to original sources

arXiv2026

P-WRFGS: Pruning 3D Gaussians for Efficient Wireless Radiance Field Construction

Wireless channel modeling is a key building block for next-generation wireless systems. Predicting the channel state information (CSI) across different transmitter locations can substantially reduce the pilot and feedback overhead of conventional channel estimation. We propose P-WRFGS, an efficient wireless radiance field modeling framework built upon 3D Gaussian splatting. P-WRFGS introduces a learnable mask for each 3D Gaussian primitive to indicate its importance, which guides the pruning of less significant primitives for more efficient rendering. The model is trained using a weighted combination of rendering and regularization losses, allowing a flexible trade-off between rendering quality and efficiency. Numerical results on the $\text{NeRF}^2$ dataset demonstrate that P-WRFGS achieves up to 100$\times$ storage reduction and 7$\times$ rendering speed-up with only mild degradation in SSIM and the achievable rate. Moreover, initializing the Gaussian primitives from a 3D point cloud of the scene further improves the entire quality-efficiency trade-off.

eess.SP↗

arXiv2026

Resolving Multi-Target Association in OFDM-based ISAC via Vision-aided Multi-Modal Learning

Orthogonal frequency division multiplexing (OFDM)-based integrated sensing and communication (ISAC) systems commonly extract target parameters by peak-searching a delay-Doppler map (DDM) constructed from reflected pilots. In multi-target scenarios, this results in ambiguity: the DDM does not reveal which physical target produced which peak, and two targets within the same delay-Doppler resolution cell cannot be separated. We propose a vision-assisted OFDM-ISAC framework that resolves both limitations by fusing wireless and visual modalities. The transmitter encodes an onboard street-view image with deep joint source-channel coding (DeepJSCC) and transmits it over the same OFDM waveform used for sensing; the receiver reconstructs the image, runs a fine-tuned YOLOv5 detector and fuses the resulting per-target features (bounding-box coordinates and class labels) with the DDM and transmitter-receiver geometry through a learned multi-modal network. To stabilize training of the high dimensional delay and Doppler classifiers, we introduce a Kullback Leibler loss against triangular soft labels centered on the ground-truth bin. On a Blender-rendered vehicular testbed, the proposed framework achieves a 16 cm localization root mean square error (RMSE) and a 10.8 ns delay RMSE. An ablation study confirms that removing the visual modality causes a 60x degradation in localization. These results highlight the potential of vision to overcome the data-association and resolution limits of single-modality ISAC.

cs.CV↗

arXiv2026

A Unified Framework for Diffusion Model Unlearning with f-Divergence

Most existing methods for concept unlearning in text-to-image diffusion models minimize a mean squared error (MSE) loss between the denoiser outputs conditioned on a target and an anchor concept, which is implicitly the KL divergence between two Gaussians. We generalize this objective to any $f$-divergence, recovering MSE as the KL instance, and identify a family of $α$-divergences whose Gaussian closed-form yields cheap, MSE-like training objectives. For the remaining $f$-divergences, we provide a min-max objective based on the variational formulation of the $f$-divergence. We theoretically analyze and numerically validate how different $f$-divergences impact the gradient magnitude and the convergence properties of the algorithm, affecting the quality of unlearning. For instance, we observe that the Hellinger closed-form instance consistently dominates MSE across multiple scenarios. More generally, the proposed unified framework offers a flexible paradigm for selecting the optimal divergence based on the application and user goal, allowing for finer control over the trade-off between unlearning efficacy and generative fidelity.

cs.LG↗

arXiv2026

Aggressive or Imperceptible, or Both: Network Pruning Assisted Hybrid Byzantines in Federated Learning

In federated learning (FL), profiling and verifying each client is inherently difficult, which introduces a significant security vulnerability: malicious clients, commonly referred to as Byzantines, can degrade the accuracy of the global model by submitting poisoned updates during training. To mitigate this, the aggregation process at the parameter server must be robust against such adversarial behaviour. Most existing defences approach the Byzantine problem from an outlier detection perspective, treating malicious updates as statistical anomalies and ignoring the internal structure of the trained neural network (NN). Motivated by this, this work highlights the potential of leveraging side information tied to the NN architecture to design stronger, more targeted attacks. In particular, inspired by insights from sparse NNs, we introduce a hybrid sparse Byzantine attack. The attack consists of two coordinated components: (i) A sparse attack component that selectively manipulates parameters with higher sensitivity in the NN, aiming to cause maximum disruption with minimal visibility; (ii) A slow-accumulating attack component that silently poisons parameters over multiple rounds to evade detection. Together, these components create a strong but imperceptible attack strategy that can bypass common defences. We evaluate the proposed attack through extensive simulations and demonstrate its effectiveness against eight state-of-the-art defence mechanisms.

