Non-Orthogonal Multiple Access (NOMA) is a Resource Sharing Mechanism (RSM) initially studied for 5G cellular networks and brought back to the agenda for 6G networks. While NOMA’s benefit at the level of a single cell has been properly established, assessing its performance at the scale of a cellular network remains an open research problem. This is mainly due to the inter-dependencies between scheduling, power control and inter-cell interference. Some algorithms have been proposed to optimize resource allocation in a multi-cell network, but they require a perfect and unrealistic knowledge of the whole channel states. In this paper, we leverage Bayesian Optimization techniques to build a versatile evaluation framework, able to assess the performance of multi-cell networks implementing a large variety of RSMs under a minimal set of assumptions. Subsequently, we illustrate how this evaluation framework can be used to compare the performance of several well-known RSMs under various fairness requirements and beamforming efficiencies. Our results show that, among the RSMs studied on a simple multi-cell network simulation, NOMA combined with a full reuse policy consistently emerges as the one able to achieve the highest end-users achievable rates under fairness constraints.

Light Fidelity (LiFi) is a recent wireless technology that offers high data rates, low power consumption, and does not interfere with other radio technologies, such as Wi-Fi. However, its short range and need for a direct line of sight require integration with other wireless technologies for effective use in a wireless local area network (WLAN). In this paper, we study the feasibility and efficiency of Hybrid LiFi/Wi-Fi Networks (HLWNet). Our study aims to determine whether HLWNet is a viable solution for WLANs in terms of throughput and power consumption. We propose an approach to evaluate HLWNet deployments, specifying the optimal number and placement of access points (APs) for each technology to ensure throughput satisfaction for stations (STAs) while minimizing overall network power consumption. Our numerical results indicate that the use of LiFi is not advantageous in all scenarios.

Monitoring the quality of streaming video applications is important for Internet service providers (ISPs) to detect network issues and facilitate capacity planning. Machine Learning (ML) inference models have emerged as an effective solution to determine service quality using network traffic. However, while much focus has been on enhancing model performance, little attention has been given to deploying these models across entire networks. This paper introduces a new placement approach of quality inference models and their associated tasks to enhance the monitoring of video streaming applications over an entire mobile traffic network. Starting from the observation that inference tasks require the deployment of multiple components to, first, calculate input features from raw traffic, and then execute the inference models, we define the placement problem as an integer programming problem and, given its NP-hardness, we provide a heuristic solution, experimentally close to the optimum, based on the relaxation and the rounding of fractional solutions. We highlight that decoupling these components for the inference of network traffic can be beneficial in terms of total accuracy of the ML inference tasks. Finally, we experimentally show that our solution outperforms state-of-the-art placement techniques by ~30% of accuracy of the deployed inference models.

Target Wake Time (TWT) is a key energy-saving mechanism in IEEE 802.11 WLANs, allowing stations (STAs) to reduce power consumption by restricting transmissions to scheduled service periods (SPs). Despite its potential, TWT remains underutilized due to concerns over performance degradation, especially under congested network conditions. In this paper, we address this limitation by proposing the Power Mode Selection Algorithm (PMSA), a lightweight and standard-compliant mechanism implemented at the Access Point (AP). PMSA operates on top of existing TWT schedulers and dynamically determines whether each STA should operate in power save (PS) mode or temporarily switch to active mode (and disable TWT) based on current network congestion. Our simulation results demonstrate that PMSA enhances the robustness of TWT scheduling by adapting to traffic conditions, maintaining low delays and avoiding packet losses even under heavy load, while preserving energy savings when feasible.

International audience

Bayesian Optimization (BO) is typically used to optimize an unknown function $f$ that is noisy and costly to evaluate, by exploiting an acquisition function that must be maximized at each optimization step. Even if provably asymptotically optimal BO algorithms are efficient at optimizing low-dimensional functions, scaling them to high-dimensional spaces remains an open problem, often tackled by assuming an additive structure for $f$. By doing so, BO algorithms typically introduce additional restrictive assumptions on the additive structure that reduce their applicability domain. This paper contains two main contributions: (i) we relax the restrictive assumptions on the additive structure of $f$ without weakening the maximization guarantees of the acquisition function, and (ii) we address the over-exploration problem for decentralized BO algorithms. To these ends, we propose DuMBO, an asymptotically optimal decentralized BO algorithm that achieves very competitive performance against state-of-the-art BO algorithms, especially when the additive structure of $f$ comprises high-dimensional factors.

