Radio Access Network Research Articles

Deep Reinforcement Learning-Based Task Assignment for Cooperative Mobile Edge Computing

Mobile edge computing (MEC) integrates computing resources in wireless access networks to process computational tasks in close proximity to mobile users with low latency. This paper investigates the task assignment problem for cooperative MEC networks in which a set of geographically distributed heterogeneous edge servers not only cooperate with remote cloud data centers but also help each other to jointly process user tasks. We introduce a novel stochastic MEC cooperation framework to model the edge-to-edge horizontal cooperation and the edge-to-cloud vertical cooperation. The task assignment optimization problem is formulated by taking into consideration dynamic network states, uncertain node computing capabilities and task arrivals, as well as the heterogeneity of the involved entities. We then develop and compare three task assignment algorithms, based on different deep reinforcement learning (DRL) approaches, value-based, policy-based, and hybrid approaches. In addition, to reduce the search space and computation complexity of the algorithms, we propose decomposition and function approximation techniques by leveraging the structure of the underlying problem. The evaluation results show that the proposed DRL-based task assignment schemes outperform the existing algorithms, and the hybrid actor-critic scheme performs the best under dynamic MEC network environments.

A Study on the Implementation of a Network Function for Real-time False Base Station Detection for the Next Generation Mobile Communication Environment

The threat posed by false base stations remains pertinent across the 4G, 5G, and forthcoming 6G generations of mobile communication. In response, this paper introduces a real-time detection method for false base stations employing two approaches: machine learning and specification-based. Utilizing the UERANSIM open 5G RAN (Radio-Access Network) test platform, we assess the feasibility and practicality of applying these techniques within a 5G network environment. Emulating real-world 5G conditions, we implement a functional split in the 5G base station and deploy the False Base Station Detection Function (FDF) as a 5G NF (Network Function) within the CU (Central Unit). This setup enables real-time detection seamlessly integrated into the network. Experimental results indicate that specification-based detection outperforms machine learning, achieving a detection accuracy of 99.6% compared to 96.75% for the highest-performing machine learning model XGBoost. Furthermore, specification-based detection demonstrates superior efficiency, with CPU usage of 1.2% and memory usage of 16.12MB, compared to 1.6% CPU usage and 182.4MB memory usage for machine learning on average. Consequently, the implementation of specification-based detection using the proposed method emerges as the most effective technique in the 5G environment.

AN OVERVIEW OF COMPONENTS FOR ENERGY EFFICIENT MULTIMEDIA NETWORKS BASED ON 5G RADIO ACCESS TECHNOLOGIES

Modern society is actively transitioning into an information-based one, and multimedia technologies have become an integral part of this process. Thanks to the proliferation of wireless access networks, users are becoming more mobile, and the development of the fifth generation of mobile networks, known as 5G, is a significant step in the advancement of information and communication technologies. 5G networks offer low latency and reliable connectivity, expanding the capabilities of mobile internet and machine communication. However, along with the opportunities provided by multimedia communication, there is a responsibility to consider the impact of associated technologies on the environment, as well as to address new challenges and the need for prudent resource utilization.
 The article defines the concept of "multimedia" and discusses various aspects of this concept, including digital storage and processing of information, components (text, photos, audio, and video), interactivity, and hypertextuality. It is noted that the transmission of multimedia data and the use of information technologies are closely linked to wireless access networks.
 The authors discuss challenges and solutions in the field of energy efficiency for networks. They provide statistics indicating that the information and communication technology (ICT) industry is responsible for a significant portion of global energy consumption and CO2 emissions, with radio access networks being a major contributor. Various components and technologies that can contribute to the development of energy-efficient multimedia networks based on 5G radio access technologies are examined. Specifically, heterogeneous networks, non-orthogonal multiple access (NOMA) technologies, and multiple-input multiple-output (MIMO) technologies are highlighted as key components for achieving energy efficiency.
 The importance of using heterogeneous networks to reduce the distance between transmitters and receivers is emphasized, along with the possibility of putting small base stations into sleep mode when there is no network load. Technologies like NOMA and MIMO are discussed as crucial components for achieving spectral efficiency and energy efficiency.
 Additionally, the article focuses on wireless sensor networks (WSNs) and suggests ways to optimize them for energy efficiency. This includes operating sensor units only when necessary, implementing wireless charging, using energy-efficient optimization methods, and applying efficient routing schemes.
 The authors also highlight the role of green data centers in reducing CO2 emissions and optimizing the use of green energy in high-performance networks. Methods such as using renewable energy sources, increasing the energy efficiency of hardware, and implementing energy-efficient routing are discussed.
 In conclusion, the article underscores the importance of energy efficiency and reduced CO2 emissions in modern multimedia networks, particularly in the context of 5G networks. It calls for interdisciplinary efforts to address these critical challenges in the field of information and communication technologies.

