A QoS Framework for Service Provision in Multi-Infrastructure-Sharing Networks Conference: The Twenty-sixth International Symposium on Theory, Algorithmic Foundations, and Protocol Design for Mobile Networks and Mobile Computing; October 27–30, 2025; Houston, TX, USA The Twenty-sixth International S
Quang Minh Nguyen, Eytan Modiano · 2025
Read A QoS Framework for Service Provision in Multi-Infrastructure-Sharing Networks Conference: The Twenty-sixth International Symposium on Theory, Algorithmic Foundations, and Protocol Design for Mobile Networks and Mobile Computing; October 27–30, 2025; Houston, TX, USA The Twenty-sixth International S by Quang Minh Nguyen, Eytan Modiano (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-01694.txt.
Introduction
Quang Minh Nguyen and Eytan Modiano build a framework for sharing network capacity while keeping service promises. Networks that many groups share must still guarantee data speed and low delay for each user. Their framework gives tunable chance based promises, meaning the promise holds with a stated likelihood. A policy at its core balances drift and penalty terms to keep the system stable over time.
The authors map which promise levels can be met at all in such networks. They prove their policy keeps average rates stable and nears the best result as the promise window grows. Tests in simulation back the theory and show strong handling of mixed demands. The work suits shared wireless and cloud setups where many tenants compete.
Catalog record
| CREATOR | Quang Minh Nguyen, Eytan Modiano |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 142 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-01694
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A QoS Framework for Service Provision in Multi-Infrastructure-Sharing Networks Conference: The Twenty-sixth International Symposium on Theory, Algorithmic Foundations, and Protocol Design for Mobile Networks and Mobile Computing; October 27–30, 2025; Houston, TX, USA The Twenty-sixth International Symposium on Theory, Algorithmic Foundations, and Protocol Design for Mobile Networks and Mobile Computing (MobiHoc ’25), October 27–30, 2025, Houston, TX, USA DOI: 10.1145/3704413.3764414 ISBN: 979-8-4007-1353-8/2025/10 CCS: Networks Network performance analysis CCS: Networks Network design principles CCS: Networks Network performance modeling CCS: Networks Network resources allocation CCS: Networks Network control algorithms CCS: Networks Network design and planning algorithms Quang Minh Nguyen, Eytan Modiano Abstract . We propose a framework for resource provisioning with QoS guarantees in shared infrastructure networks. Our novel framework provides tunable probabilistic service guarantees for throughput and delay. Key to our approach is a Modified Dirft-plus-Penalty (MDP) policy that ensures long-term stability while capturing short-term probabilistic service guarantees using linearized upper-confidence bounds. We characterize the feasible region of service guarantees and show that our MDP procedure achieves mean rate stability and an optimality gap that vanishes with the frame size over which service guarantees are provided. Finally, empirical simulations validate our theory and demonstrate the favorable performance of our algorithm in handling QoS in multi-infrastructure networks. † † cc-license: by 1. Introduction Next generation (NextG) networks ( Zeb et al. 2023 ; Nguyen and Modiano 2023 ; Nguyen et al. 2022 ; Nguyen and Modiano 2025 ) are emerging due to the rapid advancements of 5G/6G infrastructures and in response to the stringent Quality-of-Service (QoS) requirements of modern data-intensive applications. A key attribute of NextG networks is the architectural design that permits seamless integration of heterogeneous Infrastructure Providers (InP) to offer over-the-top (OTT) services to end clients ( Liu and Modiano 2022 ) (See Figure 1 ). Such utilization of underlying shared infrastructure not only enhances overall network capacity, but also facilitates robust service provisioning. These advances in network capabilities call for the development of efficient infrastructure sharing mechanisms that can meet emerging service requirements. In this paper, we focus on designing efficient algorithms for QoS-based service provision in shared multi-infrastructure networks. Efficient service provision is essential for maintaining large-scale multi-infrastructure network operation in order to maintain utilization at the InPs, while meeting clients’ service requirements ( Abdel-Malek et al. 2022 ) . From the system perspective, the implementation of a network orchestrator to facilitate the coordination of infrastructure sharing has been the de facto practice in NextG network deployment ( Zeb et al. 2023 ) . Therefore, the design of infrastructure sharing for networked systems has garnered significant research interest, encompassing both generalized resource allocation and