The Pinching Antenna System (PASS) has recently attracted a lot of attention due to its capability to reconfigure antenna positions along a waveguide dynamically. However, the resulting phase shifts introduce non-convex and highly coupled optimization problems, making closed-form solutions intractable. To overcome this, we propose a Deep Reinforcement Learning (DRL) based Two-Staged Soft Actor-Critic framework (TS-SAC). In the first stage, a Soft Actor-Critic (SAC) agent optimizes the spatial placement of antennas, while the second stage computes the optimal Non-Orthogonal Multiple Access (NOMA) power allocation coefficient using a constrained nonlinear optimization solver. This two-staged approach is evaluated in a two-user downlink NOMA scenario and benchmarked against state-of-theart DRL baselines. Our main aim is to maximize the sum rate of this system. Simulation results show that TS-SAC achieves a 23.7% improvement in average sum rate over Deep Deterministic Policy Gradient (DDPG) at 20 dBm, 11.7% improvement over Twin-Delayed Deep Deterministic Policy Gradient (TD3) and 9% improvement over SAC at the same power level. These results establish the effectiveness and reliability of the proposed approach for next-generation wireless networks.
In this work, we enhance cooperative sensing capabilities in multi Unmanned Aerial Vehicles (UAVs) systems, with each UAV mounted with a directional antenna to detect multiple targets cooperatively. To improve detection accuracy in dynamic environments with noise fluctuations, we adopt an eigenvalue-based detection technique. We maximize total detection probability by simultaneously optimizing the spatial deployment of UAVs and their antenna orientations. To address the inherent non-convexity of this problem, we propose an iterative Particle Swarm Optimization (PSO)-based approach with a penalty method for its fitness function. The proposed approach effectively navigates complex search spaces, managing spatial and antenna constraints to guide solutions toward global optima. Monte Carlo simulations demonstrate that our PSO-based algorithm outperforms current techniques, achieving superior detection performance and robust sensing capabilities. Notably, the proposed scheme achieves a 7.94% improvement with 2 UAVs and a 7.21 % improvement with 3 UAVs over the alternating direction penalty method (ADPM)-based scheme, highlighting its effectiveness even with fewer UAVs deployed for cooperative sensing.
Federated learning has emerged as an effective approach for building a unified machine learning model by training across multiple edge devices without sharing data, thus ensuring privacy. However, a large number of participating devices intensifies competition for limited network resources, making it challenging to include all devices in the training process. Selecting an optimal set of devices while managing resources is essential for maintaining the framework's resilience. In this context, this paper focuses on maximizing device participation while efficiently managing limited uplink transmission power and bandwidth under constrained latency. Through comprehensive mathematical analysis, we propose an optimized device selection and resource management method that achieves the optimal solution with reduced computational time. Experimental results show that the proposed method effectively identifies the optimal set of devices and resources while reducing computational time by a factor of 44.41 compared to existing state-of-the-art methods.
Elastic Optical Networks (EONs) are next-generation technology for growing traffic demands. Routing and spectrum assignment (RSA) algorithms run on the software-defined networking (SDN) controller and help in the efficient resource provisioning for the requested demands. The routing algorithm selects the shortest path between the source-destination nodes and the SA algorithm assigns the requested frequency slots (FSs) in the path links by satisfying spectrum constraints. The routing is of two types, static and dynamic. The static routing selects the fixed shortest path each time between the given source-destination nodes while the dynamic routing is flexible and selects the path depending upon the link status and resource availability between the source-destination node pair. The drawback of the static routing is the overburdening of the link resources because of the selection of the pre-defined fixed routes between source-destination nodes resulting in increased request-blocking probability (RBP). Therefore this paper has proposed a dynamic routing and spectrum allocation with an adaptive link weight (DRSA-ALW) algorithm that modifies the link weights with the arrived traffic and selects the path depending on the link resource availability. The simulation results demonstrate that the DRSA-ALW algorithm provides uniform traffic distribution resulting in efficient resource provisioning and reduced RBP as compared to the benchmark algorithm.
Multiple-Input Multiple-Output (MIMO) mode selection is essential for modern wireless communication systems. Usually channel state information (CSI) feedback is necessary for such operation. In fast-changing channels CSI becomes nearly infeasible due to limited coherence time and large volume of information required. To address this challenge, utilization of the channel spatial correlation matrix (SCM) is an effective solution. SCM is a function of mean angle of arrival (AoA) and variance of AoA. Our work estimates the mean AoA and its variance based on the power angular spectrum (PAS), in accordance with the 3rd Generation Partnership Project (3GPP) TR 38.901. Our approach employs the 3GPP clustered delay line (CDL) channel model, while using Orthogonal Frequency Division Multiplexing (OFDM) based channel estimation as in 4G LTE, 5G NR, and emerging 6G standards. We have achieved mean AoA and variance of AoA estimation accuracy on the order of 0.01, and our proposed AoA estimation method can be used across the microwave and mm Wave frequency range.
