arXiv subjects

Petar Popovski

Publications and source records attributed to Petar Popovski.

At least 127 records · Page 7Linked to original sources

arXiv2024

Prediction of Rare Channel Conditions using Bayesian Statistics and Extreme Value Theory

Estimating the probability of rare channel conditions is a central challenge in ultra-reliable wireless communication, where random events, such as deep fades, can cause sudden variations in the channel quality. This paper proposes a sample-efficient framework for predicting the statistics of such events by utilizing spatial dependency between channel measurements acquired from various locations. The proposed framework combines radio maps with non-parametric models and extreme value theory (EVT) to estimate rare-event channel statistics under a Bayesian formulation. The framework can be applied to a wide range of problems in wireless communication and is exemplified by rate selection in ultra-reliable communications. Notably, besides simulated data, the proposed framework is also validated with experimental measurements. The results in both cases show that the Bayesian formulation provides significantly better results in terms of throughput compared to baselines that do not leverage measurements from surrounding locations. It is also observed that the models based on EVT are generally more accurate in predicting rare-event statistics than non-parametric models, especially when only a limited number of channel samples are available. Overall, the proposed methods can significantly reduce the number of measurements required to predict rare channel conditions and guarantee reliability.

eess.SP↗

arXiv2024

TinyAirNet: TinyML Model Transmission for Energy-efficient Image Retrieval from IoT Devices

This letter introduces an energy-efficient pull-based data collection framework for Internet of Things (IoT) devices that use Tiny Machine Learning (TinyML) to interpret data queries. A TinyML model is transmitted from the edge server to the IoT devices. The devices employ the model to facilitate the subsequent semantic queries. This reduces the transmission of irrelevant data, but receiving the ML model and its processing at the IoT devices consume additional energy. We consider the specific instance of image retrieval in a single device scenario and investigate the gain brought by the proposed scheme in terms of energy efficiency and retrieval accuracy, while considering the cost of computation and communication, as well as memory constraints. Numerical evaluation shows that, compared to a baseline scheme, the proposed scheme reaches up to 67% energy reduction under the accuracy constraint when many images are stored. Although focused on image retrieval, our analysis is indicative of a broader set of communication scenarios in which the preemptive transmission of an ML model can increase communication efficiency.

cs.NI↗

arXiv2024

Performance Analysis of Uplink Rate-Splitting Multiple Access with Hybrid ARQ

Rate-splitting multiple access (RSMA) has attracted a lot of attention as a general and powerful multiple access scheme. In the uplink, instead of encoding the whole message into one stream, a user can split its message into two parts and encode them into two streams before transmitting a superposition of these two streams. The base station (BS) uses successive interference cancellation (SIC) to decode the streams and reconstruct the original messages. Focusing on the packet transmission reliability, we investigate the features of RSMA in the context of hybrid automatic repeat request (HARQ), a well-established mechanism for enhancing reliability. This work proposes a HARQ scheme for uplink RSMA with different retransmission times for a two-user scenario and introduces a power allocation strategy for the two split streams. The results show that compared with non-orthogonal multiple access (NOMA) and frequency division multiple access (FDMA), RSMA outperforms them in terms of error probability and power consumption. The results show that RSMA with HARQ has the potential to improve the reliability and efficiency of wireless communication systems.

cs.IT↗

arXiv2024

Reliable Quantum Communications based on Asymmetry in Distillation and Coding

The reliable provision of entangled qubits is an essential precondition in a variety of schemes for distributed quantum computing. This is challenged by multiple nuisances, such as errors during the transmission over quantum links, but also due to degradation of the entanglement over time due to decoherence. The latter can be seen as a constraint on the latency of the quantum protocol, which brings the problem of quantum protocol design into the context of latency-reliability constraints. We address the problem through hybrid schemes that combine: (1) indirect transmission based on teleportation and distillation; (2) direct transmission, based on quantum error correction (QEC). The intuition is that, at present, the quantum hardware offers low fidelity, which demands distillation; on the other hand, low latency can be obtained by QEC techniques. It is shown that, in the proposed framework, the distillation protocol gives rise to asymmetries that can be exploited by asymmetric quantum error correcting code (QECC), which sets the basis for unique hybrid distillation and coding design. Our results show that ad-hoc asymmetric codes give, compared to conventional QEC, a performance boost and codeword size reduction both in a single link and in a quantum network scenario.

