Deployment and Performance Evaluation of 5G Private Networks, Enabling Use Cases in Rural Remote Areas

Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000. Digital Object Identifier 10.1109/ACCESS.2024.0429000 Deployment and Performance Evaluation of 5G Private Networks, Enabling Use Cases in Rural Remote Areas MARIA RAFTOPOULOU1, MUHAMMAD FAHEEM AWAN2, ALEJANDRO RAMÍREZ-ARROYO3, MEHMET IZZET SAGLAM4, LJUPCO JORGUSESKI1, FLORIS DRIJVER1, PAWEŁ MAĆKOWIAK1, and MARIA RITA PALATTELLA5 1Department of Networks, Netherlands Organisation for Applied Scientific Research (TNO), 2595 DA The Hague, The Netherlands 2Telenor Research and Innovation, 1360, Fornebu, Norway 3Department of Electronic Systems, Aalborg University (AAU), 9220 Aalborg Øst, Denmark 4Department of R&D, Turkcell Research and Development, 34854 İstanbul, Türkiye 5Luxembourg Institute of Science and Technology, L-4422 Belvaux, Luxembourg Corresponding author: Maria Raftopoulou (e-mail: [email protected]). This paper presents ongoing work carried out within the COMMECT project, which received funding from the European Union’s Horizon Europe Research and Innovation Programme under Grant Agreement No 101060881. ABSTRACT Rural remote areas have unique challenges for providing reliable and high-speed cellular connectivity, which makes them unattractive for mobile network operators. The Horizon Europe COMMECT project has addressed connectivity in rural remote areas by 5G private network deployments, which so far have mainly been studied and deployed in industrial environment. The evaluation of the 5G private network deployments is conducted through field trials in four different environment, each of them characterized by different challenges and aiming to support different agro-forestry use cases. In particular, the evaluations relate to (i) indoor environments (e.g. hospitals and educational centers), (ii) forestry, (iii) livestock transportation, and (iv) mines. The obtained results showcase that 5G private networks can support the proposed use cases. For example, in indoor environments, a downlink and uplink throughput of 700 Mbps and 50 Mbps, respectively, were measured, which are sufficient to enable e.g. remote monitoring of hospital patients. Moreover, it was measured that at least 10 Mbps in the uplink can be achieved for up to 300 m deep in the forest. The 10 Mbps target is also met in the farm location, including a satellite backhaul, which can enable e.g. the use case of (un)loading livestock from/to the truck. Additionally, throughput of more than 100 Mbps was measured at the mine location, which is sufficient to enhance safety by e.g. operating unmanned vehicles. INDEX TERMS 5G private networks, field trials, forestry, indoor environments, livestock transportation, mines, network deployment, remote areas. I. INTRODUCTION IN recent years, the fifth generation of mobile networks (5G) has seen wide deployment around the globe. Connetwork deployments are mostly focused on urban areas. Providing connectivity in rural areas is unattractive and challenging for MNOs due to the high investment needed for the infrastructure deployment and the low population density in those areas [2]. Thus, rural areas often face limited or poor connectivity, which in return impacts the residents and businesses, hampering economic development. To allow more flexibility on the network deployment and the supported services of the 5G network, the concept of 5G private networks or 5G non-public networks (NPNs) has been introduced and standardized within the third generation tinuing the trend of the previous generations of cellular networks, 5G further improves broadband services as well as enabling a range of new services by also offering ultra-low latency and massive connectivity [1]. Typically, these socalled public 5G networks are deployed and managed by mobile network operators (MNOs), which then offer their supported services to individuals or businesses. While MNOs provide a range of services, e.g. Internet access, and video streaming, through their 5G networks, their 1 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is the preprint version of the paper published in IEEE ACCESS (Early Access), DOI: 10.1109/ACCESS.2025.3603591 Raftopoulou et al.: Evaluation of 5G Private Networks for Enabling Use Cases in Remote Areas partnership project (3GPP) [3], [4]. 5G private networks are localized networks, which are deployed within the premises of the private entity e.g. an organization or an enterprise, and they are customized to the needs of the private entity. Moreover, 5G private networks are managed by the private entity or a third-party provider, e.g. an MNO, on their behalf. Due to the flexibility offered by 5G private networks, they have gained a lot of attention for deployment in industrial environments, which require enhanced security, and high communication availability [5], [6]. To that end, the 5G alliance for connected industries and automation (5GACIA) has identified different types of 5G private network deployments [7], and multiple deployment evaluations are performed e.g. in [8], [9]. Even though there is high interest on 5G private networks for industrial environments, these networks are also attractive to provide connectivity and new services in rural areas to support community development. For example, different use cases can be enabled such as smart agriculture, rural remote healthcare, smart energy, environment monitoring, among many others [2]. A. RELATED WORK Due to the recent conception and conceptualization of 5G private networks, several state-of-the-art works focus on simulations and theoretical models. For instance, Lim et al. [10] propose a deep reinforcement learning approach to optimize content allocation strategy and cache replacement policies in 5G private networks. Chang et al. [11] present an optimization algorithm to enhance the distribution and signal coverage of 5G base stations in mine deployments. Luo et al. [12] analyze the challenges associated with 5G private network design in the frequency range 2 band based on a game theoretic user allocation algorithm to minimize co-channel interference. In parallel, practical work related to different fields such as drone scenarios and cyber security is emerging. For instance, Urama et al. [13] present the deployment of 5G private networks based on drone scenarios. Neinavaie et al. [14] exploit reference signals in private networks for cognitive navigation on ground and aerial platforms. Skokowski et al. [15] employ a 5G private network to test the robustness of the technology against low-energy and smart jamming attacks. Akgun et al. [16] study a machine learning algorithm for predicting cell interference by exploiting the channel state information reported by a 5G private network. Uitto and Heikkinen [17] evaluate video streaming in standalone 5G networks by measuring the latency and jitter under different video and network configurations. However their evaluation scenario is performed indoors and with a short distance between the video equipment and the 5G modem, rather than in rural areas. Despite previous work, very limited works are available on practical deployment of 5G private networks in rural areas. Among them, Schellenberger et al. [18] focus on an agricultural use case, where a drone and a robot are connected through a 5G nomadic network to a remote server. Even though the system architecture is validated as feasible, no detailed results on throughput and latency have been reported. To address the challenges and enable new applications in rural areas, Mendes et al. [2] propose the 5G-RANGE network, which is evaluated with