We analyze low-power short-range wireless communications through a low-rank fading channel-a bonafide use case in many communication scenarios requiring simple wireless connectivity with much relaxed constraints on throughput and data latency. This is certainly true, for instance, in low-complexity wireless channels in the low-rate wireless personal area networks (LR-WPANs). Two more wireless communication scenarios can be cited as relevant examples: low-rate communication on the control channels in the existing wireless local area networks (WLANs), as well as the low-rank Reconfigurable Intelligent Surface (RIS) assisted millimeter-wave (mmWave) communication link operating at a low transmit power budget. Specifically, we characterize the capacity of a low-rank wireless channel with varying fading severity at low signal-to-noise ratios (SNRs). The rank deficiency is incorporated by introducing pinhole condition in the channel. The channel capacity degradation with fading severity at high SNRs is well known: the distribution of deep fades increases significantly with fading severity resulting in poor performance. Our analysis of the double-fading pinhole channel at low-SNR shows a very counter-intuitive result that-higher fading severity enables higher spectral efficiency at sufficiently low SNRs. The underlying reason is that at low SNRs, ergodic capacity depends crucially on the probability distribution of channel peaks (simply tail distribution); for the pinhole channel, the tail distribution improves with increased fading severity. This allows a transmitter operating at low SNRs to exploit channel peaks efficiently resulting in net improvement in achievable spectral efficiency. We derive a new key result quantifying the above dependence for the double-Nakagami-m fading pinhole channel-that is, the ergodic capacity varies inversely to the product of fading (severity) parameters of the two independent Nakagami-m fadings involved.
This paper introduces an innovative detection strategy for IRS-aided single-user MIMO systems utilizing OTFS modulation, specifically designed to operate efficiently under hardware constraints such as Carrier Frequency Offset (CFO). The proposed approach leverages Maximum Ratio Combining (MRC) to enhance signal quality by effectively mitigating multi-path fading and inter-antenna interference, which are critical challenges in high-mobility environments. A standout feature of this detection strategy is its low computational complexity, making it highly suitable for real-time implementation in dynamic and resource-constrained wireless systems. Through extensive simulations, the performance of IRS-aided MIMO-OTFS systems with MRC detection is shown to consistently outperform traditional detection methods, demonstrating notable improvements in reliability and signal quality. These results underscore the transformative potential of combining IRS and OTFS modulation techniques to propel the development of next-generation wireless communication systems, offering enhanced performance, efficiency, and robustness.
We numerically and experimentally demonstrate an optical parametric amplifier in a partially degenerate configuration using a highly nonlinear fiber. A parametric gain exceeding 7 dB is achieved for a single-polarization signal across the C-band, with the pump wavelength maintained close to the fiber's zero-dispersion wavelength,
Accurate weather forecasting is vital for sectors such as agriculture and disaster management, where timely and precise predictions can significantly impact outcomes. Conventional forecasting methods often struggle in accurately capturing the complex spatio-temporal dependencies inherent in meteorological data. The proliferation of modern sensor-equipped devices, such as smartphones, enables decentralized and real-time data collection, providing opportunities for large-scale predictive modelling. However, the sensitive nature of user data requires privacy-preserving algorithms. In this paper, we propose a federated spatio-temporal adaptive graph neural network (FedSTAGNN) model, which combines attention-based graph convolutional network layers and temporal convolutional layers in a federated learning framework to capture spatial and temporal patterns. The proposed model employs hierarchical aggregation at both city and nationwide levels to ensure scalability across diverse regions. Experimental evaluation on real-world weather dataset demonstrates that FedSTAGNN outperforms state-of-the-art models in terms of predictive accuracy, communication efficiency, and data privacy, highlighting its potential for practical deployment in large-scale, privacy-sensitive applications.
Hyperspectral images (HSIs) are pivotal in remote sensing, providing rich spectral and spatial information for applications such as agriculture, environmental monitoring, and mineral exploration. Unlike traditional RGB images, HSIs capture data across hundreds of contiguous spectral bands, enabling precise material classification based on unique spectral signatures. However, the high dimensionality and the need for pixel-level labeling pose significant challenges, as manual annotation is labor-intensive and time-consuming. To address this, we leverage Domain Adaptation, a transfer learning technique that facilitates knowledge transfer from a well-labeled source domain to a sparsely labeled or unlabeled target domain. In this paper, we focus on few-shot Domain Adaptation for HSI classification, where only a limited number of labeled target samples are available. Our approach integrates Active Learning to iteratively fine-tune a classifier by selecting the most informative target samples, thus reducing domain shift and enhancing model performance. We utilize a framework that generates domain-invariant embeddings by aligning the spectral and spatial features of HSIs from the source and target domains. This method significantly reduces the need for manual labeling while maintaining high classification accuracy. Our experimental results demonstrate the effectiveness of combining few-shot Domain Adaptation with Active Learning, achieving a remarkable 98 % accuracy in hyperspectral image classification with limited labeled data.
