arXiv Science⌕ Search

arXiv subjects

Marcin Hoffmann

Publications and source records attributed to Marcin Hoffmann.

At least 19 recordsLinked to original sources
arXiv2026

WIP: Energy-Efficient LLM-Based Serving Cluster Formulation in Cell-Free Massive MIMO

One way to increase the Energy Efficiency (EE) of 6G wireless networks is to utilize existing network infrastructure more efficiently. This can be achieved by introducing User-Centric Cell-Free Massive Multiple-Input-Multiple-Output (UCCF MMMIMO), which allows for simultaneously serving a single user by multiple Base Stations (BSs). From this perspective, the key challenge is to decide which BSs should serve a given user, known as the Serving Cluster Formulation (SCF). In this paper, we propose to deal with this problem by using an Artificial Intelligence (AI) agent based on a Large Language Model (LLM), targeting improvement of EE. We evaluated the proposed AI agent using a complex, 3D Ray Tracer-based, cellular network simulator, comparing a few GPT models and state-of-the-art algorithms. The results show up to 32% gain in EE of the proposed AI agent compared to the baseline.

cs.NI↗
arXiv2026

A Unified E2E Energy Efficiency Testing Framework for Open RAN

Energy efficiency (EE) is one of the key challenges for contemporary and future mobile networks, including within the Open Radio Access Network (O-RAN) architecture. However, there is a significant gap in common procedures for comparing the EE of both hardware (HW) and software (SW) solutions offered by various vendors. Usually, EE improvements of both SW and HW solutions are demonstrated in a specific scenario defined by individual vendors avoiding comparisons and benchmarking under various network conditions. This paper outlines the need for unified end-to-end (E2E) EE testing for O-RAN. First, it analyzes the standards to identify missing parts. Based on the analysis, a novel O-RAN E2E EE Testing framework is proposed. The framework aims to test the EE of the xApp/rApp pair cooperating on the cell on/off switching using a commercial RAN emulator and real-world network topology data from a mobile network operator (MNO). The test results show up to 57% improvement in EE compared to the baseline.

cs.NI↗
arXiv2026

Large-scale wireless network management via Open-RAN Tandem Apps: Cell on/off switching use case

With growing mobile-network complexity, management and optimization have become increasingly difficult. Centralized algorithms face high control-data overhead and computational load, while distributed approaches often perform far from optimally. The O-RAN architecture introduces two tiers of RAN Intelligent Controllers (RICs), enabling hierarchical network-management schemes. This work proposes Tandem Apps: a pair of tightly coupled optimization mechanisms running on both controllers. We show how to design Tandem Apps through architectural and functional splitting to achieve an agile, low-complexity solution that still preserves a global network view. As an example, we implement Tandem Apps for cell on/off switching and evaluate them in a large heterogeneous network using real network data. Although the Tandem Apps concept is new, it remains fully compliant with the O-RAN standard, as validated using commercial network software.

cs.NI↗
arXiv2026

Impact of Nonlinear Power Amplifier on Massive MIMO: Machine Learning Prediction Under Realistic Radio Channel

M-MIMO is one of the crucial technologies for increasing spectral and energy efficiency of wireless networks. Most of the current works assume that M-MIMO arrays are equipped with a linear front end. However, ongoing efforts to make wireless networks more energy-efficient push the hardware to the limits, where its nonlinear behavior appears. This is especially a common problem for the multicarrier systems, e.g., OFDM used in 4G, 5G, and possibly also in 6G, which is characterized by a high Peak-to-Average Power Ratio. While the impact of a nonlinear Power Amplifier (PA) on an OFDM signal is well characterized, it is a relatively new topic for the M-MIMO OFDM systems. Most of the recent works either neglect nonlinear effects or utilize simplified models proper for Rayleigh or LoS radio channel models. In this paper, we first theoretically characterize the nonlinear distortion in the M-MIMO system under commonly used radio channel models. Then, utilizing 3D-Ray Tracing (3D-RT) software, we demonstrate that these models are not very accurate. Instead, we propose two models: a statistical one and an ML-based one using 3D-RT results. The proposed statistical model utilizes the Generalized Extreme Value (GEV) distribution to model Signal to Distortion Ratio (SDR) for victim users, receiving nonlinear distortion, e.g., as interference from neighboring cells. The proposed ML model aims to predict SDR for a scheduled user (receiving nonlinear distortion along with the desired signal), based on the spatial characteristics of the radio channel and the operation point of each PA feeding at the M-MIMO antenna array. The predicted SDR can then be used to perform PA-aware per-user power allocation. The results show about 12% median gain in user throughput achieved by the proposed ML-based power allocation scheme over the state-of-the-art, fixed operating point scheme.

