We study Constrained Online Convex Optimization with Memory (COCO-M), where both the loss and the constraints depend on a finite window of past decisions made by the learner. This setting extends the previously studied unconstrained online optimization with memory framework and captures practical problems such as the control of constrained dynamical systems and scheduling with reconfiguration budgets. For this problem, we propose the first algorithms that achieve sublinear regret and sublinear cumulative constraint violation under time-varying constraints, both with and without predictions of future loss and constraint functions. Without predictions, we introduce an adaptive penalty approach that guarantees sublinear regret and constraint violation. When short-horizon and potentially unreliable predictions are available, we reinterpret the problem as online learning with delayed feedback and design an optimistic algorithm whose performance improves as prediction accuracy improves, while remaining robust when predictions are inaccurate. Our results bridge the gap between classical constrained online convex optimization and memory-dependent settings, and provide a versatile learning toolbox with diverse applications.

International audience

Applications requiring haptic flows are expected to grow in importance for 6G networks. While the haptic codec has been recently standardized and is based on the Just Noticeable Difference (JND) principle, performance models and associated network control mechanisms are still lacking. We propose in this work a queuing theoretical control framework to minimize the cost of haptic packet delivery, in terms of network resources, while respecting constraints on consecutive packet losses, an essential metric for such applications. The framework switches between different control policies, leveraging resource reservation and/or power control, depending on the network conditions. Each set of policies is parameterized, reflecting a trade-off between system cost and performance of the haptic service. Through this adaptive approach, we derive closed-form solutions that minimize the cost while ensuring adherence to the loss constraints.

We focus on the deployment of large scale Edge Computing (EC) infrastructure, which requires substantial capital and operational costs. The Infrastructure Providers (InPs) are reluctant to make these investments, as the revenues from edge services, e.g., augmented reality, may be mainly captured by the respective Service Providers (SPs), and not the InPs. This is the main reason for the limited deployment of EC. We tackle this economic aspect by proposing a co-investment strategy, in which all players, i.e., one InP and multiple SPs, form a coalition to deploy an optimal amount of resources, dynamically allocate them over an investment period, and fairly share costs and revenues. The major challenge is to ensure that the coalition is stable, i.e., the co-investment is profitable for all players under uncertainty of revenues, which comes from the uncertainty of future user demand. We address this challenge with a novel stochastic coalitional game formulation, which allows us to analytically compute a lower bound on the probability that the grand coalition is stable. Numerical results show a large value for the lower bound, even with high levels of uncertainty, provided that the number of SPs is relatively small, their loads are comparable, and the investment period is long enough.

We study resource allocation for a metaverse user in 5G networks and beyond. To ensure an immersive experience, one should consider the multi-modality nature of the metaverse, where each user generates multiple coupled flows, namely visual and haptic, characterized by joint Quality of Service (QoS) requirements, for instance in terms of subjective Just Noticeable Difference (JND) metric. These flows can be transported via various 5G services, such as Ultra Reliable Low Latency Communications (URLLC) for the haptic flow and enhanced Mobile Broadband (eMBB) for the visual one. Furthermore, the metaverse user has to share the radio resources with classical eMBB users, characterized by an elastic nature. We formulate an optimization problem that determines the optimal resource sharing between flows, under various performance constraints. We show how to solve this problem in real-world scenarios where there is a discrete set of modulation and coding schemes (MCSs) and considering the various characteristics of the different flows.

We consider a Network Operator (NO) that owns Edge Computing (EC) resources, virtualizes them and lets third party Service Providers (SPs) run their services, using the allocated slice of resources. We focus on one specific resource, i.e., cache space, and on the problem of how to allocate it among several SPs in order to minimize the backhaul traffic. Due to confidentiality guarantees, the NO cannot observe the nature of the traffic of SPs, which is encrypted. Allocation decisions are thus challenging, since they must be taken solely based on observed monitoring information. Another challenge is that not all the traffic is cacheable. We propose a data-driven cache allocation strategy, based on Reinforcement Learning (RL). Unlike most RL applications, in which the decision policy is learned offline on a simulator, we assume no previous knowledge is available to build such a simulator. We thus apply RL in anonlinefashion, i.e., the model and the policy are learned by directly perturbing and monitoring the actual system. Since perturbations generate spurious traffic, we thus need to limit perturbations. This requires learning to be extremely efficient. To this aim, we devise a strategy that learns an approximation of the cost function, while interacting with the system. We then use such an approximation in a Model-Based RL (MB-RL) to speed up convergence. We prove analytically that our strategy brings cache allocation boundedly close to the optimum and stably remains in such an allocation. We show in simulations that such convergence is obtained within few minutes. We also study its fairness, its sensitivity to several scenario characteristics and compare it with a method from the state-of-the-art.

The commercial adoption of Edge Computing (EC) will require pricing schemes that cater to the financial interests of the operators and of the users. Pricing in EC is particularly challenging as it has to take into account the limited amount of edge resources as well as the stochasticity of user workloads due to location-specific workload characteristics and differences in user activity. We formulate the problem of maximizing the revenue of a serverless edge operator through dynamically pricing compute and memory resources under time varying workloads as a sequential decision making problem under uncertainty. We provide analytical results for the optimal pricing strategy in a Markovian setting in steady state. For the general case, we propose a novel Generalized Hidden Parameter Markov Decision Process (GHP-MDP) formulation of the revenue maximization problem, and we propose a dual Bayesian neural network approximator as a solution. The key novelty of the proposed solution is that it can be pre-trained on synthetic traces and adapts fast to previously unseen workload characteristics. We use simulations based on synthetic and real traffic traces to show that the proposed solution is sample-efficient thanks to effective transfer learning, and it outperforms state-of-the-art learning approaches in terms of revenue and learning rate by up to 50% on real traces.

