arXiv subjects

Shashi Raj Pandey

Publications and source records attributed to Shashi Raj Pandey.

At least 19 recordsLinked to original sources

arXiv2026

Energy-Latency Trade-offs in O-RAN with Distributed Baseband Processing and AI Inference

The Open Radio Access Network (O-RAN) architecture introduces flexible functional splits and open interfaces that enable distributed and centralized deployment of baseband processing. While this flexibility offers opportunities for improved resource utilization, it also introduces fundamental trade-offs between energy efficiency and latency. In this paper, we develop a throughput-based end-to-end energy consumption model for O-RAN and extend it by incorporating detailed latency modeling and application-specific Artificial Intelligence/Machine Learning inference costs. The proposed end-to-end modeling framework provides a general representation of processing, transport, and inference-related energy and delay across the access, metro, and long-haul network segments. Building on this general model, we formulate an optimization problem that selects the placement of baseband processing and AI inference tasks across candidate O-RAN configurations to analyze energy-latency tradeoffs under network load, server frequency, and energy-budget constraints. Using representative hardware platforms and realistic traffic assumptions, we evaluate multiple baseband processing placements corresponding to different O-RAN functional configurations. Our results reveal how user quality of service requirements and network load conditions jointly determine the optimal placement of baseband processing and AI inference tasks, highlighting the inherent trade-off between energy efficiency and latency. The analysis provides practical insights for latency-aware and energy-efficient O-RAN deployments supporting emerging AI-driven services.

cs.NI↗

arXiv2026

Quantum Compression for Distributed Entanglement

We study compression strategies for multipartite entanglement distribution under uncertainty in the partitioning of the quantum state. When the partition is not known at the time of state preparation, we show that a joint design of the resource state and a family of compression schemes can increase the entanglement across partitions under a fixed transmission budget. We formulate this as a source coding problem and derive non-asymptotic upper and lower bounds on the achievable average entanglement subject to an average coding rate. We furthermore design an efficient method for jointly optimizing states and lossless compression maps by exploiting the inherent symmetry of weighted Dicke states. In the bipartite case, we propose practical constructions that closely approach the derived upper bound, and more generally we provide practical constructions for multipartite settings.

quant-ph↗

arXiv2026

Dual-Mode Wireless Devices for Adaptive Pull and Push-Based Communication

This paper introduces a dual-mode communication framework for wireless devices that integrates query-driven (pull) and event-driven (push) transmissions within a unified time-frame structure. Devices typically respond to information requests in pull mode, but if an anomaly is detected, they preempt the regular response to report the critical condition. Additionally, push-based communication is used to proactively send critical data without waiting for a request. This adaptive approach ensures timely, context-aware, and efficient data delivery across different network conditions. To achieve high energy efficiency, we incorporate a wake-up radio mechanism and we design a tailored medium access control (MAC) protocol that supports data traffic belonging to the different communication classes. A comprehensive system-level analysis is conducted, accounting for the wake-up control operation and evaluating three key performance metrics: the success probability of anomaly reports (push traffic), the success probability of query responses (pull traffic) and the total energy consumption. Numerical results characterize the system's behavior and highlight the inherent trade-off between push and pull success probabilities as a function of allocated communication resources. Our analysis demonstrates that the proposed approach achieves up to a 42% reduction in energy consumption per served packet compared to traditional approaches, while maintaining reliable support for both communication paradigms.

cs.NI↗

arXiv2026

Goal-Oriented Access Optimization for ISAC-Enabled Digital Twins

Digital twins (DTs) of physical systems enable real-time remote tracking, control, and learning, but require to be updated with environmental sensory data to maintain alignment with their physical counterparts. In a network context, integrated sensing and communication (ISAC) capabilities can expand the DT's environmental awareness by linking received updates to the location where wireless sensors acquired them. Integrating localization services, however, increases the complexity of the communication system, and can only be supported through smart access optimization. To tackle this problem, we design a two-step goal-oriented approach: firstly, sensors with a high Value of Information (VoI) inform the network of their resource demands through a push-based random access; then, pull-based scheduled transmissions of the actual sensory data are optimized to satisfy ISAC performance constraints. This design allows to maximize the VoI of the information delivered to the DT while locating the transmitting nodes, significantly outperforming existing schemes.

