arXiv subjects

Hanna Bogucka

Publications and source records attributed to Hanna Bogucka.

18 recordsLinked to original sources

arXiv2026

On the Vulnerability of Deep Automatic Modulation Classifiers to Explainable Backdoor Threats

Deep learning (DL) has been widely studied for assisting applications of modern wireless communications. One of the applications is automatic modulation classification (AMC). However, DL models are found to be vulnerable to adversarial machine learning (AML) threats. One of the most persistent and stealthy threats is the backdoor (Trojan) attack. Nevertheless, most studied threats originate from other AI domains, such as computer vision (CV). Therefore, in this paper, a physical backdoor attack targeting the wireless signal before transmission is studied. The adversary is considered to be using explainable AI (XAI) to guide the placement of the trigger in the most vulnerable parts of the signal. Then, a class prototype combined with principal components is used to generate the trigger. The studied threat was found to be efficient in breaching multiple DL-based AMC models. The attack achieves high success rates for a wide range of SNR values and a small poisoning ratio.

cs.CR↗

arXiv2026

Mitigating Evasion Attacks in Fog Computing Resource Provisioning Through Proactive Hardening

This paper investigates the susceptibility to model integrity attacks that overload virtual machines assigned by the k-means algorithm used for resource provisioning in fog networks. The considered k-means algorithm runs two phases iteratively: offline clustering to form clusters of requested workload and online classification of new incoming requests into offline-created clusters. First, we consider an evasion attack against the classifier in the online phase. A threat actor launches an exploratory attack using query-based reverse engineering to discover the Machine Learning (ML) model (the clustering scheme). Then, a passive causative (evasion) attack is triggered in the offline phase. To defend the model, we suggest a proactive method using adversarial training to introduce attack robustness into the classifier. Our results show that our mitigation technique effectively maintains the stability of the resource provisioning system against attacks.

cs.CR↗

arXiv2026

Physical Backdoor Attack Against Deep Learning-Based Modulation Classification

Deep Learning (DL) has become a key technology that assists radio frequency (RF) signal classification applications, such as modulation classification. However, the DL models are vulnerable to adversarial machine learning threats, such as data manipulation attacks. We study a physical backdoor (Trojan) attack that targets a DL-based modulation classifier. In contrast to digital backdoor attacks, where digital triggers are injected into the training dataset, we use power amplifier (PA) non-linear distortions to create physical triggers before the dataset is formed. During training, the adversary manipulates amplitudes of RF signals and changes their labels to a target modulation scheme, training a backdoored model. At inference, the adversary aims to keep the backdoor attack inactive such that the backdoored model maintains high accuracy on test signals. However, if they apply the same manipulation used during training on these test signals, the backdoor is activated, and the model misclassifies these signals. We demonstrate that our proposed attack achieves high attack success rates with few manipulated RD signals for different noise levels. Furthermore, we test the resilience of the proposed attack to multiple defense techniques, and the results show that these techniques fail to mitigate the attack.

cs.CR↗

arXiv2025

Nonlinear symbols combining for Power Amplifier-distorted OFDM signal reception

Nonlinear distortion of a multicarrier signal by a transmitter Power Amplifier (PA) can be a serious problem when designing new highly energy-efficient wireless systems. Although the performance of standard reception algorithms is seriously deteriorated by the nonlinear distortion, the more advanced solutions allow the utilization of additional frequency diversity caused by nonlinear PA. However, while most of the advanced receivers are decision-aided, their gains are observed mostly in a relatively low Bit Error Rate (BER) region, not targeted by adaptive Modulation Coding Schemes utilizing Forward Error Correction (FEC). In this paper, a non-decision-aided Higher-Order Combining (HOC) reception scheme is proposed. While the analytical formulas for finding symbols combining coefficients are not known, machine learning is used for deriving them. The simulation results show an improved BER performance with respect to a standard reception and one of the established decision-aided receivers. However, as HOC has computational complexity that increases rapidly with the number of subcarriers utilized, more studies are needed to apply it in a wideband system.

cs.NI↗

arXiv2025

An Open-RAN Testbed for Detecting and Mitigating Radio-Access Anomalies

This paper presents the Open Radio Access Net-work (O-RAN) testbed for secure radio access. We discuss radio-originating attack detection and mitigation methods based on anomaly detection and how they can be implemented as specialized applications (xApps) in this testbed. We also pre-sent illustrating results of the methods applied in real-world scenarios and implementations.