Light Fidelity (LiFi) is a recent wireless technology that offers high data rates, low power consumption and do not interfere with radio frequency technologies. As a result, LiFi access points emerge as promising complements to Wi-Fi access points for hybrid LiFi/Wi-Fi networks (HLWNets). This combination can offer high data rates and energy efficiency. However, to fully benefit from hybrid networks, it is crucial to efficiently assign stations, whether with LiFi or Wi-Fi access points. In this paper, we formulate the AP assignment problem as an optimization problem. As a solution, we present an AP assignment strategy, in which we elaborate an algorithm with a low computational complexity. The algorithm allows to dynami-cally assign STAs to APs and benefits from the combination of the two technologies to enhance network performance (energy consumption and throughput satisfaction).

The Internet of Things (IoT) is transforming all economic sectors by connecting physical assets to the virtual world. The range of low-power connectivity options is continuously widening the range of possible IoT applications. However, too many possibilities often make it hard for industrial specialists to choose the right technology and configuration settings, yet these are crucial decisions. To deeply analyze and compare the performance and the scalability of various solution designs, one proven method is simulation. In this article, we show how IoT network simulation can help to future-proof an IoT connectivity design, without the burden of installing a lot of hardware and writing complicated scripts. Then, as the network simulation process is too complex for most IoT teams, we propose a no-code online IoT network simulation platform to make this powerful tool accessible to all. In particular, we explain how we hide the simulation workflow complexity via relevant abstractions and transform it into intuitive interactions. We illustrate the method and the usage of this promising approach, which can be integrated in a network digital twin. We show how it permits to easily evaluate what-if scenarios in order to answer a set of questions that may arise all along the life cycle of a smart connected solution.

The IEEE 802.11 protocol has become the de facto communication technology for WLANs (Wireless Local Area Networks). While considered reliable and efficient for applications requiring high datarate, IEEE 802.11 is often disregarded for energy-sensitive applications such as IoT (Internet of Things). In fact, the IEEE 802.11 standard and its amendments have introduced several energy-saving mechanisms over the years that are rarely used in practice. In this paper, we consider seven possible energy-saving strategies for the MAC layer of IEEE 802.11. Using two case studies, we evaluate and compare these energy-saving strategies with regard to their network performance and energy saving. We conclude that most of the energy-saving strategies manage to support the levels of workloads considered in our case studies, and at the same time, they succeed to cut energy consumption by a factor ranging from two to eight. In particular, the "DL slot'' strategy leads STAs to consume only half (or even less) of what they would without running any strategy, without trading off the attained levels of throughput. The "DL prompt + UL slot'' strategy is the most efficient energy-wise, but this can come at the expense of a loss of throughput.

In this paper, we propose different strategies to efficiently deploy RSUs in a city with the ultimate goal of having an 802.11p-based infrastructure to deliver Internet services. Unlike most existing works, (i) our strategies' only prior information is the average density of vehicles in the studied area, and (ii) they rely on the forecast of a performance model of 802.11p to assist and guide their choices regarding the location of RSUs. With the help of two simulators, namely SUMO and ns-3, we investigate the behavior of each strategy in three scenarios inspired by the street map of real-life major cities. Our findings are twofold: (i) we demonstrate that any efficient RSUs deployment is tightly tied to the specifics of the considered city (namely, the arrangement of streets and the spatial density of vehicles); (ii) the best strategy is not to position RSUs where the traffic density is at its highest, nor at the street junctions where the traffic density is often at its highest but instead where they will be able to deliver the target QoS to a maximum number of vehicles.

Systems with multiserver jobs, inspired by modern datacenters, present a challenge in terms of the analysis of their performance and, in particular, the determination of their processing capacity as some servers may remain idle even though there are jobs queued for service. We consider a generalization of the multiserver jobs model to a resource of not necessarily integer quantity and jobs of different classes requiring arbitrary fractions of that resource. We present a simple approach to the analysis of such a system in steady state. Our approach relies on a suitably chosen state description and the use of conditional probabilities. The limiting values of these conditional probabilities allow us to determine the asymptotic maximum job processing capacity without having to obtain the full solution for the system.