eess.SP↗

arXiv2026

Multi-Agent Conformal Prediction with Personalized Statistical Validity

Uncertainty quantification is essential in high-stakes machine learning tasks. However, one of the principled solutions, conformal prediction, faces challenges under limited local calibration data, privacy constraints, and data heterogeneity. In multi-agent settings, existing works do not simultaneously and satisfactorily address these challenges with guarantees either limited to averages across agents or losing validity in heterogeneous settings. Hence, we propose personalized federated weighted conformal prediction (PFWCP), a framework that combines local density ratio weighting with weighted quantile aggregation to correct for heterogeneity while preserving privacy. The method yields asymptotically valid marginal and calibration-conditional coverage guarantees for each participating agent and supports protocols with one-shot communication. Theoretical analysis presents an adjustment to the coverage variance, governed by an effective sample size expression, which is necessary in the context of weighted conformal prediction, and experiments on synthetic and real datasets show improved calibration quality over state-of-the-art federated conformal baselines.

cs.LG↗

arXiv2026

Energy-Efficient Federated Learning with Relay-Assisted Aggregation in IIoT Networks

This paper presents an energy-efficient transmission framework for federated learning (FL) in industrial Internet of Things (IIoT) environments with strict latency and energy constraints. Machinery subnetworks (SNs) collaboratively train a global model by uploading local updates to an edge server (ES), either directly or via neighboring SNs acting as decode-and-forward relays. To enhance communication efficiency, relays perform partial aggregation before forwarding the models to the ES, significantly reducing overhead and training latency. We analyze the convergence behavior of this relay-assisted FL scheme. To address the inherent energy efficiency (EE) challenges, we decompose the original non-convex optimization problem into sub-problems addressing computation and communication energy separately. An SN grouping algorithm categorizes devices into single-hop and two-hop transmitters based on latency minimization, followed by a relay selection mechanism. To improve FL reliability, we further maximize the number of SNs that meet the roundwise delay constraint, promoting broader participation and improved convergence stability under practical IIoT data distributions. Transmit power levels are then optimized to maximize EE, and a sequential parametric convex approximation (SPCA) method is proposed for joint configuration of system parameters. We further extend the EE formulation to the imperfect channel state information (ICSI). Simulation results demonstrate that the proposed framework significantly enhances convergence speed, reduces outage probability from 10-2 in single-hop to 10-6 and achieves substantial energy savings, with the SPCA approach reducing energy consumption by at least 2x compared to unaggregated cooperation and up to 6x over single-hop transmission.

eess.SP↗

arXiv2026

Energy-Efficient Federated Learning in Cooperative Communication within Factory Subnetworks

This paper investigates energy-efficient transmission protocols in relay-assisted federated learning (FL) setup within industrial subnetworks, considering latency and power constraints. In the subnetworks, devices collaborate to train a global model by transmitting their local models at the edge-enabled primary access (pAP) directly or via secondary access points (sAPs), which act as relays to optimize the training latency. We begin by formulating the energy efficiency problem for our proposed transmission protocol. Given its non-convex nature, we decompose it to minimize computational and transmission energy separately. First, we introduce an algorithm that categorizes devices into single-hop and two-hop groups to decrease transmission energy and then selects associated sAPs. Subsequently, we optimize the transmit power, aiming to maximize energy efficiency. To that end, we propose a Sequential Parametric Convex Approximation (SPCA) method to configure system parameters jointly. Numerical results demonstrate a significant reduction in outage probability and at least a twofold savings in total energy consumption, together with faster convergence, compared with single-hop transmission.

eess.SP↗

arXiv2026

Medium Access for Push-Pull Data Transmission in 6G Wireless Systems

Medium access in 5G systems was tailored to accommodate diverse traffic classes through network resource slicing. 6G wireless systems are expected to be significantly reliant on Artificial Intelligence (AI), leading to data-driven and goal-oriented communication. This leads to augmentation of the design space for Medium Access Control (MAC) protocols, which is the focus of this article. We introduce a taxonomy based on push-based and pull-based communication, which is useful to categorize both the legacy and the AI-driven access schemes. We provide MAC protocol design guidelines for pull- and push-based communication in terms of goal-oriented criteria, such as timing and data relevance. We articulate a framework for co-existence between pull and push-based communications in 6G systems, combining their advantages. We highlight the design principles and main tradeoffs, as well as the architectural considerations for integrating these designs in Open-Radio Access Network (O-RAN) and 6G systems.

cs.NI↗

arXiv2026

Online Continual Learning for Anomaly Detection in IoT under Data Distribution Shifts

In this work, we present OCLADS, a novel communication framework with continual learning (CL) for Internet of Things (IoT) anomaly detection (AD) when operating in non-stationary environments. As the statistical properties of the observed data change with time, the on-device inference model becomes obsolete, which necessitates strategic model updating. OCLADS keeps track of data distribution shifts to timely update the on-device IoT AD model. To do so, OCLADS introduces two mechanisms during the interaction between the resource-constrained IoT device and an edge server (ES): i) an intelligent sample selection mechanism at the device for data transmission, and ii) a distribution-shift detection mechanism at the ES for model updating. Experimental results with TinyML demonstrate that our proposed framework achieves high inference accuracy while realizing a significantly smaller number of model updates compared to the baseline schemes.