field demonstration. The designed 5G-RANGE network proposes new physical, medium access control and network layers that are specifically designed for rural areas. Thus, 5G-RANGE can be used as a different operation mode for beyond 5G standards. Differently, in our work we showcase how 5G private networks can be used to enable new applications in remote areas. In addition, several deployments are considered to provide an overview of the performance of these types of networks in rural areas. Section I.B describes in detail the main contributions and novelty of this study. B. CONTRIBUTIONS AND PAPER ORGANIZATION Within the COMMECT project the investigations have been focused on evaluating the possible connectivity solutions, in this case 5G private networks, for different use cases in remote areas [19]. The approach taken was via field trials that could reflect, to the extend possible, the actual conditions in these remote areas. Previously, in [20], two use cases in remote areas, namely forestry and livestock transportation, have been described. For each of the two use cases, a network deployment has been proposed and evaluated with field trials. However, the evaluation was limited to the experienced uplink throughput. This paper is an extension of our previously published work [20] and the main contributions can be summarized as follows: 1) Description of four use cases in remote areas that can be enabled by the deployment of 5G private networks, and their associated network deployment. In addition to the two previously studied use cases, the deployment of 5G private networks in indoor environments, for e.g. rural educational purposes, and in mines is also studied in this paper. 2) Illustration of the performance and limitations of 5G private network deployments through field trials. The extended field test evaluation results include coverage analysis, uplink and downlink throughput, latency, and uplink throughput performance for the case of satellite backhauling link. 3) A cost benefit analysis of the deployment of 5G private networks. In summary, this work is one of the first to report field trial-based evaluations across four different scenarios for rural use case applications, offering valuable insights for both the academic research community and engineers at commercial operators. The key performance indicators (KPIs) and metrics measured offer valuable real-world data for model validation in simulation, cross-comparison in real environments, and insights into the deployment of private 5G networks in rural environments. The remainder of this paper is organized as follows. In Section II the background of 5G private networks is presented 2 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Raftopoulou et al.: Evaluation of 5G Private Networks for Enabling Use Cases in Remote Areas TABLE 1. List of abbreviations. Abbreviation Explanation 3GPP 3rd Generation Partnership Project 5G-ACIA 5G Alliance for Connected Industries and Automation AI Artificial Intelligence AMR Autonomous Mobile Robot GPS Global Positioning System (I)IoT (Industrial) Internet of Things KPI Key Performance Indicator MIMO Multiple-Input Multiple-Output MNO Mobile Network Operator (N)LoS (Non-)Line of Sight NoW Network on Wheels NPN Non-Public Network PNI-NPN Public Network Integrated Non-Public Network RAN Radio Access Network RSRP Reference Signal Received Power SINR Signal to Interference and Noise Ratio SNPN Standalone Non-Public Network TCP Transmission Control Protocol TDD Time Division Duplex UDP User Datagram Protocol UE User Equipment including the major challenges from their deployment in remote areas. Section III presents the four use cases in remote areas that are under investigation and the associated network deployment. The performance evaluation of the proposed network deployments is presented in Section IV, whereas the relevant cost benefit analysis is presented in Section V. Finally, Section VI concludes the paper. For ease of reading, all abbreviations used across the sections are summarized in Table 1. II. 5G PRIVATE NETWORKS IN REMOTE AREAS This section provides background information on 5G private networks, its types of deployment, advantages, and challenges for deployment in rural remote areas. A. BACKGROUND health services, and remote monitoring, ensuring seamless communication and rapid response times. Finally, the larger local capacity would be very helpful as it could accommodate a vast number of connected devices simultaneously, ranging from smartphones and tablets to IoT sensors and autonomous equipment, meeting the diverse connectivity needs of the remote area communities. Another advantage is the high communication customization. 5G private networks are customized such that priority or high-quality is given to the limited set of services that are of crucial importance for the business or the enterprise. Another example is localization features (or in future sensing) that can potentially revolutionize precision farming, forestry, or remote indoor/mines practices. This feature combined with high-resolution video connectivity could also enable remote monitoring and teleworking. Furthermore, the 5G private network can be customized to seamlessly integrate IoT devices such as various sensors and actuators to satisfy the needs of the local area services and subscribers. B. TYPES OF DEPLOYMENT 5G private networks (or NPNs) can be deployed in different ways, depending on which part(s) of the network are shared (or not) with a public land mobile network [7], [21]. We identify the following two categories of private network (or NPN) deployments: Standalone non-public networks (SNPN): These private network deployments do not share any of their functions with the public network, which allows full control of the private network. SNPN deployments do not necessarily dictate that all network components are deployed at the enterprise premises. Different deployment flavours exist that can host some of the network functions outside the premises. Public network integrated non-public network (PNI-NPN): These private network deployments share some of their functions with one or more public networks. Different flavours of PNI-NPN deployments exist, depending on which functions are shared with or hosted by the public network(s) and where these network functions are hosted. For example, in [7], the ‘‘shared radio access network (RAN)’’ deployment is presented, where the private network shares its RAN with the public network. Moreover, a special case of PNI-NPNs are the public-network-hosted 5G private networks, also known as “virtual 5G private networks”, where all network functions are embedded in the public network. The two categories of 5G private network deployments are designed to accommodate different total cost of ownership and scaling/customization dependency on the MNO operating the public network. Consequently, some deployments have a low scaling/customization dependency on the public network, but a high total cost of ownership, with the extreme case of SNPNs having no dependency on the public network. On the other hand, the deployments that share many functions with the public network, have a low total cost of ownership but a high scaling/customization dependency on the public 5G private networks are intended for non-public use, meaning that their utilization is normally targeting specific services or use cases. Typically, 5G private networks allow access only to a limited set of devices (e.g. devices from employees and Internet of Things (IoT) devices of an enterprise) in a limited area (or number of areas). Additionally, due to data sensitivity and security reasons, a