cs.CR↗

arXiv2023

Open RAN xApps Design and Evaluation: Lessons Learnt and Identified Challenges

Open Radio Access Networks (RAN) offer diverse economic opportunities. A transition to a flexible, modular approach within the disaggregated RAN framework is crucial, involving careful planning of RAN architecture and the deployment of specialized software applications. Collaboration across sectors is essential for efficiency and reliability, with the open-source community driving innovation. This paper explores challenges for third-party application developers in Open RAN. It provides a comparative analysis of solutions, focusing on xApp development and implementation. Challenges arise in two areas: the complexities of xApp development, particularly for advanced use cases like beam management, and issues in low-level software implementation within open platforms. In conclusion, key challenges must promote academia-industry collaboration in Open RAN. This paper shares early lessons from xApp development, guiding the field's evolution.

cs.NI↗

arXiv2023

Uberization of telecom networks for cost-efficient communication and computing

This paper discusses the uberization of telecommunication and computing network services. The Uber-like platform business model is discussed for application in future networks together with interesting analogies of communication and computing (2C) resource-sharing models. The economy of this sharing is discussed, and some recommendations for network uberization are provided.

cs.NI↗

arXiv2023

Green Time-Critical Fog Communication and Computing

Fog computing allows computationally-heavy problems with tight time constraints to be solved even if end devices have limited computational resources and latency induced by cloud computing is too high. How can energy consumed by fog computing be saved while obeying latency constraints and considering not only computations but also transmission through wireless and wired links? This work examines the latency and energy consumption sources in fog networks and discusses models describing these costs for various technologies. Next, resource allocation strategies are discussed considering the various degrees of freedom available in such a complex system, and their influence on energy consumption and latency. Finally, a vision for a future distributed, AI-driven resources allocation strategy is presented and justified.

eess.SY↗

arXiv2023

Secure Federated Learning for Cognitive Radio Sensing

This paper considers reliable and secure Spectrum Sensing (SS) based on Federated Learning (FL) in the Cognitive Radio (CR) environment. Motivation, architectures, and algorithms of FL in SS are discussed. Security and privacy threats on these algorithms are overviewed, along with possible countermeasures to such attacks. Some illustrative examples are also provided, with design recommendations for FL-based SS in future CRs.

eess.SP↗

arXiv2022

Task Allocation for Energy Optimization in Fog Computing Networks with Latency Constraints

Fog networks offer computing resources with varying capacities at different distances from end users. A Fog Node (FN) closer to the network edge may have less powerful computing resources compared to the cloud, but processing of computational tasks in an FN limits long-distance transmission. How should the tasks be distributed between fog and cloud nodes? We formulate a universal non-convex Mixed-Integer Nonlinear Programming (MINLP) problem minimizing task transmission- and processing-related energy with delay constraints to answer this question. It is transformed with Successive Convex Approximation (SCA) and decomposed using the primal and dual decomposition techniques. Two practical algorithms called Energy-EFFicient Resource Allocation (EEFFRA) and Low-Complexity (LC)-EEFFRA are proposed. They allow for successful distribution of network requests between FNs and the cloud in various scenarios significantly reducing the average energy cost and decreasing the number of computational requests with unmet delay requirements.

cs.DC↗

arXiv2021

Beyond 5G: Big Data Processing for Better Spectrum Utilization

This article emphasizes the great potential of big data processing for advanced user- and situation-oriented, so context-aware resource utilization in future wireless networks. In particular, we consider the application of dedicated, detailed, and rich-in-content maps and records called Radio Service Maps, (RSM) for unlocking the spectrum opportunities in 6G networks. Due to the characteristics of 5G, in the future, there will be a need for high convergence of various types of wireless networks, such as cellular and the Internet-of-Things (IoT) networks, which are steadily growing and consequently considered as the studied use case in this work. We show that the 6G network significantly benefits from effective Dynamic Spectrum management (DSM) based on RSM which provides rich and accurate knowledge of the radio context; a knowledge that is stored and processed within database-oriented subsystems designed to support wireless networks for improving spectral efficiency. In this article, we discuss context-aware RSM subsystem architecture and operation for DSM in convergent 6G radio and IoT networks. By providing various use-cases, we demonstrate that the accurate definition and access to the rich context information lead to a significant improvement of the system performance. In consequence, we also claim that efficient big-data processing algorithms will be necessary for future applications.

cs.NI↗

arXiv2021

Coopetition methodology for resource sharing in distributed OFDM-based cognitive radio networks

In this paper, we present a distributed resource allocation mechanism in cognitive radio networks, based on a new coopeti-tion methodology, which combines advantages of nodes competition and cooperation. We postulate that this new method allows for fully distributed resource management between cognitive radio devices. The presented framework is generic, however, we consider it for the application in OFDMA networks. Coopetition takes the best from cooperative and competitive problem formulation and provides the opportunity to control the balance between fairness and spectral efficiency (SE) of resource allocation. Simulation results confirm that coopetition allows for efficient resource utilization, and may be used practically in wireless cognitive networks.