The Internet of Things (IoT) is the convergence of the physical and the digital worlds. It enables a large spectrum of applications such as smart building, smart tracking, smart metering, predictive maintenance, remote control, augmented reality or video surveillance. The diversity of these applications has caused a profusion of the IoT communication technologies offerings for exchanging data between IoT devices and applications. The latter technologies come with different features in terms of range, throughput, latency, scalability, energy, etc. Each technology can fit several use cases and a use case can leverage several technologies. It is complex, yet critical, for an IoT architect to evaluate the adequacy and the limits of a network technology for a targeted application and to continuously optimize its configuration as the deployment evolves. This paper introduces ADIperf, a framework to simplify and systematize the evaluation of the performance of an IoT communication technology for a given IoT use case and context. The ADIperf approach pays special attention to the energy efficiency as well as to the ability of an IoT communication technology to properly scale up with the number of end-devices, with the ultimate goal of giving guidelines and tools for IoT architects to select the technology and configure the network that fulfill their application's needs over time.

WLANs, which have overtaken wired networks to become the primary means of connecting devices to the Internet, are prone to performance issues due to the scarcity of space in the radio spectrum. As a response, IEEE 802.11ax and subsequent amendments aim at increasing the spatial reuse of a radio channel by allowing the dynamic update of two key parameters in wireless transmission: the transmission power (TX_POWER) and the sensitivity threshold (OBSS_PD). In this paper, we present INSPIRE, a distributed online learning solution performing local Bayesian optimizations based on Gaussian processes to improve the spatial reuse in WLANs. INSPIRE makes no explicit assumptions about the topology of WLANs and favors altruistic behaviors of the access points, leading them to find adequate configurations of their TX_POWER and OBSS_PD parameters for the ''greater good" of the WLANs. We demonstrate the superiority of INSPIRE over other state-of-the-art strategies using the ns-3 simulator and two examples inspired by real-life deployments of dense WLANs. Our results show that, in only a few seconds, INSPIRE is able to drastically increase the quality of service of operational WLANs by improving their fairness and throughput.

The latest amendment 802.11ax to the IEEE 802.11 standard, better known by its commercial name Wi-Fi 6, includes a feature that aims at improving the spatial reuse of a channel: each device can adapt its Clear Channel Assessment sensitivity threshold and its transmission power. In this paper, we use the Multi-Armed Bandit (MAB) framework to propose a centralized solution to dynamically adapt these parameters. We propose a new approach based on a Gaussian mixture to sample new network configurations, a specific reward function that prevents starvations when maximized, as well as a method based on Thompson Sampling to select the best network configuration. We evaluate our solution using the network simulator ns-3 and different topologies. Simulation results confirm the large benefits that 802.11ax may bring to spatial reuse. They also demonstrate the efficiency of our solution in finding appropriate parameter configurations that significantly improve the quality of service of the networks.

We consider a VoD (Video on-Demand) platform designed for vehicles traveling on a highway or other major roadway. Typically, cars or buses would subscribe to this delivery service so that their passengers get access to a catalog of movies and series stored on a back-end server. Videos are delivered through IEEE 802.11p Road Side Units deployed along the highway. In this paper, we propose a simple analytical and yet accurate solution to estimate (at the speed of a click) two key performance parameters for a VoD platform: (i) the total amount of data downloaded by a vehicle over its journey and (ii) the total "interruption time'', which corresponds to the time a vehicle spends with the playback of its video interrupted because of an empty buffer. After validating its accuracy against a set of simulations run with ns-3, we show an example of application of our analytical solution for the sizing of an IEEE 802.11p-based VoD platform.

Data plane development kit (DPDK) works as a specialized library that enables virtual switches to accelerate the processing of incoming packets by, among other things, balancing the incoming flow of packets over all the CPU cores and processing packets by batches to make a better use of the CPU cache. Although DPDK has become a de facto standard, the performance modeling of a DPDK-based vSwitch remains a challenging problem. In this paper, we present an analytical queueing model to evaluate the performance of a DPDK-based vSwitch. Such a virtual equipment is represented by a complex polling system in which packets are processed by batches, i.e., a given CPU core processes several packets of one of its attached input queues before switching to the next one. To reduce the complexity of the associated model, we develop a general framework that consists in decoupling the polling system into several queueing subsystems, each one corresponding to a given CPU core. We resort to servers with vacation to capture the interactions between subsystems. Our proposed solution is conceptually simple, easy to implement and computationally efficient. Tens of comparisons against a discrete-event simulator show that our models typically deliver accurate estimates of the performance parameters of interest (e.g., attained throughput, packet latency or loss rate). We illustrate how our models can help in determining an adequate setting of the vSwitch parameters using several real-life case studies.