quant-ph↗

arXiv2024

Batch Selection and Communication for Active Learning with Edge Labeling

Conventional retransmission (ARQ) protocols are designed with the goal of ensuring the correct reception of all the individual transmitter's packets at the receiver. When the transmitter is a learner communicating with a teacher, this goal is at odds with the actual aim of the learner, which is that of eliciting the most relevant label information from the teacher. Taking an active learning perspective, this paper addresses the following key protocol design questions: (i) Active batch selection: Which batch of inputs should be sent to the teacher to acquire the most useful information and thus reduce the number of required communication rounds? (ii) Batch encoding: Can batches of data points be combined to reduce the communication resources required at each communication round? Specifically, this work introduces Communication-Constrained Bayesian Active Knowledge Distillation (CC-BAKD), a novel protocol that integrates Bayesian active learning with compression via a linear mix-up mechanism. Comparisons with existing active learning protocols demonstrate the advantages of the proposed approach.

cs.LG↗

arXiv2024

Continual Deep Reinforcement Learning for Decentralized Satellite Routing

This paper introduces a full solution for decentralized routing in Low Earth Orbit satellite constellations based on continual Deep Reinforcement Learning (DRL). This requires addressing multiple challenges, including the partial knowledge at the satellites and their continuous movement, and the time-varying sources of uncertainty in the system, such as traffic, communication links, or communication buffers. We follow a multi-agent approach, where each satellite acts as an independent decision-making agent, while acquiring a limited knowledge of the environment based on the feedback received from the nearby agents. The solution is divided into two phases. First, an offline learning phase relies on decentralized decisions and a global Deep Neural Network (DNN) trained with global experiences. Then, the online phase with local, on-board, and pre-trained DNNs requires continual learning to evolve with the environment, which can be done in two different ways: (1) Model anticipation, where the predictable conditions of the constellation are exploited by each satellite sharing local model with the next satellite; and (2) Federated Learning (FL), where each agent's model is merged first at the cluster level and then aggregated in a global Parameter Server. The results show that, without high congestion, the proposed Multi-Agent DRL framework achieves the same E2E performance as a shortest-path solution, but the latter assumes intensive communication overhead for real-time network-wise knowledge of the system at a centralized node, whereas ours only requires limited feedback exchange among first neighbour satellites. Importantly, our solution adapts well to congestion conditions and exploits less loaded paths. Moreover, the divergence of models over time is easily tackled by the synergy between anticipation, applied in short-term alignment, and FL, utilized for long-term alignment.

cs.LG↗

arXiv2024

Two-Timescale Design for Reconfigurable Intelligent Surface-Aided URLLC

In this paper, to tackle the blockage issue in massive multiple-input-multiple-output (mMIMO) systems, a reconfigurable intelligent surface (RIS) is seamlessly deployed to support devices with ultra-reliable and low-latency communications (URLLC). The transmission power of the base station and the phase shifts of the RIS are jointly devised to maximize the weighted sum rate while considering the spatially correlation and channel estimation errors. Firstly, \textcolor{black}{the relationship between the channel estimation error and spatially correlated RIS's elements is revealed by using the linear minimum mean square error}. Secondly, based on the maximum-ratio transmission precoding, a tight lower bound of the rate under short packet transmission is derived. Finally, the NP-hard problem is decomposed into two optimization problems, where the transmission power is obtained by geometric programming and phase shifts are designed by using gradient ascent method. Besides, we have rigorously proved that the proposed algorithm can rapidly converge to a sub-optimal solution with low complexity. Simulation results confirm the tightness between the analytic results and Monte Carlo simulations. Furthermore, the two-timescale scheme provides a practical solution for the short packet transmission.

eess.SP↗

arXiv2024

Intelligent Duty Cycling Management and Wake-up for Energy Harvesting IoT Networks with Correlated Activity

This paper presents an approach for energy-neutral Internet of Things (IoT) scenarios where the IoT devices (IoTDs) rely entirely on their energy harvesting capabilities to sustain operation. We use a Markov chain to represent the operation and transmission states of the IoTDs, a modulated Poisson process to model their energy harvesting process, and a discrete-time Markov chain to model their battery state. The aim is to efficiently manage the duty cycling of the IoTDs, so as to prolong their battery life and reduce instances of low-energy availability. We propose a duty-cycling management based on K- nearest neighbors, aiming to strike a trade-off between energy efficiency and detection accuracy. This is done by incorporating spatial and temporal correlations among IoTDs' activity, as well as their energy harvesting capabilities. We also allow the base station to wake up specific IoTDs if more information about an event is needed upon initial detection. Our proposed scheme shows significant improvements in energy savings and performance, with up to 11 times lower misdetection probability and 50\% lower energy consumption for high-density scenarios compared to a random duty cycling benchmark.

eess.SY↗

arXiv2024

Deep Reinforcement Learning for Multi-User RF Charging with Non-linear Energy Harvesters