Hybrid cell handover strategy for O-RAN-based campus networks

In recent years, the concept of Open Radio Access Network (O-RAN) has become a prominent and ongoing research area in mobile communications, as mobile operators strive to enhance the intelligence, efficiency, and vendor independence of their RAN (Radio Access Network) architectures and components. O-RAN aims to virtualize and improve RAN functions in software and therefore it supports virtualized RANs where disaggregated network components of multiple vendors are connected via open interfaces. O-RAN is a concept based on interoperability and standardization of RAN elements including a unified interconnection standard for white-box hardware and open-source software elements from different vendors. Architecture integrates a modular base station software stack on off-the-shelf hardware which allows baseband and radio unit components from different suppliers to operate seamlessly together. Analyzing and examining the O-RAN architecture in terms of various network scenarios and different network policies is crucial for next-generation networks. This study begins by discussing the O-RAN architecture and its advantages in detail. Subsequently, a campus network has been proposed for the analysis and simulation of the O-RAN architecture, and performance analyses have been performed within the scope of different scenarios and network policies. As a result of these analysis, a new hybrid algorithm for cell handovers in the O-RAN architecture has been developed and compared with traditional cell handover methods based on Reference Signal Received Power (RSRP) and Quality of Service (QoS). The results indicate that the proposed hybrid handover algorithm for O-RAN architecture yields better outcomes.

A deep learning based dynamic bandwidth allocation method for XG-PON based mobile fronthaul for CRAN

In the context of 5G and beyond era, the use of Time Division Multiplexed Passive Optical Network (TDM-PON) for mobile fronthaul (MFH) in centralized/cloud radio access network (CRAN) has proven to be an optimal solution for addressing low-latency requirements. While existing literature has predominantly focused on optimizing mobile fronthaul latency rather than end-to-end latency, this paper introduces an end-to-end latency model for TDM-PON, specifically within a 10-gigabit Passive Optical Network (XG-PON) based MFH for CRAN. The paper then proposes an Intelligent Dynamic Bandwidth Allocation (DBA) scheme to minimize end-to-end latency of the network. The proposed scheme predicts the buffer occupancy reports using deep learning techniques at optical network units (ONUs). Thereafter, optical line terminal (OLT) schedules the available bandwidth using conventional DBA schemes (Group-GIANT i.e., Ggiant, Optimized Round Robin, Dynamic Service Interval). Primarily, the proposed DBA scheme transforms conventional DBA schemes (Ggiant, Optimized Round Robin, Dynamic Service Interval) into Intelligent versions (Intelligent Ggiant, Intelligent ORR, Intelligent DSI), showcasing a reduction of 28.63 %, 45.86 %, and 48.60 % in end-to-end latency in the XG-PON-based MFH for CRAN. Further, the analysis of obtained results has confirmed the supremacy of the Intelligent DSI DBA scheme over the Intelligent Ggiant and the Intelligent ORR DBA scheme.