cs.LG↗

arXiv2026

Cache-enabled Generative Joint Source-Channel Coding for Evolving Semantic Communications

Learning-based semantic communication (SemCom) has recently emerged as a promising paradigm for improving the transmission efficiency of wireless networks. However, existing methods typically rely on extensive end-to-end training, which is both inflexible and computationally expensive in dynamic wireless environments. Moreover, they fail to exploit redundancy across multiple transmissions of semantically similar content, limiting overall efficiency. To overcome these limitations, we propose a channel-aware generative adversarial network (GAN) inversion-based joint source-channel coding (CAGI-JSCC) framework that enables training-free SemCom by leveraging a pre-trained SemanticStyleGAN model. By explicitly incorporating wireless channel characteristics into the GAN inversion process, CAGI-JSCC adapts to varying channel conditions without additional training. Furthermore, we introduce a cache-enabled dynamic codebook (CDC) that caches disentangled semantic components at both the transmitter and receiver, allowing the system to reuse previously transmitted content. This semantic-level caching can continuously reduce redundant transmissions as experience accumulates. Extensive experiments on image transmission demonstrate the effectiveness of the proposed framework. In particular, our system achieves comparable perceptual quality with an average bandwidth compression ratio (BCR) of 1/224, and as low as 1/1024 for a single image, significantly outperforming baselines with a BCR of 1/128.

cs.IT↗

arXiv2026

Wireless Physical Neural Networks (WPNNs): Opportunities and Challenges

Wireless communication systems exhibit structural and functional similarities to neural networks: signals propagate through cascaded elements, interact with the environment, and undergo transformations. Building upon this perspective, we introduce a unified paradigm, termed \textit{wireless physical neural networks (WPNNs)}, in which components of a wireless network, such as transceivers, relays, backscatter, and intelligent surfaces, are interpreted as computational layers within a learning architecture. By treating the wireless propagation environment and network elements as differentiable operators, new opportunities arise for joint communication-computation designs, where system optimization can be achieved through learning-based methods applied directly to the physical network. This approach may operate independently of, or in conjunction with, conventional digital neural layers, enabling hybrid communication learning pipelines. In the article, we outline representative architectures that embody this viewpoint and discuss the algorithmic and training considerations required to leverage the wireless medium as a computational resource. Through numerical examples, we highlight the potential performance gains in processing, adaptability, efficiency, and end-to-end optimization, demonstrating the promise of reconfiguring wireless systems as learning networks in next-generation communication frameworks.

eess.SP↗

arXiv2026

Diffusion-aided Extreme Video Compression with Lightweight Semantics Guidance

Modern video codecs and learning-based approaches struggle for semantic reconstruction at extremely low bit-rates due to reliance on low-level spatiotemporal redundancies. Generative models, especially diffusion models, offer a new paradigm for video compression by leveraging high-level semantic understanding and powerful visual synthesis. This paper propose a video compression framework that integrates generative priors to drastically reduce bit-rate while maintaining reconstruction fidelity. Specifically, our method compresses high-level semantic representations of the video, then uses a conditional diffusion model to reconstruct frames from these semantics. To further improve compression, we characterize motion information with global camera trajectories and foreground segmentation: background motion is compactly represented by camera pose parameters while foreground dynamics by sparse segmentation masks. This allows for significantly boosts compression efficiency, enabling descent video reconstruction at extremely low bit-rates.