Radio frequency (RF) wireless power transfer (WPT) is a promising technology for sustainable support of massive Internet of Things (IoT). However, RF-WPT systems are characterized by low efficiency due to channel attenuation, which can be mitigated by precoders that adjust the transmission directivity. This work considers a multi-antenna RF-WPT system with multiple non-linear energy harvesting (EH) nodes with energy demands changing over discrete time slots. This leads to the charging scheduling problem, which involves choosing the precoders at each slot to minimize the total energy consumption and meet the EH requirements. We model the problem as a Markov decision process and propose a solution relying on a low-complexity beamforming and deep deterministic policy gradient (DDPG). The results show that the proposed beamforming achieves near-optimal performance with low computational complexity, and the DDPG-based approach converges with the number of episodes and reduces the system's power consumption, while the outage probability and the power consumption increase with the number of devices.

eess.SP↗

arXiv2024

EcoPull: Sustainable IoT Image Retrieval Empowered by TinyML Models

This paper introduces EcoPull, a sustainable Internet of Things (IoT) framework empowered by tiny machine learning (TinyML) models for fetching images from wireless visual sensor networks. Two types of learnable TinyML models are installed in the IoT devices: i) a behavior model and ii) an image compressor model. The first filters out irrelevant images for the current task, reducing unnecessary transmission and resource competition among the devices. The second allows IoT devices to communicate with the receiver via latent representations of images, reducing communication bandwidth usage. However, integrating learnable modules into IoT devices comes at the cost of increased energy consumption due to inference. The numerical results show that the proposed framework can save > 70% energy compared to the baseline while maintaining the quality of the retrieved images at the ES.

cs.NI↗

arXiv2024

Digital Twin of Industrial Networked Control System based on Value of Information

The paper examines a scenario wherein sensors are deployed within an Industrial Networked Control System, aiming to construct a digital twin (DT) model for a remotely operated Autonomous Guided Vehicle (AGV). The DT model, situated on a cloud platform, estimates and predicts the system's state, subsequently formulating the optimal scheduling strategy for execution in the physical world. However, acquiring data crucial for efficient state estimation and control computation poses a significant challenge, primarily due to constraints such as limited network resources, partial observation, and the necessity to maintain a certain confidence level for DT estimation. We propose an algorithm based on Value of Information (VoI), seamlessly integrated with the Extended Kalman Filter to deliver a polynomial-time solution, selecting the most informative subset of sensing agents for data. Additionally, we put forth an alternative solution leveraging a Graph Neural Network to precisely ascertain the AGV's position with a remarkable accuracy of up to 5 cm. Our experimental validation in an industrial robotic laboratory environment yields promising results, underscoring the potential of high-accuracy DT models in practice.

eess.SP↗

arXiv2024

Coexistence of Push Wireless Access with Pull Communication for Content-based Wake-up Radios

This paper considers energy-efficient connectivity for Internet of Things (IoT) devices in a coexistence scenario between two distinctive communication models: pull- and push-based. In pull-based, the base station (BS) decides when to retrieve a specific type of data from the IoT devices, while in push-based, the IoT device decides when and which data to transmit. To this end, this paper advocates introducing the content-based wake-up (CoWu), which enables the BS to remotely activate only a subset of pull-based nodes equipped with wake-up receivers, observing the relevant data. In this setup, a BS pulls data with CoWu at a specific time instance to fulfill its tasks while collecting data from the nodes operating with a push-based communication model. The resource allocation plays an important role: longer data collection duration for pull-based nodes can lead to high retrieval accuracy while decreasing the probability of data transmission success for push-based nodes, and vice versa. Numerical results show that CoWu can manage communication requirements for both pull-based and push-based nodes while realizing the high energy efficiency (up to 38%) of IoT devices, compared to the baseline scheduling method.

cs.NI↗

arXiv2024

Deterministic Patterns for Multiple Access with Latency and Reliability Guarantees

We study a scenario in which multiple uncoordinated devices aim to achieve reliable transmissions within a given time frame. The devices are intermittently active and access a shared pool of channel resources in a grant-free manner by utilizing multiple transmissions (K-repetition coding). This allows them to achieve diversity and improve the reliability within a certain latency constraint. We focus on two access methods: one where devices choose K slots at random and another one where the access patterns are deterministic and follow a specific code design, namely the Steiner System. We analyze the problem under two signal models that involve different complexity for the receiver. First, collision model is considered, where only interference-free transmissions can be used and combined. Second, a model treating interference as noise is analyzed, where the receiver is capable of utilizing all K replicas, applying maximum ratio combining (MRC). For both signal models, we investigate receivers with and without successive interference cancellation (SIC). We develop approximations and bounds for the outage probabilities that very closely match simulation results. Overall, we show that deterministic access patterns have the potential to significantly outperform random selection in terms of reliability. Furthermore, deterministic access patterns offer a simplified system design.

cs.IT↗

arXiv2024

Decentralized Multi-Party Multi-Network AI for Global Deployment of 6G Wireless Systems