Current architecture of many computer systems relies on dynamic allocation of a pool of resources according to workload conditions to meet specific performance objectives while minimizing cost (e.g., energy or billing). In such systems, different levels of operation may be defined, and switching between operating levels occurs at certain thresholds of system congestion. To avoid rapid oscillations between levels of service, “hysteresis” is introduced by using different thresholds for increasing and decreasing workload levels, respectively. We propose a model of such systems with general arrivals, arbitrary number of servers and operating levels where each higher operating level may correspond to an arbitrary number of additional servers and soft (i.e., non-deterministic) thresholds to account for “inertia” in switching between operating levels. In our model, request service times are assumed to be memoryless and server processing rates may be a function of the current operating level and of the number of requests (users) in the system. Additionally, we allow for delays in the activation of additional operating levels. We use simple mathematics to obtain a semi-numerical solution of our model. We illustrate the versatility of our model using several case study examples inspired by features of real systems. In particular, we explore optimal thresholds as a tradeoff between performance and energy consumption.

WLANs (Wireless Local Area Networks) have become ubiquitous in our everyday life, and are mostly based on IEEE 802.11 standards. In this paper, we consider the performance evaluation of an arbitrary-topology unsaturated network based on the IEEE 802.11 DCF. We present a conflict graph-based modeling approach to discover the attainable throughput of each node. Our model consists of a single Markov chain which aims at describing, at a high-level of abstraction, the current state of the entire wireless network. Owing to its low complexity, our approach is simple to implement, can cope with medium sized networks, and its execution speed is fast. We validate its accuracy against a discrete-event simulator. Results show that our approach is typically accurate, with associated relative errors generally less than 15%, and that it captures complex phenomena such as node starvation. We investigate two potential applications of our proposed approach in which, starting with a given network, we improve its performance in terms of overall throughput or fairness by throttling the throughput demand of a node, or by turning a node off altogether.

Though SDN (Software Defined Network) provides the executive building blocks for programming data-plane appliances, controller decisions must be grounded in an accurate outlook on the network topology and performance. In this context, we focus on the possibility of providing accurate measurements for the end-to-end (E2E) delay in SDN networks. In practice, like many variable quantities, a good description of the E2E delay requires characterizing its first two moments, i.e., expectation and variance. We propose to estimate the E2E delay by making use only of measurements collected locally on each node of the network. We extend a procedure that has been proposed to estimate the E2E delay in the case of one flow to handle the case of multiple competing flows. We compare its accuracy using several scenarios, with different types of traffic following real traces, different topologies and bandwidth. Also, an analysis of the computational and networking costs of our solution is proposed.

With Software-Defined Networking (SDN), computer networks will gain a better control and management on their physical resources and on flows requiring a specific type of QoS. To fulfil these objectives, networks must be able to characterize the end-to-end performance of flows, which are usually unknown and not directly measurable. Instead, networks have typically at their disposal only local measurements, collected at each node. In this paper, we propose a new method to evaluate the variance of the end-to-end delay based only on measurements collected on nodes. The core of our solution is to link samples of waiting times at different nodes in order to approximate the corresponding covariance terms, needed to refine the approximate value of the sought variance. We evaluate the accuracy of our solution using two different scenarios. The obtained results show that the method is generally good, with a relative error on the estimated standard deviation of the end-to-end delay usually close to 5%, though it may exceed 10% when the network is facing high levels of load and experiencing packet losses.