eess.IV↗

arXiv2025

In-Context Learning for Deep Joint Source-Channel Coding Over MIMO Channels

Large language models have demonstrated the ability to perform \textit{in-context learning} (ICL), whereby the model performs predictions by directly mapping the query and a few examples from the given task to the output variable. In this paper, we study ICL for deep joint source-channel coding (DeepJSCC) in image transmission over multiple-input multiple-output (MIMO) systems, where an ICL denoiser is employed for MIMO symbol estimation. We first study the transceiver without any hardware impairments and explore the integration of transformer-based ICL with DeepJSCC in both open-loop and closed-loop MIMO systems, depending on the availability of channel state information (CSI) at the transceiver. For both open-loop and closed-loop scenarios, we propose two MIMO transceiver architectures that leverage context information, i.e., pilot sequences and their outputs, as additional inputs, enabling the DeepJSCC encoder, DeepJSCC decoder, and the ICL denoiser to jointly learn encoding, decoding, and estimation strategies tailored to each channel realization. Next, we extend our study to a more challenging scenario where the transceiver suffers from in-phase and quadrature (IQ) imbalance, resulting in nonlinear MIMO estimation. In this case, the context information is also exploited, facilitating joint learning across the DeepJSCC encoder, decoder, and the ICL denoiser under hardware impairments and varying channel conditions. Numerical results demonstrate that the ICL denoiser for MIMO estimation significantly outperforms the conventional least-squares method, with even greater advantages under IQ imbalance. Moreover, the proposed transformer-based ICL framework, integrated with contextual information, achieves significant improvements in end-to-end image reconstruction quality under transceiver IQ imbalance.

eess.SP↗

arXiv2025

TransCoder: A Neural-Enhancement Framework for Channel Codes

Reliable communication over noisy channels requires the design of specialized error-correcting codes (ECCs) tailored to specific system requirements. Recently, neural network-based decoders have emerged as promising tools for enhancing ECC reliability, yet their high computational complexity prevents their potential practical deployment. In this paper, we take a different approach and design a neural transmission scheme that employs the transformer architecture in order to improve the reliability of existing ECCs. We call this approach TransCoder, alluding both to its function and architecture. TransCoder operates as a code-adaptive neural module aimed at performance enhancement that can be implemented flexibly at either the transmitter, receiver, or both. The framework employs an iterative decoding procedure, where both noisy information from the channel and updates from the conventional ECC decoder are processed by a neural decoder block, utilizing a block attention mechanism for efficiency. Through extensive simulations with various conventional codes (LDPC, BCH, Polar, and Turbo) and across a wide range of channel conditions, we demonstrate that TransCoder significantly improves block error rate (BLER) performance while maintaining computational complexity comparable to traditional decoders. Notably, our approach is particularly effective for longer codes (block length >64) and at lower code rates, scenarios in which existing neural decoders often struggle (despite their formidable computational complexity). The results establish TransCoder as a promising practical solution for reliable communication among resource-constrained wireless devices.

cs.IT↗

arXiv2025

Implicit Communication in Linear Quadratic Gaussian Control Systems

This paper studies implicit communication in linear quadratic Gaussian control systems. We show that the control system itself can serve as an implicit communication channel, enabling the controller to transmit messages through its inputs to a receiver that observes the system state. This communication is considered implicit because (i) no explicit communication channels are needed; and (ii) information is transmitted while simultaneously fulfilling the controller's primary objective--maintaining the control cost within a specified level. As a result, there exists an inherent trade-off between control and communication performance. This trade-off is formalized through the notion of implicit channel capacity, which characterizes the supremum reliable communication rate subject to a constraint on control performance. We investigate the implicit channel capacity in two settings. When both the controller and the receiver have noiseless observations of the system state, the channel capacity admits a closed-form expression. When both the controller and the receiver have noisy observations, we establish a lower bound on the implicit channel capacity. Surprisingly, in the noiseless observation case, the capacity-achieving input policy adheres to a separation principle, allowing the control and channel coding tasks to be addressed independently, without loss of optimality. While this separation principle no longer holds in the noisy observation setting, we show that linear Gaussian input policies still decouple the channel coding problem from control, and can thus greatly simplify the practical implementation of implicit communication.

cs.IT↗

arXiv2025

LISAC: Learned Coded Waveform Design for ISAC with OFDM

We propose deep learning based coded waveform design for integrated sensing and communication (ISAC) with orthogonal frequency-division multiplexing (OFDM). Our goal is to design a coded waveform capable of delivering accurate target parameter estimation while maintaining high communication quality measured in terms of bit error rate (BER). In the proposed learned coded waveform for ISAC (LISAC), the pilot and data encoding functions at the encoder are parameterized by recurrent neural networks (RNNs) and are trained jointly in an end-to-end fashion. The communication receiver estimates the channel and performs residual-assisted minimum mean square error (MMSE) channel equalization, where a neural network is introduced to calibrate the coarse estimate produced by the standard MMSE channel equalizer. Then, an RNN-based channel decoder is employed to decode the information bits using the equalized signal. Two different sensing loss functions are considered, one calculates the mean square error (MSE) between the original and the estimated sensing parameters, while the other calculates the Cramer-Rao lower bound (CRLB). The LISAC modules are optimized using a weighted combination of communication and sensing losses and different trade-off points between the sensing and communication performances are achieved by adjusting the weights. Simulation results show that the proposed LISAC waveform achieves a better trade-off curve compared to existing alternatives for both AWGN and multi-path fading scenarios. Ablation studies are carried out to demonstrate the gain brought by each design component for a better understanding of the proposed scheme.