Multiple visions of 6G networks elicit Artificial Intelligence (AI) as a central, native element. When 6G systems are deployed at a large scale, end-to-end AI-based solutions will necessarily have to encompass both the radio and the fiber-optical domain. This paper introduces the Decentralized Multi-Party, Multi-Network AI (DMMAI) framework for integrating AI into 6G networks deployed at scale. DMMAI harmonizes AI-driven controls across diverse network platforms and thus facilitates networks that autonomously configure, monitor, and repair themselves. This is particularly crucial at the network edge, where advanced applications meet heightened functionality and security demands. The radio/optical integration is vital due to the current compartmentalization of AI research within these domains, which lacks a comprehensive understanding of their interaction. Our approach explores multi-network orchestration and AI control integration, filling a critical gap in standardized frameworks for AI-driven coordination in 6G networks. The DMMAI framework is a step towards a global standard for AI in 6G, aiming to establish reference use cases, data and model management methods, and benchmarking platforms for future AI/ML solutions.

cs.NI↗

arXiv2024

Coexistence of Pull and Push Communication in Wireless Access for IoT Devices

We consider a setup with Internet of Things (IoT), where a base station (BS) collects data from nodes that use two different communication modes. The first is pull-based, where the BS retrieves the data from specific nodes through queries. In addition, the nodes that apply pull-based communication contain a wake-up receiver: upon a query, the BS sends wake-up signal (WuS) to activate the corresponding devices equipped with wake-up receiver (WuDs). The second one is push-based communication, in which the nodes decide when to send to the BS. Consider a time-slotted model, where the time slots in each frame are shared for both pull-based and push-based communications. Therein, this coexistence scenario gives rise to a new type of problem with fundamental trade-offs in sharing communication resources: the objective to serve a maximum number of queries, within a specified deadline, limits the transmission opportunities for push sensors, and vice versa. This work develops a mathematical model that characterizes these trade-offs, validates them through simulations, and optimizes the frame design to meet the objectives of both the pull- and push-based communications.

eess.SP↗

arXiv2024

Impact of Inter-Operator Interference via Reconfigurable Intelligent Surfaces

A wireless communication system is studied that operates in the presence of multiple reconfigurable intelligent surfaces (RISs). In particular, a multi-operator environment is considered where each operator utilizes an RIS to enhance its communication quality. Although out-of-band interference does not exist (since each operator uses isolated spectrum resources), RISs controlled by different operators do affect the system performance of one another due to the inherently rapid phase shift adjustments that occur on an independent basis. The system performance of such a communication scenario is analytically studied for the practical case where discrete-only phase shifts occur at RIS. The proposed framework is quite general since it is valid under arbitrary channel fading conditions as well as the presence (or not) of the transceiver's direct link. Finally, the derived analytical results are verified via numerical and simulation trial as well as some novel and useful engineering outcomes are manifested.

cs.IT↗

arXiv2024

Time, Simultaneity, and Causality in Wireless Networks with Sensing and Communications

Wireless systems beyond 5G evolve towards embracing both sensing and communication, resulting in increased convergence of the digital and the physical world. The existence of fused digital-physical realms raises critical questions regarding temporal ordering, causality, and the synchronization of events. This paper addresses the temporal challenges arising from the fact that the wireless infrastructure becomes an entity with multisensory perception. With the growing reliance on real-time interactions and applications such as digital twins, extended reality, and the metaverse, the need for accurate timestamping and temporal forensics becomes crucial. The paper introduces a model that incorporates Temporal Windows of Integration (TWI) to emulate human multisensory perception and discusses the implications for setting timing constraints in real-time applications and enabling temporal forensics. The analysis explores trade-offs, probabilities, and bounds for simultaneity and causality violation in the context of wireless systems evolving towards perceptive networks. This work underscores the significance of timestamping in the evolving wireless landscape, provide insights into system-level implications, and points out new research avenues for systems that combine sensing and communications.

cs.NI↗

arXiv2024

Experimental Study of Spatial Statistics for Ultra-Reliable Communications

This paper presents an experimental validation for prediction of rare fading events using channel distribution information (CDI) maps that predict channel statistics from measurements acquired at surrounding locations using spatial interpolation. Using experimental channel measurements from 127 locations, we demonstrate the use case of providing statistical guarantees for rate selection in ultra-reliable low-latency communication (URLLC) using CDI maps. By using only the user location and the estimated map, we are able to meet the desired outage probability with a probability between 93.6-95.6% targeting 95%. On the other hand, a model-based baseline scheme that assumes Rayleigh fading meets the target outage requirement with a probability of 77.2%. The results demonstrate the practical relevance of CDI maps for resource allocation in URLLC.

eess.SP↗