private 5G network can be organized such that both the communication data and/or operations and management data are kept within the enterprise premises. One of the main advantages of deploying 5G private networks in remote areas is that it allows for high performance connectivity. Reliable and high-speed connectivity is challenging in remote areas. Introducing 5G private networks could be a game-changer with several notable benefits. Firstly, residents and businesses could enjoy faster download and upload speeds, facilitating more efficient access to online services, video streaming, and large file downloads. Secondly, latency performance can be improved that is crucial for real-time applications like video conferencing, tele3 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Raftopoulou et al.: Evaluation of 5G Private Networks for Enabling Use Cases in Remote Areas network. For example, the extreme case of virtual 5G private networks are fully dependent on the public network. C. CHALLENGES For the deployment of 5G private networks in remote areas, different challenges need to be addressed, for example, acquisition of spectrum licenses for SNPNs, coverage for PNINPN with shared RAN, bandwidth size selection, and backhauling for SNPN. Some of these challenges are addressed in this paper through the performance evaluation of different 5G private network deployments to support use cases in remote areas. One of the challenges of 5G private networks is to identify the most appropriate deployment, i.e. SNPN or PNI-NPN, based on the requirements of the use case as well as the total cost of ownership, which includes both capital and operational expenditure. Typically, there is a trade-off between network performance and total cost of ownership. Therefore, a clear understanding of this trade-off is required per use case, as well as the definition of business models. This paper addresses this challenge by proving a cost benefit analysis in Section V. Moreover, the performance of the deployed 5G private network, depends on a number of parameters. For example, the network coverage should be large enough such that all devices in the intended area can communicate with the network. Typically a location is considered as ‘‘covered’’ by the network if the reference signal received power (RSRP) at that location exceeds a threshold. Therefore, to guarantee sufficient coverage and capacity, the location of the base stations, and the coverage optimization of the sectored antennas per allocated carrier frequency are essential. However, for PNI-NPN deployments, where the RAN (and thus, the base stations) of the public network is shared with the private network, sufficient coverage and capacity tailored to the localized ‘‘private’’ area of the particular enterprise could be challenging. This is due to the dependency on the public network’s RAN infrastructure that is potentially optimized to serving other ‘‘public’’ users outside the ‘‘private’’ area. On the contrary, with SNPNs, there is full control within the 5G private network in terms of base station location and antenna direction to achieve the required coverage and capacity. For instance, in dense environments with strong multipath components, such as indoor environments, signal blocking or non-line-of-sight (NLoS) can strongly affect the dynamic range and RSRP, leading to poor coverage [22]. In this paper, coverage results are presented for each of the deployed networks, covering both SNPNs and PNI-NPNs. Apart from coverage, the network performance also depends from the RAN configuration. In particular, in this paper, we address the challenges of identifying the most appropriate bandwidth size and time division duplex (TDD) frame configuration, which are directly related to the use case requirements. Both the bandwidth size and TDD frame configuration have a direct effect on the achievable uplink and downlink throughput, with the TDD frame configuration defining how many time slots are allocated to the uplink and downlink channels within a given time period. When deploying SNPNs, there are strict regulations in terms of 5G frequency license, which only allow for bandwidth sizes in a multiple of 10 MHz [23] and/or having a maximum allowed bandwidth size. Moreover, the license conditions include requirements in terms of TDD configuration, due to synchronization restrictions with other network operating in close proximity [24]. This could also restrict the performance of 5G private networks, especially for use cases where there is mostly uplink-oriented traffic. Both of this aspects are studied in this paper. Last but not least, the availability of backhaul links is also critical, as they are required for the connection of the 5G private network to the Internet. This can be especially difficult in remote areas, where the terrestrial infrastructure is limited or non-existent. A potential solution for backhauling is the use of satellite communications. However, the performance of satellite communications may still be limited, regardless of the many recent technological advancements. Considering that the backhaul could be the limiting factor in the performance of 5G private networks, it has been chosen for study in this paper. III. 5G PRIVATE NETWORK ENABLED USE CASES This section describes the use cases, the challenges of each use case, and their associated 5G private network deployments. The methodology and setups presented allow for the reproducibility or comparison of deployments under similar conditions, enabling future evaluations in similar contexts. A. INDOOR ENVIRONMENTS 1) USE CASES The concept of wireless factories has spread significantly in the last decade. Due to advances in cellular connectivity, highly reliable automated production lines can now be implemented according to the concepts introduced by Industry 4.0 and the industrial Internet of things (IIoT) [25]. Among the connectivity solutions proposed to provide high reliability in wireless environments are 5G private networks due to dedicated spectrum and infrastructure [26]. Among the use cases, we highlight the massive connection of sensors with IoT devices for real-time monitoring, allowing maintenance work and efficient resource management. Additionally, this connectivity can be extended to the control of autonomous mobile robots (AMRs) in charge of logistics in factories [27]. Finally, this highly reliable connectivity could enable augmented reality and digital twins for remote assistance and control in the production chain. This first use case, focused on indoor environments, is presented as a baseline for the performance of a private 5G network compared to outdoor environments (see following use cases). Establishing this indoor baseline is essential for subsequent comparisons with deployments in more challenging rural or remote environments, where factors such as propagation, interference, and infrastructure limitations can 4 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Raftopoulou et al.: Evaluation of 5G Private Networks for Enabling Use Cases in Remote Areas significantly impact performance. From a high-level perspective, a 5G private indoor network can offer multiple use cases beyond those already explained in industrial environments, especially in rural areas where indoor environments like greenhouses or local healthcare and education centers can benefit greatly from advanced connectivity. Some examples are: (i) automation and sensor networks in agro-forestry applications (e.g., smart greenhouses requiring reliable lowlatency connections for precise climate control or automated irrigation), (ii) connectivity in healthcare centers and hospitals for remote monitoring or remote surgery, which have critical ultra-low latency requirements, or (iii) connectivity in education centers where virtual and augmented reality classes are held, which may require high bandwidths [28]–[30]. Given the previous use cases, this paper analyzes an indoor 5G private network to assess the coverage and performance in terms of latency and throughput that such a deployment can provide in a realistic indoor scenario such as an industrial environment. 