Model-Based Deep Reinforcement Learning Framework for Channel Access in Wireless Networks

Model-Based Deep Reinforcement Learning Framework for Channel Access in Wireless Networks

Efficient User Association Strategy for Maximizing User Satisfaction and Resource Utilization in Heterogeneous Cloud Radio Access Networks (H-CRANs)

The primary aim of this study is to present a User Association Strategy that can effectively optimize the Heterogeneous Cloud Radio Access Network (H CRAN). The primary objective of this strategy is to enhance customer satisfaction by optimizing the utilization of available network resources, such as transmission power and bandwidth. To achieve this objective, the proposed approach employs a logarithmic barrier function to address inequality constraints and transform the optimization problem into a formulation that facilitates effective convergence towards the optimal solution.

Virtualization of a 4G Evolved Packet Core Network Using Network Function Virtualization (NFV) Technology with NS3 for Enterprise and Educational Purpose

The current networks of many operators dispose an increasing variety of purpose built, vertically integrated, and vendor-locked hardware equipment. This makes it difficult to easily scale their radio access and core networks as it requires yet another variety of equipment, leading to increased Time-to-Market and inefficient resource utilization. Furthermore, network equipment are expensive to procure and upgrade (increased CAPEX), and are difficult to adapt and program for new services (increased OPEX). These trends have recently spurred several efforts to redesign various components of mobile networks, notably the 4G Evolved Packet Core (EPC). With regards to this, the present work proposes a solution based on Network Function Virtualization (NFV), which can be deployed for educational or enterprises purpose like in Data Centers, or network nodes with a fine-grained QoS, while maintaining the scalability of the virtualized network function entities, and optimum resource utilization. The actual work develops a mobile network simulation module centered on the 4G Core network comprising its major components; MME, SGW, and PGW (vEPC), as well as the networking amongst these entities, core network related aspects such as its interaction with the Enhanced Radio Access Network (E-UTRAN) and the Packet Data Network Services (Internet). This solution gives to core network engineers and EPC network agents the possibility to design, analyze and test variable types of network scenarios. The solution can be also integrated in engineering schools and college for labs’ practical work or to any e-laboratory initiative to allow students in virtual environment to analyze signaling, modify the network configuration and better understand some theoretical concepts taught during the courses. The results obtained from the present work can be of great help during the final deployment phase of the network for full production for engineers of carriers’ network or improve the engineering understanding of core network in academic domain. In this light, the network can easily be scaled, adjust virtually for the time to commercialize a service reduced for carrier or enterprises’ solutions or for testing in e-laboratories.

Context-Based e2e Autonomous Operation in B5G Networks.

The research and innovation related to fifth-generation (5G) networks that has been carried out in recent years has decided on the fundamentals of the smart slice in radio access networks (RANs), as well as the autonomous fixed network operation. One of the most challenging objectives of beyond 5G (B5G) and sixth-generation (6G) networks is the deployment of mechanisms that enable smart end-to-end (e2e) network operation, which is required for the achievement of the stringent service requirements of the envisioned use cases to be supported in the short term. Therefore, smart actions, such as dynamic capacity allocation, flexible functional split, and dynamic slice management need to be performed in tight coordination with the autonomous capacity management of the fixed transport network infrastructure. Otherwise, the benefits of smart slice operation (i.e., cost and energy savings while ensuring per-slice service requirements) might be cancelled due to uncoordinated autonomous fixed network operation. Notably, the transport network in charge of supporting slices from the user equipment (UE) to the core expands across access and metro fixed networks. The required coordination needs to be performed while keeping the privacy of the radio and fixed network domains, which is important in multi-tenant scenarios where both network segments are managed by different operators. In this paper, we propose a novel approach that explores the concept of context-aware network operation, where the slice control anticipates the aggregated and anonymized information of the expected slice operation that is sent to the fixed network orchestrator in an asynchronous way. The context is then used as the input for the artificial intelligence (AI)-based models used by the fixed network control for the predictive capacity management of optical connections in support of RAN slices. This context-aware network operation aims at enabling accurate and reliable autonomous fixed network operation under extremely dynamic traffic originated by smart RAN operation. The exhaustive numerical results show that slice context availability improves the benchmarking fixed network predictive methods (90% reduction in prediction maximum error) remarkably in the foreseen B5G scenarios, for both access and metro segments and in heterogeneous service demand scenarios. Moreover, context-aware network operation enables robust and efficient operation of optical networks in support of dense RAN cells (>32 base stations per cell), while the benchmarking methods fail to guarantee different operational objectives.