In the domain of predictive modeling, achieving high accuracy with reduced computational complexity is imperative, particularly for applications requiring real-time processing or operating under stringent resource constraints. Conventional neural networks, despite their efficacy, often encounter limitations due to high structural complexity and inefficiencies in information propagation. This paper introduces an optimized small world neural network architecture (oSWNN), which systematically enhances predictive accuracy by optimizing the number and strategic placement of new links between neurons. By incorporating the small world property, characterized by short path lengths and efficient connectivity, our model facilitates seamless information flow, boosting performance in regression tasks. Tested on multiple regression datasets, including UCI Compressive Concrete Strength, UCI Energy Efficiency, and Oxford Parkinson's Disease dataset, the proposed oSWNN demonstrates significant gains in both accuracy and computational efficiency, underscoring the potential of small world neural network for robust predictive analysis.
We consider a wireless networked control system scenario consisting of a plant and a remote controller. The plant is modelled as a Markov decision process (MDP). The plant is connected to the remote controller through a wireless access point. At a time, the wireless access point can either transmit state observations from the plant to the controller through an uplink queue or control actions from the controller to the plant through a downlink queue. We consider a resource allocation problem for the uplink and downlink queues. The objective is to maximize control performance measured using average reward for the MDP. We also investigate whether the resource allocation problem can be equivalently solved by minimizing the Age-of-Loop (AoL), which is a recently proposed information freshness metric for closed-loop control applications. We analytically characterize the average reward in terms of the joint distribution of uplink and downlink Age-of-Information for two service disciplines and a special class of MDP models. The analytical characterization is useful for optimal uplink and downlink resource allocation. We also analytically characterize the average AoL. In contrast to prior work, we conclude that resource allocation to minimize the average AoL generally does not lead to maximization of the average reward for remote control of MDPs. We present examples of MDPs for which minimization of average AoL leads to a minimum for the average reward.
This paper presents the implementation of the direction of arrival (DoA) estimation in the processing system of RFSoC for real-time applications. The multiple signal classification (MUSIC) algorithm is a high-resolution and more accurate DoA estimation algorithm. However, the complexity is very high due to complex operations like eigenvalue decomposition and spectrum peak search. Implementing such a complex algorithm in hardware is challenging for real-time applications. The radio frequency system on chip (RFSoC) is suitable for implementing complex algorithms more efficiently in which FPGA, multi-core ARM processors and RF data converters are present in a single chip. Some of the challenges in implementing complex algorithms in RFSoC are achieving minimum resource utilization, high precision, and automatic synchronization of data converters. In this work, the MUSIC algorithm is implemented in an embedded ARM processor of RFSoC, which will not require any resources and provides high accuracy by using floating-point data types. To validate the design, DoA measurement is performed in an anechoic chamber by receiving LTE signals through a 1x4 antenna array from different directions.
Water quality monitoring (WQM) is indispensable as several industries discharge hazardous chemicals and waste directly into the rivers, contaminating the water resources. Numerous existing systems provide WQM but are not energy-efficient solutions as they transmit huge amounts of data on cloud servers at regular intervals. Additionally, incorporating Deep Neural Network (DNN) models on the cloud for data inferencing results in high-cost consumption. Thus, this study provides a dynamic payload-based energy-efficient solution that integrates a lightweight Message Queuing Telemetry Transport (MQTT) protocol for wireless communication between the WQM sensor node and the cloud server. The proposed system is built on a resource-constrained ATMEGA2560-16AU microcontroller platform that provides real-time data and makes a cost-effective solution. Quectel EC200U-CN module provides a 4G connection for data transmission and consumes maximum current during transmission. Hence, the presented architecture establishes a reliable connection and transmits dynamic payload constructing water quality (WQ) parameters to the cloud server whenever a significant change is noticed via the proposed set of sequences of AT commands. The lower root mean squared error (RSME) and higher R-squared value comparison between the sensor node data at the edge and stored data at the cloud server remarks the efficiency of the proposed architecture with reduced energy consumption. Moreover, the system sends alert notifications on the developed web application for quick remedial actions at alarming conditions.
