Cloud computing data centers handle highly variable workloads: job resource requirements can range from just one or few cores to thousands, and job service times can range from milliseconds to hours or days. This variability significantly limits the maximum achievable utilization of infrastructures. Queuing theory has addressed the study of these systems with the definition of the Multiserver-Job Queuing Model (MJQM), where s identical servers are present and job n requires alpha(n) of the s servers for a random amount of time sigma(n). The alpha(n) servers are occupied and released simultaneously. Unfortunately, despite its simple formulation, the MJQM remains elusive. For example, the MJQM stability condition has been derived only in particular cases. As a consequence, even applying Discrete-Event Simulation (DES) under high load becomes challenging, because stability cannot be determined a priori. In this paper, we analyze the MJQM with general independent arrival processes and service times under FCFS scheduling, using stochastic recurrence equations (SREs) and ergodic theory. Starting from the definition of the MJQM SRE, we prove the monotonicity and separability properties that allow us to apply an extension of Loynes' theorem, known as the monotone-separable framework, and formally define the MJQM stability condition. From these results, we introduce and implement two algorithms: the first one is used to draw sub-perfect samples (SPS) of the system's workload and the second one estimates the system's stability condition given the statistics of the jobs' input stream. The nature of the SPS algorithm allows for a massive GPU parallelization, thus greatly improving the efficiency in the estimation of performance metrics. The algorithm for the estimation of the stability condition solves an important problem for the analysis of MJQMs. We also define new metrics that capture the synchronization loss in MJQM systems and we show how these metrics can be efficiently evaluated using the SRE approach. Finally, we show that the approach proposed in this paper can be extended to more complicated systems, including MJQMs where resources have types.
The emergence of eXtended Reality (XR) technologies is revolutionizing Mission Critical (MC) operations by enhancing situational awareness and decision-making. However, the high computational demands of XR MC applications, coupled with the limited capabilities of battery-powered wearable XR devices worn, e.g., by first responders, necessitate offloading strategies to more processing-powerful network nodes. Traditional terrestrial networks, while supporting XR MC services, may not be reliable in all scenarios, especially during emergencies or in remote areas. To address this, the integration of Non-Terrestrial Networks (NTNs) with Terrestrial Networks (TNs) offers various options to place and run in-network computing tasks, e.g., Low Earth Orbit (LEO) satellites and Unmanned Aerial Vehicles (UAVs). The potential of these offloading options for XR MC services has not yet been fully explored. In this work, we close this gap and analyze the performance of application-driven offloading of computational tasks of XR MC services at different locations in the integrated TN/NTN environment. Through system-level simulations, we assess the end-to-end latency cost under different traffic loads at the various system layers and analyze the energy consumption of XR device, identifying practical insights for system designers. For a small number of requests, offloading is more effective than local computing, improving performance by up to 93%, whereas, for a high number of requests, local computing is preferred but constrained by battery limitations.
Frequency Range 3 (FR3) in the 7–24 GHz band will be the new spectrum for 6G wireless networks. The bandwidth availability and diversity of FR3 offer unprecedented opportunities for coherent multiband Integrated Sensing And Communication (ISAC), which aggregates the carrier phase information from multiple frequency bands to increase the sensing resolution to the cm-level. However, the frequency anisotropy of sensing targets over GHz-wide bands and the non-contiguity of the 6G spectrum, pose critical challenges to the application of existing multiband ISAC techniques. We present the first study on coherent multiband sensing in FR3. We experimentally characterize the frequency anisotropy of targets and propose new phase coherence metrics for multiband processing. Then, we analyze the impact of non-contiguous FR3 bands considered by 3GPP, and design a new algorithm to mitigate the resulting sensing artifacts, outperforming existing techniques. Our results represent a first step toward fully developing multiband ISAC for FR3.
Integrated sensing and communication (ISAC) has gained significant traction in recent years, primarily because it allows existing communication infrastructure to support sensing applications with minimal additional costs. In particular, millimeter-wave (mmWave) ISAC has the potential to offer improved sensing performance in applications such as pose estimation and gesture recognition. For complex sensing tasks and environments, data-driven sensing, which relies on deep learning, is becoming increasingly popular and has shown promising results. However, deep learning models for these tasks require large labeled datasets to achieve high accuracy. Dataset collection and labeling are labor-intensive and time-consuming. Consequently, there is growing interest in leveraging unlabeled data to overcome these challenges. To address this, we propose mmGAN, a semi-supervised method for ISAC-based gesture recognition. We propose a novel loss function for mmGAN based on softplus, feature matching, and manifold regularization to significantly improve gesture recognition performance. We evaluate mmGAN on a 5G Orthogonal Frequency Division Multiplexing (OFDM) mmWave dataset comprising power per beam pair measurements. When training both mmGAN and the supervised baseline with only 0.6% of the labeled data, mmGAN demonstrates up to 25 percentage points higher accuracy than the supervised baseline. Our method serves as a strong foundation for cross-subject transfer learning, demonstrating the significant value of leveraging unlabeled data to enhance cross-domain sensing performance in ISAC systems. Our results demonstrate that the proposed loss function achieves superior performance across diverse subjects. Further, mmGAN significantly narrows the performance gap between semi-supervised and fully supervised models on the publicly available Widar dataset. Moreover, we provide an interpretable analysis of mmGAN performance through saliency maps and ablation studies, revealing key insights into the model’s behavior and generalization. This work is the first to evaluate gesture recognition performance in 5G OFDM mmWave ISAC systems using a semi-supervised learning approach, covering the entire pipeline from testbed implementation to model evaluation.
Deep reinforcement learning (DRL) promises adaptive control for future mobile networks but conventional agents remain reactive: they act on past and current measurements and cannot leverage short-term forecasts of exogenous KPIs such as bandwidth. Augmenting agents with predictions can overcome this temporal myopia, yet uptake in networking is scarce because forecast-aware agents act as closed-boxes; operators cannot tell whether predictions guide decisions or justify the added complexity. We propose SIA, the first interpreter that exposes in real time how forecast-augmented DRL agents operate. SIA fuses Symbolic AI abstractions with per-KPI Knowledge Graphs to produce explanations, and includes a new Influence Score metric. SIA achieves sub-millisecond speed, over 200x faster than existing XAI methods. We evaluate SIA on three diverse networking use cases, uncovering hidden issues, including temporal misalignment in forecast integration and reward-design biases that trigger counter-productive policies. These insights enable targeted fixes: a redesigned agent achieves a 9% higher average bitrate in video streaming, and SIA's online Action-Refinement module improves RAN-slicing reward by 25% without retraining. By making anticipatory DRL transparent and tunable, SIA lowers the barrier to proactive control in next-generation mobile networks.