An Intelligent Bayesian Inference Based Learning Automaton Approach for Traffic Management in Radio Access Network

An Intelligent Bayesian Inference Based Learning Automaton Approach for Traffic Management in Radio Access Network

A QoS Improving Downlink Scheduling Scheme for Slicing in 5G Radio Access Network (RAN)

A QoS Improving Downlink Scheduling Scheme for Slicing in 5G Radio Access Network (RAN)

Hierarchical Intelligent Radio Access Network Slicing for Differential Service Level Agreement Guaranteeing

Hierarchical Intelligent Radio Access Network Slicing for Differential Service Level Agreement Guaranteeing

The drift handling framework for open radio access networks: An experimental evaluation

The Open Radio Access Network (i.e., Open RAN) aims to transform the inflexible, proprietary RAN into one that is flexible, programmable, and disaggregated to provide intelligent solutions at the network edge for delivering the highly dynamic service demands of Beyond 5G (B5G) networks. Open RAN Working Group-2 (WG2) focuses on the architecture and standards of Artificial Intelligence/Machine Learning (AI/ML) to meet the Service Level Agreement (SLA) requirements for emerging use cases. The inherent dynamic nature of B5G networks and their applications often leads to AI/ML model performance degradation (i.e., drift), resulting in violations of SLAs, over- or under-provisioning of resources, etc.This paper proposes a drift handling framework which includes drift detection, analysis, and its adaptation to minimize the SLA violations and efficient resource utilization. The proposed drift handling framework is evaluated across three diverse use cases: (i) Quality of Service prediction, (ii) Channel Quality Indicator prediction, and (iii) User traffic prediction. The experimental evaluations are performed by using the Open RAN Software Community, and OpenAirInterface FlexRIC platforms with a real-time dataset. Results demonstrate that the proposed drift handling framework outperforms all other considered state-of-the art frameworks for all the scenarios in terms of the drift detection time, accuracy, precision, F1-score, etc.

Multi-Objective Multi-Dimensional Resource Allocation for Categorized QoS Provisioning in Beyond 5G and 6G Radio Access Networks

Multi-Objective Multi-Dimensional Resource Allocation for Categorized QoS Provisioning in Beyond 5G and 6G Radio Access Networks

Hierarchical Slicing: A New Paradigm of Radio Resource Management for Mobile Networks

Radio access network slicing is considered to be highly challenging due to the complexity of mobile environments and the diversity of mobile services. Existing works in this area mainly decompose the radio resource management problem into slice-level resource management and user-level resource scheduling. In this paper, we propose a hierarchical three-tier slicing architecture, where an additional tier is introduced to merge the gap between the conventional two tiers in time and spatial scales. Specifically, the medium-scale tier provides intercell and inter-slice coordination to address short-term dynamics caused by local user mobility and traffic variations, which highly simplifies the upper and lower tiers by allowing them to focus on long-term traffic distribution and instant user demands, respectively. Thus, it allows each independent iter to apply more flexible and more efficient radio resource management methods with appropriate time and spatial scales. A proof of concept is provided to show that the proposed three-tier architecture can achieve a better tradeoff between slice isolation and slice capacity as compared to the state-of-the-art two-tier solution.