eess.SP↗

arXiv2025

Learning to Speak on Behalf of a Group: Medium Access Control for Sending a Shared Message

The rapid development of Internet of Things (IoT) technologies has not only enabled new applications, but also presented new challenges for reliable communication with limited resources. In this work, we define a novel problem that can arise in these scenarios, in which a set of sensors need to communicate a joint observation. This observation is shared by a random subset of the nodes, which need to propagate it to the rest of the network, but coordination is complex: as signaling constraints require the use of random access schemes over shared channels, sensors need to implicitly coordinate, so that at least one transmission gets through without collisions. Unlike the majority of existing medium access schemes, the goal is to make sure that the shared message gets through, regardless of the sender. We analyze this coordination problem theoretically and provide low-complexity solutions. While a clustering-based approach is near-optimal if the sensors have prior knowledge, we provide a distributed multi-armed bandit (MAB) solution for the more general case and validate it by simulation.

cs.NI↗

arXiv2025

Compression Beyond Pixels: Semantic Compression with Multimodal Foundation Models

Recent deep learning-based methods for lossy image compression achieve competitive rate-distortion performance through extensive end-to-end training and advanced architectures. However, emerging applications increasingly prioritize semantic preservation over pixel-level reconstruction and demand robust performance across diverse data distributions and downstream tasks. These challenges call for advanced semantic compression paradigms. Motivated by the zero-shot and representational capabilities of multimodal foundation models, we propose a novel semantic compression method based on the contrastive language-image pretraining (CLIP) model. Rather than compressing images for reconstruction, we propose compressing the CLIP feature embeddings into minimal bits while preserving semantic information across different tasks. Experiments show that our method maintains semantic integrity across benchmark datasets, achieving an average bit rate of approximately 2-3* 10(-3) bits per pixel. This is less than 5% of the bitrate required by mainstream image compression approaches for comparable performance. Remarkably, even under extreme compression, the proposed approach exhibits zero-shot robustness across diverse data distributions and downstream tasks.

cs.CV↗

arXiv2025

A Theory of Semantic Communication

Semantic communication is an emerging research area that has gained a wide range of attention recently. Despite this growing interest, there remains a notable absence of a comprehensive and widely-accepted framework for characterizing semantic communication. This paper introduces a new conceptualization of semantic communication and formulates two fundamental problems, which we term language exploitation and language design. Our contention is that the challenge of language design can be effectively situated within the broader framework of joint source-channel coding theory, underpinned by a comprehensive end-to-end distortion metric. To tackle the language exploitation problem, we put forth three approaches: semantic encoding, semantic decoding, and a synergistic combination of both in the form of combined semantic encoding and decoding. Furthermore, we establish the semantic distortion-cost region as a critical framework for assessing the language exploitation problem. For each of the three proposed approaches, the achievable distortion-cost region is characterized. Overall, this paper aims to shed light on the intricate dynamics of semantic communication, paving the way for a deeper understanding of this evolving field.

cs.IT↗

arXiv2025

Implementing Neural Networks Over-the-Air via Reconfigurable Intelligent Surfaces

In this paper, we investigate reconfigurable intelligent surface (RIS)-aided multiple-input-multiple-output (MIMO) OAC systems designed to emulate the fully-connected (FC) layer of a neural network (NN) via analog OAC, where the RIS and the transceivers are jointly adjusted to engineer the ambient wireless propagation environment to emulate the weights of the target FC layer. We refer to this novel computational paradigm as AirFC. We first study the case in which the precoder, combiner, and RIS phase shift matrices are jointly optimized to minimize the mismatch between the OAC system and the target FC layer. To solve this non-convex optimization problem, we propose a low-complexity alternating optimization algorithm, where semi-closed-form/closed-form solutions for all optimization variables are derived. Next, we consider training of the system parameters using two distinct learning strategies, namely centralized training and distributed training. In the centralized training approach, training is performed at either the transmitter or the receiver, whichever possesses the channel state information (CSI), and the trained parameters are provided to the other terminal. In the distributed training approach, the transmitter and receiver iteratively update their parameters through back and forth transmissions by leveraging channel reciprocity, thereby avoiding CSI acquisition and significantly reducing computational complexity. Subsequently, we extend our analysis to a multi-RIS scenario by exploiting its spatial diversity gain to enhance the system performance. Simulation results show that the AirFC system realized by the RIS-aided MIMO configuration achieves satisfactory classification accuracy.