cs.LG↗
arXiv2026

Badanie Sieci Massive MIMO o Architekturze Zorientowanej na Uzytkownika

The future 6G networks are expected to utilize large antenna arrays and follow the user-centric architecture, where the user is being served by all base stations. This work evaluates such a system within an advanced system-level simulator, which utilizes an accurate 3D Ray-Tracing radio channel model. Results show that the novel user-centric network architecture can increase the cell-edge users throughput by a fold of 3.

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↗
arXiv2025

O-RAN for Energy-Efficient Serving Cluster Formulation in User-Centric Cell-Free MMIMO

The 6G Massive Multiple-Input Multiple-Output (MMIMO) networks can follow the so-called User-Centric Cell-Free (UCCF) architecture, where a single user is served by multiple Access Points (APs) coordinated by the Central Processing Unit (CPU). In this paper, we propose how O-RAN functionalities, i.e., rApp-xApp pair, can be used for energy-efficient Serving Cluster Formulation (SCF). Simulation studies show up to 37\% gain in Energy Efficiency (EE) of the proposed solution over the state-of-the-art Network-Centric (NC) designs.

cs.IT↗
arXiv2025

Formowanie klastrów obsługujących w sieci 6g open ran o architekturze zorientowanej na użytkownika

One of the main challenges associated with the implementation of an Open RAN User-Centric Cell-Free network is the appropriate formulation of serving clusters. This problem can be effectively solved using a proposed algorithm that leverages cooperation between xApp and rApp. Simulation results have shown that the proposed solution allows for a 16% increase in the median of the user throughput distribution compared to the state-of-the-art network-centric architecture

cs.NI↗
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

Why white noise is not enough? On using radio front-end models while designing 6G PHY

From generation to generation there are increasing requirements for wireless standards both in terms of spectral and energy efficiency. While up to now the layered wireless transceiver architecture worked allowing for, e.g., separation of channel decoding algorithms from front-end design, this may need reconsideration in the 6G era. Especially the hardware-originated distortions have to be taken into account while designing other layer algorithms as the high throughput and energy efficiency requirements will push these devices to their limit revealing their nonlinear characteristics. This position paper will shed some light on new degrees of freedom while cross-layer designing and controlling multicarrier and multiantenna transceivers of 6G systems.

cs.NI↗
arXiv2023

Beam Management Driven by Radio Environment Maps in O-RAN Architecture

The Massive Multiple-Input Multiple-Output (M-MIMO) is considered as one of the key technologies in 5G, and future 6G networks. From the perspective of, e.g., channel estimation, especially for high-speed users it is easier to implement an M-MIMO network exploiting a static set of beams, i.e., Grid of Beams (GoB). While considering GoB it is important to properly assign users to the beams, i.e., to perform Beam Management (BM). BM can be enhanced by taking into account historical knowledge about the radio environment, e.g., to avoid radio link failures. The aim of this paper is to propose such a BM algorithm, that utilizes location-dependent data stored in a Radio Environment Map (REM). It utilizes received power maps, and user mobility patterns to optimize the BM process in terms of Reinforcement Learning (RL) by using the Policy Iteration method under different goal functions, e.g., maximization of received power or minimization of beam reselections while avoiding radio link failures. The proposed solution is compliant with the Open Radio Access Network (O-RAN) architecture, enabling its practical implementation. Simulation studies have shown that the proposed BM algorithm can significantly reduce the number of beam reselections or radio link failures compared to the baseline algorithm.

cs.NI↗
arXiv2023

Signaling Storm Detection in IIoT Network based on the Open RAN Architecture

The Industrial Internet of Things devices due to their low cost and complexity are exposed to being hacked and utilized to attack the network infrastructure causing a so-called Signaling Storm. In this paper, we propose to utilize the Open Radio Access Network (O-RAN) architecture, to monitor the control plane messages in order to detect the activity of adversaries at its early stage.