Virtual switches are a key elements within the new paradigms of Software Defined Networking (SDN) and Network Function Virtualization (NFV). Unlike proprietary networking appliances, virtual switches come with a high level of flexibility in the management of their physical resources such as the number of CPU cores, their allocation to the switching function, and the capacities of the RX queues, which gives the opportunity for an efficient sizing of the system resources. We propose a model for the performance evaluation of a virtual switch. Our model resorts to servers with vacation to capture the involved interactions between queues resulting from the implemented polling strategies. The solution to the model is found using a simple fixed-point iteration and it provides estimates for customary performance metrics such as the attained throughput, the packet latency, the buffer occupancy and the packet loss rate. In the tens of explored examples, the predictions of the model were found to be accurate, thereby allowing their use for the purpose of sizing problems.

In this paper, we propose an algorithm to estimate the second moment of the end-to-end delay experienced by the packets of a flow based only on delay measurements locally collected by the network nodes. Our solution estimates the standard deviation of the end-to-end delay in an easy and computationally efficient way. Based on thousands of simulations using a real-life trace, our solution is found to be accurate, typically differing by only a few percent from the actual value of the standard deviation of the end-to-end delay.

In this paper we consider the problem of combining calibrated performance models of system components in order to predict overall system performance. We focus on open workload system models, in which, under certain conditions, obtaining and validating the overall system performance measures can be a simple application of Little's law. We discuss the conditions of applicability of such a simple validation methodology, including examples of successful application, as well as examples where this approach fails. Additionally, we propose to analyze the deviations between the model predictions and system measurements, so as to decide if they correspond to "measurement noise" or if an important system component has not been correctly represented. This approach can be used as an aid in the design of validated system performance models.

Cloud providers need to size their systems to determine the right amount of resources to allocate as a function of customer's needs so as to meet their SLAs (Service Level Agreement), while at the same time minimizing their costs and energy use. Queueing theory based tools are a natural choice when dealing with performance aspects of the QoS (Quality of Service) part of the SLA and forecasting resource utilization. The characteristics of a cloud center lead to a queueing system with multiple servers (nodes) in which there is potentially a very large number of servers and both the arrival and service process can exhibit high variability. We propose to use a G/G/c-like model to represent a cloud system and assess expected performance indices. Given the potentially high number of servers in a cloud system, we present an efficient, fast and easy-to-implement approximate solution. We have extensively validated our approximation against discrete-event simulation for several QoS performance metrics such as task response time and blocking probability with excellent results. We apply our approach to examples of system sizing and our examples clearly demonstrate the importance of taking into account the variability of the tasks arrivals and thus expose the risk of under- or over-provisioning if one relies on a model with Poisson assumptions.

In this paper, we follow up an existing modeling framework to analytically evaluate the performance of multi-hop flows along a wireless chain of four nodes. The proposed model accounts for a non-perfect physical layer, handles the hidden node problem, and is applicable under workload conditions ranging from flow(s) with low intensity to flow(s) causing the network to saturate. Its solution is easily and quickly obtained and delivers estimates for the expected throughput and for the datagram loss probability of the chain with a good accuracy.

A large number of real-life systems can be viewed as instances of the classical M/G/c/N queue. The exact analytical solution of this queueing model is not known, and a frequently-used approach is to replace the general service time distribution by a phase-type distribution. The advantage of this approach is that the resulting M/Ph/c/N queue can be described by familiar balance equations. The downside is that the size of the resulting state space suffers from the "dimensionality curse", i.e., exhibits combinatorial growth as the number of servers and/or phases increases. To circumvent this complexity issue, we propose to use, instead of the classical full state description, a reduced state description in which the state of only one server is represented explicitly, while the other servers are accounted for through their rate of completions. The accuracy of the resulting approximation is generally good and, moreover, tends to improve as the number of servers in the system increases. Its computational complexity in terms of the number of states grows only linearly in the number of servers and phases, thus making the numerical solution of such queues with hundreds of servers and a reasonable number of phases computationally affordable.

We propose a modelling framework, that can be used to reproduce the workload volatility of a Video on Demand (VoD) system. Based on numerical simulations, we evaluate the precision of the estimation procedure we derive to calibrate our parametric model. We also compare its performance to that of other existing models examining the goodness-of-fit of the steady state distribution and of the autocorrelation function of real workload traces. We then give each parameter of the model an interpretation in terms of the workload volatility, that enlightens on some origins of the system dynamics, like the users behaviour.