cs.IT↗

arXiv2025

GINO-Q: Learning an Asymptotically Optimal Index Policy for Restless Multi-armed Bandits

The restless multi-armed bandit (RMAB) framework is a popular model with applications across a wide variety of fields. However, its solution is hindered by the exponentially growing state space (with respect to the number of arms) and the combinatorial action space, making traditional reinforcement learning methods infeasible for large-scale instances. In this paper, we propose GINO-Q, a three-timescale stochastic approximation algorithm designed to learn an asymptotically optimal index policy for RMABs. GINO-Q mitigates the curse of dimensionality by decomposing the RMAB into a series of subproblems, each with the same dimension as a single arm, ensuring that complexity increases linearly with the number of arms. Unlike recently developed Whittle-index-based algorithms, GINO-Q does not require RMABs to be indexable, enhancing its flexibility and applicability. Our experimental results demonstrate that GINO-Q consistently learns near-optimal policies, even for non-indexable RMABs where Whittle-index-based algorithms perform poorly, and it converges significantly faster than existing baselines.

cs.LG↗

arXiv2025

Multimodal LLM Integrated Semantic Communications for 6G Immersive Experiences

6G networks promise revolutionary immersive communication experiences including augmented reality (AR), virtual reality (VR), and holographic communications. These applications demand high-dimensional multimodal data transmission and intelligent data processing in real-time, which is extremely challenging over resource-limited wireless communication systems. Moreover, a joint understanding of the environment, context, and user intent is essential to deliver task-relevant content effectively. This article presents a novel multimodal large language model (MLLM) integrated semantic communications framework, termed MLLM-SC, which fully leverages reasoning and generative capabilities of pre-trained foundation models for context-aware and task-oriented wireless communication. The MLLM-SC framework adopts a device-edge collaborative architecture. At the edge, MLLM-empowered semantic guidance module analyzes multimodal inputs, user intents, and channel conditions to generate importance-aware attention maps prioritizing semantically critical information. An importance-aware semantic encoder and a resource-adaptive semantic decoder are jointly designed and optimized, which can utilize the semantic guidance for adaptive bandwidth allocation and high-quality content reconstruction or generation. Extensive case studies on visual question answering for AR/VR applications and diffusion-driven image generation validate the effectiveness of MLLM-SC.

cs.LG↗

arXiv2025

Over-the-Air Learning-based Geometry Point Cloud Transmission

This paper presents novel solutions for the efficient and reliable transmission of point clouds over wireless channels for real-time applications. We first propose SEmatic Point cloud Transmission (SEPT) for small-scale point clouds, which encodes the point cloud via an iterative downsampling and feature extraction process. At the receiver, SEPT decoder reconstructs the point cloud with latent reconstruction and offset-based upsampling. A novel channel-adaptive module is proposed to allow SEPT to operate effectively over a wide range of channel conditions. Next, we propose OTA-NeRF, a scheme inspired by neural radiance fields. OTA-NeRF performs voxelization to the point cloud input and learns to encode the voxelized point cloud into a neural network. Instead of transmitting the extracted feature vectors as in SEPT, it transmits the learned neural network weights in an analog fashion along with few hyperparameters that are transmitted digitally. At the receiver, the OTA-NeRF decoder reconstructs the original point cloud using the received noisy neural network weights. To further increase the bandwidth efficiency of the OTA-NeRF scheme, a fine-tuning algorithm is developed, where only a fraction of the neural network weights are retrained and transmitted. Noticing the poor generality of the OTA-NeRF schemes, we propose an alternative approach, termed OTA-MetaNeRF, which encodes different input point clouds into the latent vectors with shared neural network weights. Extensive numerical experiments confirm that the proposed SEPT, OTA-NeRF and OTA-MetaNeRF schemes achieve superior or comparable performance over the conventional approaches, where an octree-based or a learning-based point cloud compression scheme is concatenated with a channel code. Finally, the run-time complexities are evaluated to verify the capability of the proposed schemes for real-time communications.

eess.SP↗