specific contexts such as spectrum sharing ( Joshi et al. 2017 ) and network slicing ( Leconte et al. 2018 ) . Most of the previous work on QoS modeling focused on capturing simple requirements such as throughput ( Neely et al. 2008 ) or energy level ( Huang and Neely 2013 ) , which can be cast as long-term linear constraints. Unfortunately, more sophisticated requirements such as delay and reliability involve non-linear probabilistic constraints that are difficult to model and provision. While there has been a rich literature on modeling real-world QoS requirements (see ( Hou et al. 2009 ) and the references therein), modeling realistic QoS requirements may result in non-linear problems whose solutions are often NP-hard ( Rajkumar et al. 1998 ) and where relaxation techniques result in drastic sub-optimality ( Xin and Xiang 2009 ) . Moreover, there has been very limited work on QoS-based OTT service provision for multi-infrastructure networks, mostly focusing on empirical evaluation ( Samdanis et al. 2013 ) . Figure 1 . Infrastructure sharing in NextG networks. To exemplify a realistic shared multi-infrastructure network, consider a setting of two 5G cellular InPs and five clients running various applications, including video streaming, environmental monitoring and data backup, as in Figure 1 . The InPs dynamically schedule links to serve the clients every 1 ms, which is the supported switching frequency of 5G cellular ( Pedersen et al. 2016 ) . Operating at a slower time-scale, the network orchestrator gathers the clients’ job requests and computes the scheduling decisions for the InPs to serve the clients every time-frame. A common approach in the literature is to treat each time-frame independently to enforce QoS within that frame ( da Silva et al. 2020 ) . For example, to support real-time video streaming with delay requirement of 150 ms and negligible failure probability tolerance ( Khan et al. 2024 ) , the scheduler can set the time-frame size to be 150 ms and provision full delivery of the required throughput per frame. However, this approach neglects inter-frame QoS. Specifically, environmental monitoring can allow delays of seconds to minutes, which correspond to thousands of orchestration time-frames, and does not demand immediate throughput delivery ( Ho et al. 2005 ) , while data backup typically has no hard deadline requirement, in which the data merely needs to be served eventually. The design of QoS-based infrastructure sharing mechanisms thus requires a new QoS framework for OTT services that is powerful enough to ensure broad applicability yet analytically tractable enough to enable the provisioning of service guarantees. To address the aforementioned challenges, we propose a novel QoS framework that captures delay and probabilistic service guarantees for OTT services. The framework is specified by three parameters: time-frame size, guaranteed delivery throughput and reliability. Specifically, our service provisioning is within a fixed time-frame, whose length T_{s} can be chosen to capture stringent latency requirements. For every client i with type- c jobs, our framework specifies a guaranteed delivery throughput \gamma^{c}_{i} that is satisfied with probability at least q^{c}_{i} in each time-frame. Moreover, our model also takes into account queue stability of job requests to ensure long-term QoS for applications such as data backup, where timely but eventual service fulfillment is sufficient. Closest to our work are ( Neely et al. 2008 ; Hou et al. 2009 ) . In particular, ( Neely et al. 2008 ) formulated a carefully designed network utility maximization problem (NUM) with stability constraints to jointly provision fairness and throughput guarantees and ( Hou et al. 2009 ) provides QoS in the form of long-term guaranteed delivery ratio of deadline-constrained packets with probability 1. Our model extends NUM by incorporating both stability constraints and our proposed QoS framework. By setting q^{c}_{i}=0 for all clients and jobs, our model reduces to the classical NUM formulation in ( Neely et al. 2008 ) . For the required \gamma^{c}_{i} and q^{c}_{i}\to 1 , our model aligns with ( Hou et al. 2009 ) in providing guaranteed delivery ratio, while additionally satisfying queue stability. Furthermore, our QoS model with appropriately chosen \gamma^{c}_{i} and q^{c}_{i} provides multi-time-frame delay guarantees. This enables the jobs to be served over multiple time-frames to model applications with soft delay requirements (e.g. average delay) such as environmental monitoring and traditional Internet services. Finally, our framework’s modeling flexibility in tuning T_{s},\gamma^{c}_{i} and q^{c}_{i} allows for various combinations of all these models to provide heterogeneous QoS for different classes of applications. To this end, we present a new QoS-based infrastructure sharing paradigm for NextG networks. Our contributions are as follows: • We present a