2) NETWORK DEPLOYMENT (a) (b) FIGURE 1. (a) Photograph of the SIMCOM modem and the next unit of computing (NUC) assembled on the AMR, and the 5G SA core located in the 5G SmartLab, and (b) scenario through which the AMR navigates. AirScale indoor radio (ASiR) is illustrated on the roof of the factory. sitioning within the laboratory, the modem stores data related to network performance and signal quality. Therefore, by synchronizing the timestamps from both ends it is possible to accurately map the network performance parameters at each location in the industrial scenario. The analysis of the radio propagation characteristics and performance of the 5G private indoor network is presented throughout Section IV. B. FORESTRY 1) USE CASES In Norway, forests cover roughly 33% of the country’s land area, making them a crucial part of both the economy and environmental management efforts. Sustainable forest management is central to Norway’s strategy, supporting the longterm vitality of the forests while enabling a wide range of economic activities. However, the forestry sector in Norway faces a significant challenge in terms of digitalization. The The AAU 5G Smart Production Lab is a 1200 m2 research laboratory located at Aalborg University (Denmark), which emulates the physical conditions of an industrial environment. Therefore, these facilities emulate an industrial setting with multiple production lines, robotic arms or autonomous mobile robots. Due to the research nature of the environment, there are multiple connectivity solutions deployed such as MULTEfire, Wi-Fi 6 or private 4G and 5G networks. In the framework of this work, a 5G private network deployed in the lab has been assessed. This private network is connected to the public core of a Danish operator through a dedicated core-RAN over a fiber connection. This solution, deployed in the 3.7 GHz in frequency range 1 n78 band, has a bandwidth of 100 MHz, a TDD frame configuration of 3 uplink and 7 downlink slots, and a subcarrier spacing configuration of 30 kHz. The private network hardware deployment consists of a Nokia Mxie 5G SA core and a Nokia AirScale indoor radio unit, which can be visualized in Figs. 1(a) and 1(b), respectively. The indoor radio unit, which acts as an access point, provides coverage to one of the halls of the facility with an approximate area of 400 m2, and is located at a height of 5 m. To analyze the performance of the described private network, a SIMCOM SIM8380G-M2 modem with support for 5G NSA/SA in the network operating band is used as user equipment (UE). Four antennas are connected to this UE to take advantage of the modem’s multiple-input multipleoutput (MIMO) capabilities, as illustrated in Fig. 1(a). To monitor UE operations, the UE is connected to a small computing box (next unit of computing) from which ping and iperf commands run to evaluate network performance. Both the modem and the computing box are placed on an AMR with the ability to navigate the 5G SmartLab following predefined routes due to a LiDAR-based positioning system (see Fig. 1(a)). While the AMR stores data related to its po5 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Raftopoulou et al.: Evaluation of 5G Private Networks for Enabling Use Cases in Remote Areas sector’s value chain has not yet fully embraced technological innovations, particularly with regard to connectivity solutions [31]. As Norway continues to foster innovation across various sectors, there is an opportunity to boost the forestry industry’s resilience and sustainability by adopting digital transformation and keeping pace with technological developments. The COMMECT project has identified two key use cases where 5G private networks could play a pivotal role, as depicted in Fig. 2. Remote Operational Support for Forest Machinery Operators: The forestry industry faces a critical need for remote expert assistance during thinning and logging activities, where operators in the field require immediate guidance and support. This challenge is compounded by the limited coverage and performance of public mobile networks in forested regions. To address this, the proposed solution involves deploying a mobile 5G private network on wheels (NoW), specifically designed for forestry environments, as shown in Fig. 2. This 5G private NoW facilitates high-quality video transmission from forestry machinery, enabling real-time remote assistance through online cameras. Experts from the forest operator’s company can guide on-siteoperators in making informed decisions. Additionally, this solution facilitates remote maintenance, diagnostics, and repairs of machinery, with vendor experts providing supervision via high-definition video feeds. Enhanced Situational Awareness in Forests: The increasing risk of forest fires, exacerbated by climate change, underscores the need for a sophisticated monitoring and surveillance system in forested areas. Forests are vulnerable to a range of emergencies, including wildfires, landslides, and floods, necessitating advanced solutions to ensure the safety of both the environment and emergency responders, including police, ambulances, fire departments, and volunteers. Current systems require rapid reporting of incidents, such as fires, through digital channels. In response, the NoW solution shown in Fig. 2, integrated with drone technology and onthe-ground sensors, offers a comprehensive monitoring and surveillance platform. This system enables the swift notification of emergency personnel during critical incidents. High-resolution, multi-spectral cameras mounted on drones capture aerial imagery, which is transmitted in real-time via the private 5G NoW to an external edge server. The inclusion of on-map analytics further enhances the ability to visualize forest conditions, ensuring timely and effective responses to any alarming changes. 2) NETWORK DEPLOYMENT The COMMECT forestry use cases require substantial uplink throughput to support the transmission of high-quality video streams from remote forestry machinery and drones. To meet this demand, the concept of a 5G private network, specifically the NoW, has been introduced. The NoW functions as a self-contained system, integrating 5G radio, 5G core, and related applications within a single, highly mobile platform, as depicted in Fig. 3(a). This configuration enables the rapid deployment of a standalone 5G private network FIGURE 2. Depiction of forestry use cases. that delivers seamless 5G coverage. The traffic generated by NoW is backhauled through either terrestrial networks or satellite links, depending on the operational requirements and location, allowing remote access to cloud services or the internet for expert consultations. To enhance real-time forest monitoring and situational awareness, the NoW’s Edge server can be equipped with artificial intelligence (AI) and machine learning algorithms to detect and respond to incidents such as forest fires or other anomalies efficiently. To evaluate the suitability of NoW for forestry applications, multiple tests were performed with the NoW, providing both 5G radio and 5G core functionalities, deployed in a selected forest location. The tests