cs.NI↗
arXiv2022

Radio Environment Map and Deep Q-Learning for 5G Dynamic Point Blanking

Dynamic Point Blanking (DPB) is one of the Coordinated MultiPoint (CoMP) techniques, where some Base Stations (BSs) can be temporarily muted, e.g., to improve the cell-edge users throughput. In this paper, it is proposed to obtain the muting pattern that improves cell-edge users throughput with the use of a Deep Q-Learning. The Deep Q-Learning agent is trained on location-dependent data. Simulation studies have shown that the proposed solution improves cell-edge user throughput by about 20.6%.

cs.NI↗
arXiv2022

Federated Learning-Based Interference Modeling for Vehicular Dynamic Spectrum Access

A platoon-based driving is a technology allowing vehicles to follow each other at close distances to, e.g., save fuel. However, it requires reliable wireless communications to adjust their speeds. Recent studies have shown that the frequency band dedicated for vehicle-to-vehicle communications can be too busy for intra-platoon communications. Thus it is reasonable to use additional spectrum resources, of low occupancy, i.e., secondary spectrum channels. The challenge is to model the interference in those channels to enable proper channel selection. In this paper, we propose a two-layered Radio Environment Map (REM) that aims at providing platoons with accurate location-dependent interference models by using the Federated Learning approach. Each platoon is equipped with a Local REM that is updated on the basis of raw interference samples and previous interference model stored in the Global REM. The model in global REM is obtained by merging models reported by platoons. The nodes exchange only parameters of interference models, reducing the required control channel capacity. Moreover, in the proposed architecture platoon can utilize Local REM to predict channel occupancy, even when the connection to the Global REM is temporarily unavailable. The proposed system is validated via computer simulations considering non-trivial interference patterns.

cs.NI↗
arXiv2021

Frequency Selection for Platoon Communications in Secondary Spectrum Using Radio Environment Maps

Platoon-based driving is an idea that vehicles follow each other at a close distance, in order to increase road throughput and fuel savings. This requires reliable wireless communications to adjust the speeds of vehicles. Although there is a dedicated frequency band for vehicle-to-vehicle (V2V) communications, studies have shown that it is too congested to provide reliable transmission for the platoons. Additional spectrum resources, i.e., secondary spectrum channels, can be utilized when these are not occupied by other users. Characteristics of interference in these channels are usually location-dependent and can be stored in the so-called Radio Environment Maps (REMs). This paper aims to design REM, in order to support the selection of secondary spectrum channel for intra-platoon communications. We propose to assess the channel's quality in terms of outage probability computed, with the use of estimated interference distributions stored in REM. A frequency selection algorithm that minimizes the number of channel switches along the planned platoon route is proposed. Additionally, the REM creation procedure is shown that reduces the number of database entries using (Density-Based Spatial Clustering of Applications with Noise) DBSCAN algorithm. The proposals are tested using real IQ samples captured on a real road. Application of the DBSCAN clustering to the constructed REM provided 7% reduction in its size. Utilization of the proposed channel selection algorithm resulted in a 35 times reduction of channel switches concerning channel assignment performed independently in every location.

cs.NI↗
arXiv2021

Similarity Measures for Location-Dependent MMIMO, 5G Base Stations On/Off Switching Using Radio Environment Map

The Massive Multiple-Input Multiple-Output (MMIMO) technique together with Heterogeneous Network (Het-Net) deployment enables high throughput of 5G and beyond networks. However, a high number of antennas and a high number of Base Stations (BSs) can result in significant power consumption. Previous studies have shown that the energy efficiency (EE) of such a network can be effectively increased by turning off some BSs depending on User Equipments (UEs) positions. Such mapping is obtained by using Reinforcement Learning. Its results are stored in a so-called Radio Environment Map (REM). However, in a real network, the number of UEs' positions patterns would go to infinity. This paper aims to determine how to match the current set of UEs' positions to the most similar pattern, i.e., providing the same optimal active BSs set, saved in REM. We compare several state-of-the-art distance metrics using a computer simulator: an accurate 3D-Ray-Tracing model of the radio channel and an advanced system-level simulator of MMIMO Het-Net. The results have shown that the so-called Sum of Minimums Distance provides the best matching between REM data and UEs' positions, enabling up to 56% EE improvement over the scenario without EE optimization.

cs.NI↗