cs.LG↗

arXiv2026

Type-Based Unsourced Federated Learning With Client Self-Selection

We address the client-selection problem in federated learning over wireless networks under data heterogeneity. Existing client-selection methods often rely on server-side knowledge of client-specific information, thus compromising privacy. To overcome this issue, we propose a client self-selection strategy based solely on the comparison between locally computed training losses and a centrally updated selection threshold. Furthermore, to support robust aggregation of clients' updates over wireless channels, we integrate this client self-selection strategy into the recently proposed type-based unsourced multiple-access framework over distributed multiple-input multiple-output (D-MIMO) networks. The resulting scheme is completely unsourced: the server does not need to know the identity of the clients. Moreover, no channel state information is required, neither at the clients nor at the server side. Simulation results conducted over a D-MIMO wireless network show that the proposed self-selection strategy matches the performance of a comparable state-of-the-art server-side selection method and consistently outperforms random client selection.

cs.IT↗

arXiv2026

Reliable Explanations or Random Noise? A Reliability Metric for XAI

In recent years, explaining decisions made by complex machine learning models has become essential in high-stakes domains such as energy systems, healthcare, finance, and autonomous systems. However, the reliability of these explanations, namely, whether they remain stable and consistent under realistic, non-adversarial changes, remains largely unmeasured. Widely used methods such as SHAP and Integrated Gradients (IG) are well-motivated by axiomatic notions of attribution, yet their explanations can vary substantially even under system-level conditions, including small input perturbations, correlated representations, and minor model updates. Such variability undermines explanation reliability, as reliable explanations should remain consistent across equivalent input representations and small, performance-preserving model changes. We introduce the Explanation Reliability Index (ERI), a family of metrics that quantifies explanation stability under four reliability axioms: robustness to small input perturbations, consistency under feature redundancy, smoothness across model evolution, and resilience to mild distributional shifts. For each axiom, we derive formal guarantees, including Lipschitz-type bounds and temporal stability results. We further propose ERI-T, a dedicated measure of temporal reliability for sequential models, and introduce ERI-Bench, a benchmark designed to systematically stress-test explanation reliability across synthetic and real-world datasets. Experimental results reveal widespread reliability failures in popular explanation methods, showing that explanations can be unstable under realistic deployment conditions. By exposing and quantifying these instabilities, ERI enables principled assessment of explanation reliability and supports more trustworthy explainable AI (XAI) systems.

cs.LG↗

arXiv2026

Context Dependence and Reliability in Autoregressive Language Models

Large language models (LLMs) generate outputs by utilizing extensive context, which often includes redundant information from prompts, retrieved passages, and interaction history. In critical applications, it is vital to identify which context elements actually influence the output, as standard explanation methods struggle with redundancy and overlapping context. Minor changes in input can lead to unpredictable shifts in attribution scores, undermining interpretability and raising concerns about risks like prompt injection. This work addresses the challenge of distinguishing essential context elements from correlated ones. We introduce RISE (Redundancy-Insensitive Scoring of Explanation), a method that quantifies the unique influence of each input relative to others, minimizing the impact of redundancies and providing clearer, stable attributions. Experiments demonstrate that RISE offers more robust explanations than traditional methods, emphasizing the importance of conditional information for trustworthy LLM explanations and monitoring.

cs.CL↗

arXiv2026

1Q: First-Generation Wireless Systems Integrating Classical and Quantum Communication

We introduce the concept of 1Q, the first wireless generation of integrated classical and quantum communication. 1Q features quantum base stations (QBSs) that support entanglement distribution via free-space optical links alongside traditional radio communications. Key new components include quantum cells, quantum user equipment (QUEs), and hybrid resource allocation spanning classical time-frequency and quantum entanglement domains. Several application scenarios are discussed and illustrated through system design requirements for quantum key distribution, blind quantum computing, and distributed quantum sensing. A range of unique quantum constraints are identified, including decoherence timing, fidelity requirements, and the interplay between quantum and classical error probabilities. Protocol adaptations extend cellular connection management to incorporate entanglement generation, distribution, and handover procedures, expanding the Quantum Internet to the cellular wireless.

