5th Generation (5G) New Radio Unlicensed (NR-U) enables Mobile Network Operators (MNOs) to extend their network capacity by leveraging nearly 2 GHz of mid-band unlicensed spectrum. However, delivering delay-sensitive services over shared spectrum is challenging due to unpredictable contention delays and variable channel conditions. To address this, the 3rd Generation Partnership Project (3GPP) introduced Channel Access Priority Classes (CAPCs), which allow multiple concurrent Listen-Before-Talk (LBT) processes with varying access probabilities. This design aims to balance diverse Quality-of-Service (QoS) requirements alongside fair access in shared spectrum environments. Traffic classes are mapped to one of the four CAPCs, allowing probabilistic time-domain resource slicing. Mapping complex, multi-dimensional QoS requirements to CAPCs, however, remains a non-trivial task, as no single CAPC optimally supports a given traffic flow under varying channel conditions and QoS demands. In this paper, we propose a QoS-aware scheduler that decouples traffic flows from specific CAPC processes. This scheduler prioritizes flows for each Transmit-Time-Interval (TTI) within a valid Channel Occupancy Time (COT) regardless of which CAPC wins contention. Our experimental results demonstrate over a 50% reduction in average delay for high-priority traffic and more than a 43% delay reduction across all traffic classes using this decoupled scheduling mechanism, achieved without modifying contention parameters. By leveraging NR-U's synchronized slot structure and refined QoS controls inherited from licensed-access frameworks, this approach significantly improves COT utilization and the delivery of delay-critical services.

In recent years, the availability of open-source 5G stacks that support the use of Software-Defined Radios (SDRs) as Radio Frequency (RF)-frontends has significantly accelerated experimental 5G research. This has enabled researchers to deploy customized 5G networks and utilize them with both SDR-based and Commercial Off-The-Shelf (COTS) User Equipments (UEs) for various applications. However, SDRs, unlike commercial 5G Radio Units (RUs), are designed to support a wide range of frequencies and protocols, leading to limitations when used specifically for 5G deployments. These limitations include supported Sampling Rates (srates), inherent Local Oscillator (LO) leakages, and limited Transmit (Tx) and Receive (Rx) gains. In this work, we benchmark the throughput performance of 5G standalone networks using the srsRAN 5G stack, Open5GS core, and three SDRs from National Instruments (NI) (X310, N310, and B210) in an indoor Over-the-Air (OTA) environment at varying distances. Measured throughput is compared to the estimated theoretical maximum under identical configurations. Also, we demonstrate how srates and LO leakage affect OTA throughput, comparing results across the three SDR platforms. Additionally, we show that similar Received Signal Strength (RSS) can result in significantly different throughput due to LO leakage, correlating these effects with overall performance. We also provide the dataset of the measured OTA throughput and RSS values.

Blockchain technology relies on Peer-to-Peer (P2P) ledgers that require a mechanism to maintain security and ensure consensus. The Proof of Work (PoW) consensus algorithm requires participants to dedicate significant computation to solve a cryptography puzzle. The Proof of Work Difficulty (PoW-D) in large public blockchains, such as Ethereum, is adjusted based on the time taken by the winning miner to mine the last block since the system is unaware of the computation power in the blockchain. The consortium and private blockchains provide greater visibility of the nodes involved, which may have variable compute power over time. Such heterogeneity in miners makes it challenging to maintain fairness leading to super linear profits. To address these issues, we propose two novel Reinforcement Learning (RL)-based techniques to intelligently select miners and adaptively update the PoW-D based on the variable compute power of the heterogeneous miners - RL based Miner Selection (RL-MS) and RL based Miner and Difficulty Selection (RL-MDS). Compared to the existing PoW techniques, the proposed approach lowered the winning percentage of the miners below a fairness factor threshold. Furthermore, the block mining time constraint violation is significantly reduced by up to 97% and up to 85% in RL-MS and RL-MDS, respectively.

This article investigates the deployment of a Near-Real-Time Radio Access Network (RAN) Intelligent Controller (near-RT RIC) on a distributed cloud infrastructure composed of multiple physical sites with different amounts of resources and associated costs. The challenge is dynamically adapting the near-RT RIC deployment to the most cost-effective arrangement while meeting the latency requirements between the near-RT RIC and the controlled nodes. We introduce an optimization model to solve the disaggregated near-RT RIC placement problem, considering a cloud-native infrastructure to minimize the placement cost while satisfying the latency-sensitive control loop requirements across the cloud-edge continuum. Moreover, we describe an experimental environment we created using geographically disparate cloud sites. We present data detailing the latencies of the communication links among these sites and the costs incurred in using this real-world infrastructure. We conduct a performance evaluation of the near-RT RIC deployment, comparing the distributed approach versus a traditional monolithic strategy and evaluating positioning costs, deployment, setup and registration times, and the control loop latency considering three scenarios. Our results show that in a cloud-native environment, the disaggregated near-RT RIC allows cost savings of up to 60% in comparison to a monolithic near-RT RIC while satisfying the control loop latency and achieving time efficiency in terms of deployment and registration of xApps and near-RT RIC components.