Energy Sustainability in Dense Radio Access Networks via High Altitude Platform Stations

Energy Sustainability in Dense Radio Access Networks via High Altitude Platform Stations

Multistage Decimator Design for the Low-PHY PRACH Receiver

This paper describes the design of a multistage decimator for the Physical Random Access Channel (PRACH) receiver based on the hybrid time/frequency domain architecture. Flexible structure of the multistage decimator allows to support all combinations of PRACH subcarrier spacings and carrier bandwidths defined in the 3rd Generation Partnership Project (3GPP) Release 15 for Frequency Range (FR) 1. Impulse responses of low pass filters for each decimation stage do not depend on a selected combination allowing to facilitate hardware implementation. The multistage decimator jointly uses the Cascaded Integrator-Comb (CIC) and Nyquist filters to minimize number of required multipliers. Acceptable degradation level due to decimation in the receiver performance is estimated based on requirements defined by the Radio Access Network (RAN) working group 4. Assuming no specific detection algorithm, it is shown that the proposed design introduces negligible degradation, less than 0.1 dB in the operating Signal-to-Noise Ratio (SNR) region.

Queuing Network Models of Multiservice RANs

In this paper, we present a new queuing network model for the analysis of a portion of a radio access network (RAN) comprising macro cell base stations (BSs) and small cell BSs offering ”streaming” and ”elastic” services. Streaming services require a certain data rate for a random time. The required data rates depend on the type of service, e.g., audio and video. Elastic services require the transfer of random data volumes, and their data rate adjusts dynamically based on the capacity not utilized by the streaming services. To derive performance measures for the proposed model we develop a computationally efficient framework that exploits a new product form result for streaming services, relying on a well-known blocking policy, and an approximate product form for elastic services. Insensitivity to the distribution of service requirements holds in the case of negligible end user mobility. We show the high accuracy of our model in predicting the performance of practical system configurations by conducting a thorough comparison between the model’s results and those obtained from a detailed discrete-event simulator. Through this analysis, we uncover significant counter-intuitive behaviors that arise from the competition between streaming services with diverse demands, and that are effectively captured and predicted by our modeling approach. Our computationally efficient queuing model is a useful new tool to support design and planning of multiservice RANs whose complex structures result from the coexistence of BSs of different generations in dense areas.

Relay-empowered beyond 5G radio access networks with edge computing capabilities

Relevant services envisaged for beyond 5G (B5G) systems, such as extended reality and holographic communications, have extremely demanding user experience requirements with significant computational and communication demands. While edge computing aims to address the computation requirements by offloading the computational tasks to edge servers near the user, the communication will take advantage of the technologies developed for 5G New Radio jointly with an never-before-seen degree of network densification. This paper proposes the use of relays with edge computing capabilities. The approach's potential for B5G are identified, and a system model is defined to characterize both computational and communications viewpoints. Based on this, results are provided to highlight the gains and limitations of the proposed approach from a system-level perspective. Finally, the main challenges for enabling relays with computing capabilities in B5G deployments are discussed.

Scheduling of Industrial Control Traffic for Dynamic RAN Slicing with Distributed Massive MIMO

Industry 4.0, with its focus on flexibility and customizability, is pushing in the direction of wireless communication in future smart factories, in particular, massive multiple-input-multiple-output (MIMO) and its future evolution of large intelligent surfaces (LIS), which provide more reliable channel quality than previous technologies. At the same time, network slicing in 5G and beyond systems provides easier management of different categories of users and traffic, and a better basis for providing quality of service, especially for demanding use cases such as industrial control. In previous works, we have presented solutions for scheduling industrial control traffic in LIS and massive MIMO systems. We now consider the case of dynamic slicing in the radio access network, where we need to not only meet the stringent latency and reliability requirements of industrial control traffic, but also minimize the radio resources occupied by the network slice serving the control traffic, ensuring resources are available for lower-priority traffic slices. In this paper, we provide mixed-integer programming optimization formulations for radio resource usage minimization for dynamic network slicing. We tested our formulations in numerical experiments with varying traffic profiles and numbers of nodes, up to a maximum of 32 nodes. For all problem instances tested, we were able to calculate an optimal schedule within 1 s, making our approach feasible for use in real deployment scenarios.