were conducted using 80 MHz of bandwidth in the 3.3–3.4 GHz frequency range, with a transmit power of 35 dBm and an antenna height of 2.5 m. An uplink-heavy TDD frame configuration of 3 uplink and 7 downlink slots was utilized to prioritize uplink performance. The measurement locations within the forest are shown in Fig. 3(b). Due to the dense forest canopy, measurements could not be taken farther from the NoW deployment point. The measurements were conducted using a Huawei P40 smartphone, acting as a UE, held at a typical natural height of 1.5 m. The tests employed the web-based OpenSpeedTest tool1, which was hosted on the NoW’s Edge server. This tool measures uplink and downlink throughput, and latency, directly through a browser, without the need for additional software. These measurements are performed using multiple transmission control protocol (TCP) streams. The distance between the NoW and the UE was calculated using global positioning system (GPS) coordinates, with an accuracy of approximately 3 m. Notably, no backhaul was involved in this phase of testing, as the focus was on evaluating the performance of the 5G private network itself. The performance evaluation is presented in Section IV. C. LIVESTOCK TRANSPORTATION 1) USE CASES One of the major industries in Denmark is the pig industry, which is highly regulated by both the European Union and the Danish authorities to ensure the welfare of animals before, during and after their transport [32], [33]. Within the COMMECT project, discussions have been taken place 1https://openspeedtest.com/selfhosted-speedtest 6 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Raftopoulou et al.: Evaluation of 5G Private Networks for Enabling Use Cases in Remote Areas FIGURE 3. (a) Network on Wheels (NoW), and (b) measurement points in a dense forest with high vegetation, where the base station marks the location of the NoW. 2) NETWORK DEPLOYMENT To support the livestock transportation use cases, the connectivity solution in Fig. 4 is proposed. Specifically, local connectivity at the (un)loading location is provided with a 5G private network, which is then connected to the remote server via terrestrial or satellite Internet access. The choice of backhaul depends on the availability of terrestrial networks at the (un)loading location, which is typically in rural areas with little or no cellular coverage. In this work, and in relation to the livestock transportation use case, we assess the performance of a 5G private network in terms of coverage and uplink throughput. Moreover, we assess the performance of the backhaul via satellites, as the backhaul can significantly limit the experienced end-to-end uplink throughput. For our evaluations, we performed field tests at a farm location in The Netherlands, where the Amarisoft CallBox Classic2, illustrated in Fig. 5(a), was used to act as an SNPN, providing both 5G radio and 5G core functionalities. The Amarisoft was configured in the n77 band, in particular between 3.825 GHz and 3.875 GHz and its maximum transmission power was -20 dBm/MHz. For the measurements two different bandwidth sizes were considered, i.e. 20 MHz and 50 MHz. Moreover, two different TDD frame configurations were considered. The first TDD configuration is typically recommended for mostly downlink oriented traffic and consists of 3 downlink slots, 1 special slot and 1 uplink slot [24]. The second TDD configuration is an uplink heavy configuration as it consists of 5 downlink slots, 2 special slot and 3 uplink slots during a 10 slots frame [24]. To perform the relevant coverage and uplink throughput measurements, the Quectel RM502Q-GL has been used to act as a UE at different locations at the farm, which are illustrated in Fig. 5(b). For the assessment of the backhaul link, in terms 2https://www.amarisoft.com/test-and-measurement/devicetesting/device-products/amari-callbox-classic with stakeholders from the livestock transportation industry and two potential use cases have been identified that will be beneficial to the industry. Automatic license plate recognition: For a transportation truck to be allowed to enter a piglet (un)loading location, a check on the license plate number should be performed. This process is important as trucks need to go through a disinfection process before being allowed at a (un)loading location and the truck’s clearance is registered to a database, coupled to their license plate number. As a solution, it is envisioned that a camera will be installed on and control the road barrier leading to the (un)loading location. The camera will upload the image of the license plate to a remote server, which is connected to the database and where the relevant software runs. Automatic animal counting and monitoring: The animals are being counted during their (un)loading from/to the truck, and thus an efficient and correct animal counting is important. Moreover, during the (un)loading process, it is beneficial to monitor and detect the status of the animals, e.g. injuries, to ensure that their health status remained unchanged during their transport. The envisioned solution includes the deployment of a camera at the (un)loading location, which will live stream a video of the (un)loading process to a remote server, where the relevant software will run. The envisioned deployment solutions for both use cases is similar i.e. deployment of a camera that will upload an image/video to a remote server. From a communications point of view, sufficient uplink throughput, and thus coverage, should be ensured both at the road barrier and at the (un)loading location. In terms of uplink throughput, the 10 Mbps [34] have been identified as a minimum requirement, whereas for coverage, typically, an RSRP target of -120 dBm is chosen. 7 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Raftopoulou et al.: Evaluation of 5G Private Networks for Enabling Use Cases in Remote Areas FIGURE 4. Livestock transportation use case with a 5G private network. of uplink throughput, the Starlink Standard V3 terminal3has been connected to the Amarisoft CallBox to provide connectivity between the UE and a server located in Amsterdam. For all uplink throughput measurements, the iPerf34has been used in user datagram protocol (UDP) mode, which is a commonly used protocol for video streaming codecs due to its latency characteristics and packet loss tolerance. The results on coverage, uplink throughput, and backhaul are presented in Section IV. D. MINES 1) USE CASES Open-pit or opencast mine extracts minerals from the surface instead of underground. Open-pit mining is the most common method used in Türkiye for mineral mining and does not require extractive methods or tunnels. This surface mining technique is used when minerals or deposits are relatively close to the surface of the earth. Open pits are sometimes called quarries when building materials and dimension stones are produced. Open-pit mines are dug on different benches depending on the excavation machinery. The walls of the open-pit mines are dug at an angle and include steps to prevent avalanches from occurring inside the building site. These walls in the openpit mines are dangerous for the operators of the excavators and drivers of the heavy dump trucks. The goal is to decrease accidents and mistakes in the risky environment of Türkiye’s open-pit mines with unmanned collaborative vehicles connected to a private network. The private network solution is developed for automated and teleoperated vehicles in open-pit mines located in extremely suburban areas. With the support of 5G infrastructure, AI-supported central task planning, and smart sensors, unmanned collaborative vehicles will be developed and operated for excavation works. 