In this paper, we present a framework for experimentation in next-generation Initial Access (IA) procedures for Millimeter Wave (mmWave) and Terahertz (THz) communications called SofTwAre-defined Mmwave INitial Access (STAMINA). The IA procedure is one of the essential components for communication systems in high frequencies, enabling directional transmitters and receivers to acquire each other's relative orientation before data transmission. While effective in establishing communication, the existing IA procedure standardized by 3GPP consumes a significant amount of radio resources. Many research efforts have proposed enhancements over the current-generation IA procedure, e.g., leveraging non-uniform beam sweep sequences or adaptive codebooks. However, no existing experimental mmWave platforms support modifications in their standard-compliant IA procedures, preventing their utilization for conducting experimental research on next-generation IA procedures. Our software-defined mmWave framework addresses this gap by combining the flexibility of Software-defined Radios (SDRs) with the directionality of mmWave front-ends to perform customizable IA procedures. We demonstrate STAMINA's ability to control mmWave frontends correctly, its increased performance over traditional static experiments, and its flexibility to customize the IA parameters to achieve different objectives. Our results show that STAMINA provides experimenters with a flexible platform for performing experiments on next-generation IA procedures.

This paper presents an experimental study on mmWave beam profiling on a mmWave testbed, and develops a machine learning model for beamforming based on the experiment data. The datasets we have obtained from the beam profiling and the machine learning model for beamforming are valuable for a broad set of network design problems, such as network topology optimization, user equipment association, power allocation, and beam scheduling, in complex and dynamic mmWave networks. We have used two commercial-grade mmWave testbeds with operational frequencies on the 27 Ghz and 71 GHz, respectively, for beam profiling. The obtained datasets were used to train the machine learning model to estimate the received downlink signal power, and data rate at the receivers (user equipment with different geographical locations in the range of a transmitter (base station). The results have showed high prediction accuracy with low mean square error (loss), indicating the model's ability to estimate the received signal power or data rate at each individual receiver covered by a beam. The dataset and the machine learning based beamforming model can assist researchers in optimizing various network design problems for mmWave networks.

In this article, we present the design and implementation of the software-defined IoT management (SDIM) framework based on software-defined networking (SDN)-enabled architecture that is purposely built for the edge computing multidomain wireless sensor networks (WSNs). This framework can dynamically provision the IoT devices to enable machine-to-machine (M2M) communication as well as continuous operational fault detection for WSNs. Unlike the existing approaches in the literature, SDIM is mainly deployed at multiaccess edge computing (MEC) nodes and is integrated with the cloud by aggregating multidomain topology information. Backed by the experimental results over the University of Bristol 5G test network, we demonstrate in practice that our framework outperforms the implementations of the lightweight M2M (LWM2M) and NETCONF Light IoT device management protocols when deployed autonomously at the network edge and/or the cloud. Specifically, SDIM edge deployments can lower the average device provisioning time as high as 46% compared to LWM2M and 60.3% compared to NETCONF Light. Moreover, it can decrease the average operational fault detection time by approximately 33% compared to LWM2M and roughly 40% compared to NETCONF Light. Also, SDIM reduces control operations time up to 27%, posing a powerful feature for use cases with time-critical control requirements. Last, SDIM manages to both reduce CPU consumption and to have important energy consumption gains at the network edge, which can reach as high as 20% during device provisioning and 4.5%-4.9% during fault detection compared to the benchmark framework deployments.

In this paper, we propose a novel congestion control framework for delay- and disruption tolerant networks (DTNs). The proposed framework, called Smart-DTN-CC, adjusts its operation automatically as a function of the dynamics of the underlying net- work. It employs reinforcement learning, a machine learning technique known to be well suited to problems in which the environment, in this case the network, plays a crucial role; yet, no prior knowledge about the target environment can be assumed, i.e., the only way to acquire information about the environment is to interact with it through continuous online learning. Smart-DTN-CC nodes get input from the environment (e.g., its buffer occupancy, set of neighbors, etc), and, based on that information, choose an action to take from a set of possible actions. Depending on an action's effectiveness in controlling congestion, it will be given a reward. Smart-DTN-CC's goal is to maximize the over- all reward which translates to minimizing congestion. To our knowledge, Smart-DTN-CC is the first DTN congestion control framework that has the ability to automatically and continuously adapt to the dynamics of the target environment. As demonstrated by our experimental evaluation, Smart-DTN- CC is able to consistently outperform existing DTN congestion control mechanisms under a wide range of network conditions and characteristics.

In this paper, we propose a novel modeling framework to study congestion in delay- and disruption tolerant networks (DTNs). The proposed model is based on directed site-bond percolation where sites represent space-time positions of DTN nodes, and bonds are contact opportunities, i.e. Communication links that can be established whenever nodes come in range of each other. To the best of our knowledge, this is the first model of DTN congestion using percolation theory. The proposed modeling framework is simple yet general and can be used to evaluate different DTN congestion control mechanisms in a variety of scenarios and conditions. We validate our model by showing that its results match quite well results obtained from the ONE DTN simulation platform. We also show that our model can be used to understand how parameters like buffer management policy, buffer size, routing mechanism, and message time-to-live affect network congestion.