cs.NI↗

arXiv2026

Explainability of Complex AI Models with Correlation Impact Ratio

Complex AI systems make better predictions but often lack transparency, limiting trustworthiness, interpretability, and safe deployment. Common post hoc AI explainers, such as LIME, SHAP, HSIC, and SAGE, are model agnostic but are too restricted in one significant regard: they tend to misrank correlated features and require costly perturbations, which do not scale to high dimensional data. We introduce ExCIR (Explainability through Correlation Impact Ratio), a theoretically grounded, simple, and reliable metric for explaining the contribution of input features to model outputs, which remains stable and consistent under noise and sampling variations. We demonstrate that ExCIR captures dependencies arising from correlated features through a lightweight single pass formulation. Experimental evaluations on diverse datasets, including EEG, synthetic vehicular data, Digits, and Cats-Dogs, validate the effectiveness and stability of ExCIR across domains, achieving more interpretable feature explanations than existing methods while remaining computationally efficient. To this end, we further extend ExCIR with an information theoretic foundation that unifies the correlation ratio with Canonical Correlation Analysis under mutual information bounds, enabling multi output and class conditioned explainability at scale.

cs.LG↗

arXiv2026

Experimental Study of Low-Latency Video Streaming in an ORAN Setup with Generative AI

Current Adaptive Bit Rate (ABR) methods react to network congestion after it occurs, causing application layer buffering and latency spikes in live video streaming. We introduce a proactive semantic control channel that enables coordination between Open Radio Access Network (ORAN) xApp, Mobile Edge computing (MEC), and User Equipment (UE) components for seamless live video streaming between mobile devices. When the transmitting UE experiences poor Uplink (UL) conditions, the MEC proactively instructs downscaling based on low-level RAN metrics, including channel SNR updated every millisecond, preventing buffering before it occurs. A Generative AI (GAI) module at the MEC reconstructs high-quality frames from downscaled video before forwarding to the receiving UE via the typically more robust Downlink (DL). Experimental validation on a live ORAN testbed with 50 video streams shows that our approach reduces latency tail behavior while achieving up to 4 dB improvement in PSNR and 15 points in VMAF compared to reactive ABR methods. The proactive control eliminates latency spikes exceeding 600 ms, demonstrating effective cross-layer coordination for latency-critical live video streaming.

cs.NI↗

arXiv2025

Link-Aware Energy-Frugal Continual Learning for Fault Detection in IoT Networks

The use of lightweight machine learning (ML) models in internet of things (IoT) networks enables resource constrained IoT devices to perform on-device inference for several critical applications. However, the inference accuracy deteriorates due to the non-stationarity in the IoT environment and limited initial training data. To counteract this, the deployed models can be updated occasionally with new observed data samples. However, this approach consumes additional energy, which is undesirable for energy constrained IoT devices. This letter introduces an event-driven communication framework that strategically integrates continual learning (CL) in IoT networks for energy-efficient fault detection. Our framework enables the IoT device and the edge server (ES) to collaboratively update the lightweight ML model by adapting to the wireless link conditions for communication and the available energy budget. Evaluation on real-world datasets show that the proposed approach can outperform both periodic sampling and non-adaptive CL in terms of inference recall; our proposed approach achieves up to a 42.8% improvement, even under tight energy and bandwidth constraint.

cs.LG↗

arXiv2025

ISAC-Powered Distributed Matching and Resource Allocation in Multi-band NTN

Scalability is a major challenge in non-geostationary orbit (NGSO) satellite networks due to the massive number of ground users sharing the limited sub-6 GHz spectrum. Using K- and higher bands is a promising alternative to increase the accessible bandwidth, but these bands are subject to significant atmospheric attenuation, notably rainfall, which can lead to degraded performance and link outages. We present an integrated sensing and communications (ISAC)-powered framework for resilient and efficient operation of multi-band satellite networks. It is based on distributed mechanisms for atmospheric sensing, cell-to-satellite matching, and resource allocation (RA) in a 5G Non-Terrestrial Network (NTN) wide-area scenario with quasi-Earth fixed cells and a beam hopping mechanism. Results with a multi-layer multi-band constellation with satellites operating in the S- and K-bands demonstrate the benefits of our framework for ISAC-powered multi-band systems, which achieves 73% higher throughput per user when compared to single S- and K-band systems.

cs.NI↗

arXiv2025

Real-Time Inference for Distributed Multimodal Systems under Communication Delay Uncertainty

Connected cyber-physical systems perform inference based on real-time inputs from multiple data streams. Uncertain communication delays across data streams challenge the temporal flow of the inference process. State-of-the-art (SotA) non-blocking inference methods rely on a reference-modality paradigm, requiring one modality input to be fully received before processing, while depending on costly offline profiling. We propose a novel, neuro-inspired non-blocking inference paradigm that primarily employs adaptive temporal windows of integration (TWIs) to dynamically adjust to stochastic delay patterns across heterogeneous streams while relaxing the reference-modality requirement. Our communication-delay-aware framework achieves robust real-time inference with finer-grained control over the accuracy-latency tradeoff. Experiments on the audio-visual event localization (AVEL) task demonstrate superior adaptability to network dynamics compared to SotA approaches.

cs.LG↗