3https://www.starlink.com/specifications 4https://iperf.fr/ 2) NETWORK DEPLOYMENT With the scalability of the 5G open-pit mine private network architecture, two experience zone activations have been carried out at high-profile events, where participants were able to experience the control of excavator from a remote location 800 km away, with a latency of one-tenth of the blink of a human eye. In addition to enabling 5G standalone edge network in the open-pit mine private network, the quality of services is guaranteed by AI-based resource optimisation of network slicing technology. The open-pit mine 5G private network solution is covered by five road side units and two fibre base stations with advanced cellular vehicle-to-everything use cases that AI on the edge applications and 5G standalone edge network scalable architecture. The open-pit mine private network deployment uses 100 MHz of bandwidth in the 3.5–3.6 GHz frequency range, with a transmit power of 55 dBm and an antenna height of 18 m. The deployment of five road site units is shown in Fig. 6. The proposed private network aims to enable 5G connectivity along with an edge network architecture that provides AIassisted perceived zero latency and ultra-reliability in openpit mining areas. The private network deployment includes 5G radio equipment mounted on the base stations in the demo location, the transformation of 5G sites into 5G fibre sites, and conducting fault tests. AI-assisted perceived zero latency and ultra-reliability algorithm was developed with the data collection platform designed and located at open-pit mine private network. This platform was also designed to support custom data generation. The performance evaluation of the proposed private network is presented in Section IV. IV. PERFORMANCE EVALUATION This section presents the performance evaluation of the proposed 5G private networks to support the previously described use cases. The results are presented per KPI rather than per use case to more easily showcase the difference in performance between the different network deployments on a specific KPI. Moreover, network deployments have only been evaluated for the KPIs that are relevant for the associated use cases. Therefore, not all network deployments are compared for each KPI. The KPIs that are under investigation are coverage (in terms of RSRP measurements), uplink throughput, downlink throughput, latency, and uplink throughput for a satellite backhaul. A. COVERAGE Coverage is an important KPI for all network deployments as it defines the area where the UEs can connect to the network. Typically, to determine the coverage area, RSRP measurements are performed at different locations, based on a reference signal transmitted by the network. The RSRP, and hence the coverage, depend among others on the transmit power, antenna gains, and the propagation conditions in the environment. Therefore, for each network deployment, the RSRP values over different distances are presented. Note that 8 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Raftopoulou et al.: Evaluation of 5G Private Networks for Enabling Use Cases in Remote Areas FIGURE 5. (a) Measurement setup with Amarisoft CallBox Classic and (b) measurement locations at the farm. FIGURE 6. Network deployment of open-pit mine private network. typically a location is considered as covered when the RSRP exceeds -120 dBm. 1) Indoor Environments where the signal drops to −80 dBm due to the blocking of the signal given the machinery and racks present in the scenario. To analyze in more detail the behavior of the RSRP over distance, Fig. 8(a) shows the relation between both parameters, where a decrease of the RSRP is observed as the distance increases due to propagation attenuation. To quantitatively evaluate the decrease in RSRP, the path gain can be modeled simply through a slope-intercept model which accounts for the path loss exponent n[35]. Considering the linear relation between RSRP and path gain given the fixed transmit power and the omnidirectional gain of the antennas in UE and AP, RSRP curves in terms of distance can be established for different values of path loss exponent n. Therefore, radio propagation conditions can be determined taking into account a loss exponent n= 2 for free space as a baseline. Referring to Fig. 8(a), it can be observed that the scatter plot tends to converge to values close to the curves denoted by n= 3 and n= 4. These values are similar to those observed in industrial environments with high machinery density in the state of the art [22], [36]. Finally, it is worth noting the region located within the range of 10 to 13 m, where a sharp drop in RSRP is observed. This region corresponds to the previously described area that experiences total signal blockage, relying on signal reflections and diffractions in the environment rather than on LoS. 2) Forestry Fig. 8(b) shows the RSRP values against the varying distance inside the deep forest. The RSRP measurements in a deep forest environment demonstrate a clear attenuation of signal strength with increasing distance from the NoW location. At a short distance of 15 m, the RSRP is strong at -35 dBm, indicating minimal signal loss. However, as the distance increases, the signal strength diminishes significantly. At 133 m, the RSRP drops to -68 dBm, highlighting the impact of dense foliage on signal propagation. Beyond 200 m, the signal degradation becomes more pronounced, with RSRP values reaching -87 dBm at 207 m and -98 dBm at 286 m. These measurements underline the challenging nature of Given the setup explained in Section III.A and illustrated in Fig. 1, we evaluated the indoor coverage as the strength signal received along the hall in the industrial scenario. For this purpose, we proceed to navigate a route with the AMR in which the UE measures the RSRP level at different locations in the hall. Fig. 7(a) shows the range between the UE−access point link given the different AMR locations within the lab. This range varies between 8 m when the AMR is in the vicinity of the access point and 22 m when the AMR navigates the northern part of the lab. Notably, there is line-of-sight (LoS) between the UE and the access point in most locations of the hall, except in the western area, where a rack with heavy machinery blocks the link. Fig. 7(b) shows the RSRP acquired on the navigated route. Note that this value is calculated as the average received signal in the synchronization signal blocks that the network transmits with 20 ms periodicity. Regarding the south location of the hall, the RSRP reaches values of around −60 dBm due to the LoS condition, and the short range of the link. In the northern part, these values decrease to values below −70 dBm due to the larger link range. It is notable that in the western part, there is an abrupt transition 9 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Raftopoulou et al.: Evaluation of 5G Private Networks for Enabling Use Cases in Remote Areas proposed to address the connectivity challenges in remote areas and enable new use cases. In particular, use cases within four different environments have been considered, namely indoor environments (for e.g. healthcare or educational centers), forestry, livestock transportation, and mines. The proposed network deployments are diverse in terms of hardware (and thus cost) and configuration, and they have been evaluated via field trials. The results have shown that the deployed networks can support the presented use cases and that there is a clear trade-off between performance and cost. This highlights the need for