This paper explores the integration of unmanned aerial vehicles (UAV s) with reconfigurable intelligent surfaces (RIS) and relay networks to enhance the performance of the communication systems. We introduce a system model that combines relay and RIS functionalities on a UAV, operating in an integrated relay-RIS mode with selection combining. Our work focuses on critical performance metrics, such as outage probability (OP) and average symbol error rate (ASER), under realistic channel conditions modeled with Nakagami-m fading. We derive analytical expressions for OP and SER, validated through comprehensive simulations, demonstrating substantial improvements in reliability over standalone systems. This re-search provides a robust framework for deploying integrated relay-RIS systems, addressing the challenges of dynamic and complex communication scenarios, and paving the way for transformative applications in smart cities, emergency response, and remote connectivity.
Most Single Image Super-Resolution (SISR) meth-ods rely on paired training data, typically generated through bicubic downsampling, which fails to capture the complex degra-dations seen in real-world scenarios. This reliance on synthetic data creates a performance gap when deploying SISR models on actual degraded images. To address this limitation, we pro-pose a Robust Zero-Shot learning-based Generative Adversarial Network for Super-Resolution (RZSGAN-SR) framework that leverages zero-shot learning to handle unknown degradations effectively. Our approach incorporates a Zero-Shot learning-based Degradation Correction Network (ZSDCN) to translate real-world degraded Low-Resolution (LR) images into synthetic LR images with known degradations. These translated images are then fed into a lightweight, robust Generative Adversarial Network (GAN)-based SR network to generate high-quality, visually realistic Super-Resolved (SR) images. More specifically, the proposed RZSGA-SR is a two-phase framework consisting of zero-shot degradation correction and efficient GAN-based upsampling. This hybrid model leverages the adaptability of Zero-Shot Learning (ZSL) with the realism of a robust GAN-based SR network with high fidelity and perceptual quality SR reconstruction. Extensive experiments show that RZSGAN-SR surpasses state-of-the-art methods, achieving superior reconstruction (PSNR, SSIM) and perceptual quality (LPIPS) on real-world degraded images.
Modeling of heritage monuments in 3D poses a computationally complex task. One approach to reduce complexity lies in modeling architecturally significant parts and integrating those into a whole. To complete that task based on crowdsourced images without annotations and with varying viewpoints and lighting, we consider semantic part segmentation of heritage monuments using the state-of-theart Segment Anything Model (SAM). However, SAM does not produce semantically consistent labels across multiple images. As a remedy, we propose an SfM-guided framework that combines sparse 3D reconstruction from Structure-from-Motion (SfM) with SAM to achieve consistent part-level segmentation of architectural elements. Using 3D spatial context and camera poses, our method aligns segmentations across diverse images, bridging the gap between fine-grained segmentation and semantic consistency. This scalable approach demonstrates the potential to integrate 3D geometry with foundational models for cultural heritage analysis.
Ambient IoT (A-IoT) is a new paradigm of extremely low complex devices that enables devices to operate with extremely low energy consumption, often harvesting energy from the surrounding environment rather than relying on traditional batteries or wired energy sources. Further, the communication from the A-IoT device is based on modulation and backscattering on a signal, a.k.a carrier wave, provided externally. This paper discusses various aspects associated with A-IoT communication system, i.e., capabilities and types of A-IoT devices, the deployment scenarios, modulation schemes, characteristics of the carrier wave and backscattering mechanism. The document further analyses the performance of the A-IoT communication system under various scenarios.
This study conducts a detailed performance analysis of free-space optical (FSO) communication systems in high atmospheric turbulence, where a negative exponential model charac-terizes channel conditions. The research focuses on understanding how inaccuracies in channel estimation critically affect FSO system performance in such turbulent settings. We present novel analytical expressions for the probability density function (PDF) and cumulative distribution function (CDF) of signal-to-noise ratio (SNR) under channel estimation errors. The derived expressions provide insights into key performance metrics such as outage probability, average bit error rate (BER), and ergodic capacity. Through numerical simulations, we demonstrate the extent of performance deterioration due to imperfect channel knowledge as compared to ideal estimates. Finally, Monte Carlo simulations validate our findings, highlighting the essential role of precise channel estimation for dependable FSO performance under severe turbulence conditions.
Orthogonal Time Frequency Space (OTFS) mod-ulation has been proposed to provide better bit error rate (BER) performance over orthogonal frequency division multi-plexing (OFDM) under high-mobility wireless channels. In this paper, a multi-mode index modulation aided OTFS-based spatial modulation (MMIM-OTFS-SM) system is proposed to improve spectral efficiency (SE) and transmission reliability in mobile communication environments. The proposed system exploits the benefits of MMIM, OTFS, and SM. Moreover, we discuss the discrete-input continuous-output memoryless channel (DCMC) capacity of the MMIM-OTFS-SM system. A framework for analyzing the average bit error rate (ABER) performance is also introduced. The analytical expression is validated through simulation results. Finally, our simulation results show that the proposed MMIM-OTFS-SM system performs well over conventional index modulation-based SM-OTFS schemes.