IEEE 802.11 is implemented in many wireless networks, including multi-hop networks where communications between nodes are conveyed along a chain. We present a modeling framework to evaluate the performance of flows conveyed through such a chain. Our framework is based on a hierarchical modeling composed of two levels. The lower level is dedicated to the modeling of each node, while the upper level matches the actual topology of the chain. Our approach can handle different topologies, takes into account Bit Error Rate and can be applied to multi-hop flows with rates ranging from light to heavy workloads. We assess the ability of our model to evaluate loss rate, throughput, and end-to-end delay experienced by flows on a simple scenario, where the number of nodes is limited to three. Numerical results show that our model accurately approximates the performance of flows with a relative error typically less than 10%.

Many methods have been proposed in the literature to perform admission control in order to provide a sufficient level of Quality of Service (QoS) to accepted flows. In this paper, we introduce a novel data-driven method based on a time-varying model that we refer to as Knowledge-Based Admission Control solution (KBAC). Our KBAC solution consists of three main stages: (i) collect measurements on the on-going traffic over the communication link; (ii) maintain an up-to-date broad view of the link behavior, and feed it to a Knowledge Plane; (iii) model the observed link behavior by a mono-server queue whose parameters are set automatically and which predicts the expected QoS if a flow requesting admission were to be accepted. Our KBAC solution provides a probabilistic guarantee whose admission threshold is either expressed, as a bounded delay or as a bounded loss rate. We run extensive simulations to assess the behavior of our KBAC solution in the case of a delay threshold. The results show that our KBAC solution leads to a good trade-off between flow performance and resource utilization. This ability stems from the quick and automatic adjustment of its admission policy according to the actual variations on the traffic conditions.

We propose a simple approximation to assess the steady-state probabilities of the number of customers in Ph/Ph/1 and Ph/Ph/1/N queues, as well as probabilities found on arrival, including the probability of buffer overflow for the Ph/Ph/1/N queue. The phase-type distributions considered are assumed to be acyclic. Our method involves iteration between solutions of an M/Ph/1 queue with state-dependent arrival rate and a Ph/M/1 queue with state-dependent service rate. We solve these queues using simple and efficient recurrences. By iterating between these two simpler models our approximation divides the state space, and is thus able to easily handle phase-type distributions with large numbers of stages (which might cause problems for classical numerical solutions). The proposed method converges typically within a few tens of iterations, and is asymptotically exact for queues with unrestricted queueing room. Its overall accuracy is good: generally within a few percent of the exact values, except when both the inter-arrival and the service time distributions exhibit low variability. In the latter case, especially under moderate loads, the use of our method is not recommended.

Admission control is a mechanism used to restrict access to a computer network to some flows based on the current utilization level of the network resource. By regulating the number of on-going flows, admission control aims at preventing over loading, congestion and performance collapses, so that, accepted flows receive a sufficient level of Quality of Service (QoS). In this paper, we evaluate three existing measurement-based admission control (MBAC) solutions, and we compare their efficiency in the context of semantic networks. Semantic networks refer to networks that autonomously acquire a knowledge on the on-going traffic as well as on any new incoming flow requesting admission. In this framework, we configure the three MBAC solutions in a way they have an identical target in terms of maximum tolerable packet loss rate or maximum tolerable packet queueing delay. We evaluate the solutions performance analytically or by simulation, and compare them to the "ideal" admission control. The results show that one solution, outperforms the others in meeting the target performance.

In this paper, we explore a way to find the right scheduling policy for WiMAX networks, that achieves the best compromise between an efficient use of the resource and a relative fairness among users. This problem is of primary importance as no scheduling policy has been recommended in the WiMAX standard. To do so, we develop an extension of our previous analytical model for WiMAX networks, that takes into account a more general scheduling policy than those previously studied (i.e., instantaneous throughput fairness, slot sharing fairness and opportunistic scheduling). We show that this general policy covers the two extreme cases, namely the instantaneous throughput fairness policy and the opportunistic policy, and offers intermediate policies that are good candidates for finding the right trade-off. In order to formulate the decision criterion, we introduce a new performance parameter, the mean throughput obtained by a user depending on its efficiency to use the resource. The model has a closed-form solution, and all performance parameters can be obtained instantaneously. This allows us to carry out dimensioning studies that require several thousands of evaluations, which would not be tractable with any simulation tool.