designing 5G private networks based on the use case and the respective requirements, which could potentially lead to new business models and opportunities within the telecom sector. Therefore, this work has performed an empirical assessment and validation across challenging and underexplored remote scenarios, which are essential for understanding the feasibility and limitations of 5G private networks in rural contexts. For future work, we propose to study the performance of other technologies which could potentially support the use cases, e.g. Wi-Fi. Then, a more in-depth cost benefit analysis will provide further insights on the applicability of 5G private networks, the opportunities that they bring and their challenges. Additionally, further analysis is recommended on the backhaul performance, as it can significantly limit the 5G private network performance. Apart from satellite communications for the backhaul, an alternative worth investigating is fixed wireless access. Finally, it is recommended to perform end-to-end measurements to investigate the impact of all the components in the network and not only the performance of the 5G private network. REFERENCES [1] S. Ahmadi, ‘‘5G NR: Architecture, technology, implementation, and operation of 3GPP new radio standards,’’ 1st ed. Academic Press, 2019. [2] L. L. Mendes et al., ‘‘Enhanced remote access communications: The missing scenario for 5G and beyond 5G networks,’’ IEEE Access, vol. 8, pp. 219859 - 219880, 2020. [3] M. Wen et al., ‘‘Private 5G networks: Concepts, architectures, and research landscape,’’ IEEE Journal of Selected Topics in Signal Processing, vol. 16, no. 1, pp. 7-25, 2022. [4] J. Prados-Garzon, P. Ameigeiras, J. Ordonez-Lucena, P. Muñoz, O. Adamuz-Hinojosa and D. Camps-Mur, ‘‘5G non-public networks: Standardization, architectures and challenges,’’ IEEE Access, vol. 9, pp. 153893-153908, 2021. [5] A. Mahmood et al., ‘‘Industrial IoT in 5G-and-beyond networks: Vision, architecture, and design trends,’’ IEEE Transactions on Industrial Informatics, vol. 18, no. 6, pp. 4122-4137, 2022. [6] A. Aijaz, ‘‘Private 5G: The future of industrial wireless,’’ IEEE Industrial Electronics Magazine, vol. 14, no. 4, pp. 136-145, 2020. [7] 5G-ACIA, ‘‘5G non-public networks for industrial scenarios,’’ White Paper, Sep. 2021. [Online.] Available: https://5gacia.org/media/2021/04/WP_5G_NPN_2019_01.pdf. [8] X. Li et al., ‘‘Multi-domain solutions for the deployment of private 5G networks,’’ IEEE Access, vol. 9, pp. 106865-106884, 2021. [9] A. Karaagac et al., ‘‘Managing 5G non-public networks from industrial automation systems,’’ in Proceedings of IEEE 19th International Conference on Factory Communication Systems (WFCS), 2023. [10] J. Lim, D. Kim and Y. Yoo, ‘‘Joint cache allocation and replacement for content-centric network-based private 5G networks: Deep reinforcement learning approach,’’ IEEE Access, vol. 12, pp. 56214-56225, 2024. [11] Z. Chang, Q. Gu, C. Lu, Y. Zhang, S. Ruan and S. Jiang, ‘‘5G private network deployment optimization based on RWSSA in open-pit mine,’’ IEEE Transactions on Industrial Informatics, vol. 18, no. 8, pp. 5466-5476, 2022. [12] H. Luo, A. Bishnu and T. Ratnarajah, ‘‘Design and analysis of in-band fullduplex private 5G networks using FR2 band,’’ IEEE Access, vol. 9, pp. 166886-166905, 2021. [13] J. Urama et al., ‘‘UAV-aided interference assessment for private 5G NR deployments: Challenges and solutions,’’ IEEE Communications Magazine, vol. 58, no. 8, pp. 89-95, 2020. [14] M. Neinavaie, J. Khalife and Z. M. Kassas, ‘‘Cognitive opportunistic navigation in private networks with 5G signals and beyond,’’ IEEE Journal of Selected Topics in Signal Processing, vol. 16, no. 1, pp. 129-143, 2022. [15] P. Skokowski et al., ‘‘Practical trial for low-energy effective jamming on private networks with 5G-NR and NB-IoT radio interfaces,’’ IEEE Access, vol. 12, pp. 51523-51535, 2024. [16] B. Akgun et al., ‘‘Interference-aware intelligent scheduling for virtualized private 5G networks,’’ IEEE Access, vol. 12, pp. 7987-8003, 2024. [17] M. Uitto, and A. Heikkinen, ‘‘Evaluating 5G uplink performance in low latency video streaming,’’ in Proceedings of Joint European Conference on Network and Communications (EuCNC) and 6G Summit: Operational and Experimental Insights, 2022. [18] C. Schellenberger, C. Hobelsberger, B. Kolb-Grunder, F. Herrmann, and H. D. Schotten, ‘‘Leveraging 5G private networks, UAVs and robots to detect and combat broad-leaved dock (Rumex obtusifolius) in feed production,’’ Mobile Communication - Technologies and Applications; 27. ITGSymposium, 2023. [19] Horizon Europe COMMECT project. 2025. [Online.] Available: https://www.horizoneurope-commect.eu/ [20] M. F. Awan et al., ‘‘5G private networks for rural communities: Use cases and field trials,’’ in Proceedings of IEEE Wireless Communications and Networking Conference (WCNC), 2025. [21] 5G-PPP, ‘‘Non-public-networks – State of the art and way forward,’’ White Paper, 2022. [Online.] Available: https://zenodo.org/records/7230191. [22] A. Ramírez-Arroyo et al., ‘‘FR2 5G networks for industrial scenarios: Experimental characterization and beam management procedures in operational conditions,’’ IEEE Transactions on Vehicular Technology, vol. 73, no. 9, pp. 13513-13525, 2024. [23] L. M. Bartolín-Arnau, J. Vera-Pérez, V. Sempere-Payá, and J. SilvestreBlanes, ‘‘Are European initiatives related to local spectrum allocation for private 5G networks ready for use in industrial cases?,’’ IEEE Access, vol. 12, pp. 19887 - 19901, 2024. [24] GSMA, ‘‘5G TDD synchronisation,’’ 2020. [Online.] Available: https://www.gsma.com/connectivity-for-good/spectrum/wpcontent/uploads/2020/04/3.5-GHz-5G-TDD-Synchronisation.pdf. [25] L. Chettri and R. Bera, ‘‘A comprehensive survey on Internet of Things (IoT) toward 5G wireless systems,’’ IEEE Internet of Things Journal, vol. 7, no. 1, pp. 16-32, 2020. [26] M. Wen et al., ‘‘Private 5G networks: Concepts, architectures, and research landscape,’’ IEEE Journal of Selected Topics in Signal Processing, vol. 16, no. 1, pp. 7-25, 2022. [27] D. S. Fowler et al., ‘‘A 5G automated-guided vehicle SME testbed for resilient future factories,’’ IEEE Open Journal of the Industrial Electronics Society, vol. 4, pp. 242-258, 2023. [28] P. I. Tebe et al., ‘‘5G-Enabled medical data transmission in mobile hospital systems,’’ IEEE Internet of Things Journal, vol. 9, no. 15, pp. 13679-13693, 2022. [29] J. Rischke, P. Sossalla, S. Itting, F. H. P. Fitzek and M. Reisslein, ‘‘5G Campus networks: A first measurement study,’’ IEEE Access, vol. 9, pp. 121786-121803, 2021. [30] B. Kizilkaya, G. Zhao, Y. A. Sambo, L. Li and M. A. Imran, ‘‘5G-Enabled education 4.0: Enabling technologies, challenges, and solutions,’’ IEEE Access, vol. 9, pp. 166962-166969, 2021. [31] P. J. Nesse, N. V. Nielsen, S. Davidson, L. Gillund, and M. F. Awan, ‘‘5G in rural forest enables real time decision support and new remote operation solutions,’’ in Proceedings of 3rd International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME), 2023. [32] ‘‘Council Regulation (EC) No 1/2005,’’ 2008. [Online]. Available: https://eur-lex.europa.eu/legalcontent/en/ALL/?uri=CELEX%3A32005R0001 [33] ‘‘Danish agriculture and food council.’’ [Online]. Available: https://pigresearchcentre.dk/DANISH-quality-assurance-scheme/- /media/D5F89185691F4D3EB282A0F1312DC9EB.ashx [34] Horizon Europe COMMECT project, ‘‘Deliverable D1.2: COMMECT requirements and KPIs,’’ Tech. Rep., 2024. 