cs.NI↗

arXiv2021

Out-of-Band Power Reduction in NC-OFDM with Optimized Cancellation Carriers Selection

In this letter, we propose a computationally efficient method for joint selection of cancellation carriers (CCs) and calculation of their values minimizing the out-of-band (OOB) power in non-contiguous (NC-) OFDM transmission. The proposed new CCs selection method achieves higher OOB power attenuation than algorithms known from literature as well as noticable reception performance improvement.

cs.SI↗

arXiv2021

Energy Savings by Task Offloading to a Fog Considering Radio Front-End Characteristics

Fog computing can be used to offload computationally intensive tasks from battery powered Internet of Things (IoT) devices. Although it reduces energy required for computations in an IoT device, it uses energy for communications with the fog. This paper analyzes when usage of fog computing is more energy efficient than local computing. Detailed energy consumption models are built in both scenarios with the focus set on the relation between energy consumption and distortion introduced by a Power Amplifier (PA). Numerical results show that task offloading to a fog is the most energy efficient for short, wideband links.

cs.NI↗

arXiv2021

In-Band-Interference Robust Synchronization Algorithm for an NC-OFDM System

We consider receiver synchronization in the non-continguous orthogonal frequency division multiplexing (NC-OFDM)-based radio system in the presence of in-band interfering signal, which occupies the frequency-band between blocks of subcarriers (SCs) used by this system, i.e. in-band of NC-OFDM receiver spectrum range. This paper proposes a novel preamble-based synchronization algorithm for estimation of the time and frequency offset based on the received signal cross-correlation with the reference preamble. Contrary to the existing algorithms, it is robust against in-band interference including narrowband interference at the cost of increased complexity. Moreover, in the interference-free system, the probability of frame synchronization error is improved in comparison to all simulated algorithms.

cs.NI↗

arXiv2021

Low Complex, Narrowband-Interference Robust Synchronization for NC-OFDM Cognitive Radio

This paper presents a new, low-complexity algorithm for time and frequency synchronization in a non-contiguous-orthogonal frequency division multiplexing (NC-OFDM) radio communication system in the presence of narrowband interference (NBI). This interference scenario is plausible for cognitive radio systems. The computational complexity of the proposed synchronization method is similar to the known Schmidl&Cox algorithm. The algorithm is partially blind, i.e., it does not require information on the NC-OFDM subcarriers used, nor on the center-frequency of the NBI. Moreover, it does not require filtering to remove the NBI. It performs best for the NBI of constant carrier frequency, but it also provides good results for the frequency-modulated interfering signal.

cs.SI↗

arXiv2018

Context-based spectrum sharing in 5G wireless networks based on Radio Environment Maps

Dynamic spectrum sharing can provide many benefits to wireless networks operators. However, its efficiency requires sophisticated control mechanisms. The more context information is used by it, the higher performance of networks is expected. A facility for collecting this information, processing it and controlling base stations managed by various network operators is a so-called Radio Environment Map (REM) subsystem. This paper proposes REM-based schemes for the allocation of base stations power levels in 4G/5G networks, while considering interference generated to a licensed network. It is assumed that both networks have different profiles of served users, e.g., area of their positions and movement, which opens opportunities for spectrum sharing. The proposed schemes have been evaluated by means of extensive system-level simulations and compared with two widely adopted policy-based spectrum sharing reference schemes. Simulation results show that dynamic schemes utilizing rich context information, outperforms static, policy-based spectrum sharing schemes.

cs.NI↗

arXiv2009

On the Benefits of Bandwidth Limiting in Decentralized Vector Multiple Access Channels

We study the network spectral efficiency of decentralized vector multiple access channels (MACs) when the number of accessible dimensions per transmitter is strategically limited. Considering each dimension as a frequency band, we call this limiting process bandwidth limiting (BL). Assuming that each transmitter maximizes its own data rate by water-filling over the available frequency bands, we consider two scenarios. In the first scenario, transmitters use non-intersecting sets of bands (spectral resource partition), and in the second one, they freely exploit all the available frequency bands (spectral resource sharing). In the latter case, successive interference cancelation (SIC) is used. We show the existence of an optimal number of dimensions that a transmitter must use in order to maximize the network performance measured in terms of spectral efficiency. We provide a closed form expression for the optimal number of accessible bands in the first scenario. Such an optimum point, depends on the number of active transmitters, the number of available frequency bands and the different signal-to-noise ratios. In the second scenario, we show that BL does not bring a significant improvement on the network spectral efficiency, when all transmitters use the same BL policy. For both scenarios, we provide simulation results to validate our conclusions.

cs.IT↗