16 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Raftopoulou et al.: Evaluation of 5G Private Networks for Enabling Use Cases in Remote Areas [35] A. Ramírez-Arroyo, T. B. Sørensen, P. Beltoft, H. Christiansen, J. F. Valenzuela-Valdés and P. Mogensen, ‘‘Observations on Large-Scale Attenuation Effects in a 26 GHz Urban Micro-Cell Environment,’’ IEEE Wireless Communications Letters, vol. 13, no. 9, pp. 2611-2615, 2024. [36] T. Jiang et al., ‘‘3GPP standardized 5G channel model for IIoT scenarios: A survey,’’ IEEE Internet of Things Journal, vol. 8, no. 11, pp. 8799-8815, 2021. [37] D. Ebehard and E. Voges, ‘‘Digital single sideband detection for interferometric sensors,’’ in Proceedings of 2nd International Conference on Optical Fiber Sensors, Stuttgart, Germany, Jan. 2-5, 1984. [38] L. Kehoe. ‘‘Starlink shines in Europe as constellation investments boost performance.’’ ookla.com. 2025. [Online.] Available: https://www.ookla.com/articles/starlink-europe-q1-2024. [39] EU4Digital, ‘‘5G private networks development; EU best practice report,’’ 2024. [Online.] Available: https://eufordigital.eu/wpcontent/uploads/2024/03/EU4D-Telecom-Rules-Report-on-5G-privatenetwork-development-1.pdf. MARIA RAFTOPOULOU received the MEng degree in Electrical and Computer Engineering from the National Technical University of Athens, Greece, in 2016 and the MSc degree cum laude in Telecommunications and Sensing Systems from the Delft University of Technology, The Netherlands, in 2018. She also received her PhD degree at the Delft University of Technology, in 2024. Between 2018-2019, she worked as a Technology Young Talent at KPN, The Netherlands. She is currently working as a Scientist Innovator at the Netherlands Organisation for Applied Scientific Research (TNO), The Netherlands. Her research interests include radio resource management and optimisation over wireless networks. MUHAMMAD FAHEEM AWAN is working as Research Scientist at Telenor Research and Innovation Unit, Fornebu Norway. He has received his PhD degree in Electronics and Telecommunications from Norwegian University of Science and Technology, NTNU, Norway, in 2020. His current research interests include 5G and Beyond 5G radio access technologies, advance 5G/B5G use cases, 5G Positioning and mission critical communications. MEHMET IZZET SAĞLAM received his B.S. degree in electrical and electronics engineering from Çukurova University in 2001, and his M.S. and PhD degrees in wireless communication from Istanbul Technical University, in 2003 and 2018, respectively. His executive MBA degree is from Yıldız Technical University. He is a researcher and 3GPP RAN WG2 delegate at Turkcell R&D. His current research interests include 5G Advanced and 6G radio access networks. LJUPCO JORGUSESKI graduated (Dipl. Ing., 1996) at the University Cyril and Methodius, Skopje, Republic of Macedonia , and obtained a Ph.D. degree (2008) from the Aalborg University, Denmark. From 1997 to 1999 he worked as applied researcher at TU Delft followed by a research position on wireless 3G/4G systems at KPN Research, The Netherlands, till 2003. Since 2003 he is a senior consultant on wireless access at TNO focusing on radio planning and (self-)optimization of wireless networks, and 3GPP and O-RAN Alliance standardization activities. He has (co-)authored over 30 publications in conferences, journals, book chapters and patent applications. FLORIS DRIJVER was born in Leiderdorp, The Netherlands on 20th of February 1994. He received his B.S.E. degree in Electrical Engineering from Delft University of technology, and the MSc. degree in Electrical Engineering: Telecommunications and Sensing systems also from Delft University of Technology. From 2017 he is with the Networks department of TNO (Netherlands Organisation for Applied Scientific Research) as a Senior Scientist. His focus has been on 4G and later 5G and 6G field labs. He is currently the technical lead of the Future Network Services (FNS) National 6G testbed. PAWEŁ MAĆKOWIAK was born in Poznań, Poland on 14th of September 1995. He received the B.S.E. degree in Electronics and Telecommunications from Poznan University of Technology, and the MSc. degree in Electrical Engineering from Delft University of Technology. From 2019 to 2025, he was with the Networks department of TNO (Netherlands Organisation for Applied Scientific Research) as a Research Scientist. He is currently a Field Software Engineer in the Data Center Enterprise team of Canonical. His research interests include 5G/6G and cloud computing. ALEJANDRO RAMÍREZ-ARROYO was born in Córdoba, Spain, in 1997. He received the B.Sc. degree, M.Sc degree and Ph.D in telecommunication engineering from the University of Granada (UGR), Spain, in 2019, 2021 and 2023, respectively. From 2019 to 2024, he was with the Smart Wireless Applications and Technologies (SWAT) Research Group in the Department of Signal Theory, Telematics, and Communications, University of Granada. In 2022, he was with the Wireless Communication Networks section, Aalborg University (AAU), Denmark, as an invited Ph.D. Student. Since 2024, he is a PostDoctoral Researcher in the Department of Electronic Systems, Aalborg University (AAU), Denmark. His current research interests include optimization techniques, radio propagation and channel characterization for mmWaves and 5G communications. 17 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Raftopoulou et al.: Evaluation of 5G Private Networks for Enabling Use Cases in Remote Areas MARIA RITA PALATTELLA holds a bachelor’s degree in Telecommunication Engineering (2004) and a master’s degree in Telecommunication Engineering (2007), both with honor, from Politecnico di Bari, Italy. She obtained her PhD in Electronics Engineering from ‘‘Scuola Interpolitecnica di Dottorato’’ (SIPD) and Politecnico di Bari, Italy, in February 2011. From 2011 to 2016 she was a Research Associate at the Interdisciplinary centre for Security, Reliability and Trust (SnT), at the University of Luxembourg. She contributed to the design of MAC protocols and energy-efficient scheduling algorithms for Wireless Sensor Networks. She was involved in several EU FP7 and H2020 projects, including OUTSMART, IoT6, F-Interop. Since 2016, she joined the Luxembourg Institute of Science and Technology (LIST), in the Environmental Research and Innovation (ERIN) department, as a Senior Research and Technology Associate. Since October 2024, she is a Principal Scientist, leading the work on the design of innovative communication systems and network architectures for different Internet of Things (IoT) applications, with a focus on digital agriculture. Her current research interest focuses on IoT-NTN, and she has contributed to this research topic through several projects: ESA ITT M2MSAT, FNR LORSAT, SMC Lux5GCloud, SatNex V ONION, and INVENTIVE. Currently she is coordinating the HEU COMMECT project, aiming to extend connectivity in rural remote area, integrating 5G public and private networks, IoT cellular and not cellular (NB-IoT, LoRa, etc.), NTN (satellites and drones), to make rural communities more sustainable and competitive. She is also the PI of the GEH LIFE and FNR 5G-AGROBOT projects, investigating the use of Digital Twins and Robotics in controlled agriculture environments. She is strongly engaged with Telecommunication and Space Industry, and IoT companies in Luxembourg, and abroad, as evidence by her collaboration in several national and internation projects with local and external stakeholders. She sits on the Editorial Board of the Transactions on Emerging Telecommunications Technologies, and the EAI Transactions on IoT. She has (co)-authored several papers published in well known, high-impact journals and international conferences, and two patents. 18 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2025.3603591 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/