We address radio resource allocation for the transport of extreme Ultra Reliable Low Latency and Reliability (URLLC) traffic. One illustrative use case is factory automation using 6G networks. In this context, extreme URLLC has very stringent Quality of Service (QoS) requirements: 0.1 ms for the delay and $10^{-7}$ for the reliability. Reliability can be even higher for other use cases. We model QoS in terms of outage probability, that is the likelihood of failing to serve at least one packet due to insufficient resources, and derive the minimal resource reservation that would meet such requirement. We formulate the problem as chance-constrained optimization and solve it assuming partial knowledge of arriving traffic distribution. We treat the case where traffic is described through its mean and variance, and make use of three approaches to find the optimal solution: distributionally robust using worst-case value at risk approach, distribution-based approximation and bounds from large deviation theory. We also solve the optimization problem using a data driven approach and propose a sliding window mechanism to perform it online. We compare the performance of the aforementioned approaches numerically and show the effectiveness of the data driven approach, accounting for user radio condition heterogeneity and thus different Modulation and Coding Schemes.

We propose a novel resource allocation framework for latency-critical traffic, namely Ultra Reliable Low Latency Communications (URLLC), in mobile networks which meets stringent latency and reliability requirements while minimizing the allocated resources. The Quality of Service (QoS) requirement is formulated in terms of the probability that the latency exceeds a maximal allowed budget. We develop a discrete-time queuing model for the system, in the case where the URLLC reservation is fully-flexible, and when the reservation is made on a slot basis while URLLC packets arrive in mini-slots. We then exploit this model to propose a control scheme that dynamically updates the amount of resources to be allocated per time slot so as to meet the QoS requirement. We formulate an optimization framework that derives the policy which achieves the QoS target while minimizing resource consumption and propose offline algorithms that converge to the quasi optimal reservation policy. In the case when traffic is unknown, we propose online algorithms based on stochastic bandits to achieve this aim. Numerical experiments validate our model and confirm the efficiency of our algorithms in terms of meeting the delay violation target at minimal cost.

We propose in this paper a coinvestment plan between several stakeholders of different types, namely a physical network owner, operating network nodes, e.g. a network operator or a tower company, and a set of service providers willing to use these resources to provide services as video streaming, augmented reality, autonomous driving assistance, etc. One such scenario is that of deployment of Edge Computing resources. Indeed, although the latter technology is ready, the high Capital Expenditure (CAPEX) cost of such resources is the barrier to its deployment. For this reason, a solid economical framework to guide the investment and the returns of the stakeholders is key to solve this issue. We formalize the coinvestment framework using coalitional game theory. We provide a solution to calculate how to divide the profits and costs among the stakeholders, taking into account their characteristics: traffic load, revenues, utility function. We prove that it is always possible to form the grand coalition composed of all the stakeholders, by showing that our game is convex. We derive the payoff of the stakeholders using the Shapley value concept, and elaborate on some properties of our game. We show our solution in simulation.

We propose a new performance model for transport of time-critical Ultra Reliable Low Latency Communications (URLLC) traffic in 5G networks and Beyond and apply it to dimensioning of such systems. The Quality of Service (QoS) requirement is formulated in terms of an outage probability which is defined as the probability that the latency exceeds a maximal allowed budget, and which should be kept very low. We develop a generic queuing model to compute this outage probability and adapt it to integrate the specificity of the 5G radio interface, taking into account the heterogeneity of users radio conditions and thus their Modulation and Coding Schemes (MCS) as well as retransmissions due to errors on the radio link. We also propose a low complexity method to calculate it using a geometric tail approach to approximate the tail distribution of the queue, for relevant arrival distributions: Poisson and Binomial. We show numerically the performance of our exact model and approximation and that they yield very accurate performance against simulations, and in comparison with other models from the state of the art. We also show the system dimensioning in terms of required resources to satisfy the outage constraint.

We consider in this paper multiple 5G base stations (BSs) implementing Advanced Sleep Modes (ASM) wherein each base station is able to deactivate some of its components when it does not transport any traffic and save thus energy. Thanks to so-called lean carrier, ASM define four levels of sleep, the deeper the level the larger the energy gain but the more delay to wake-up and serve the incoming user. We specifically study this energy saving versus delay performance trade-off taking into account the effect of inter-cell interference and its impact on whether to wake-up and serve the transmission request immediately upon arrival or to continue to sleep; this latter decision is a main novelty of our work. We treat the case where arrivals of those requests are unknown and a reinforcement learning agent is implemented in each BS in order to (selfishly) derive the optimal sleep policy that achieves a target energy saving versus delay performance trade-off. Our results show the optimal policies in terms of the value of the timer after which the BS goes into sleep, the time spent in each sleep level, and whether the BS should continue to sleep or wake up immediately upon request arrival. We eventually show the corresponding achieved power saving and delay performance.

Edge Computing (EC) allows users to access computing resources at the network frontier, which paves the way for deploying delay-sensitive applications such as Mobile Augmented Reality (MAR). Under the EC paradigm, MAR users connect to the EC server, open sessions and send continuously frames to be processed. The EC server sends back virtual information to enhance the human perception of the world by merging it with the real environment. Resource allocation arises as a critical challenge when several MAR Service Providers (SPs) compete for limited resources at the edge of the network. In this paper, we consider EC in a multi-tenant environment where the resource owner, i.e., the Network Operator (NO), virtualizes the resources and lets SPs run their services using the allocated slice of resources. Indeed, for MAR applications, we focus on two specific resources: CPU and RAM, deployed in some edge node, e.g., a central office. We study the decision of the NO about how to partition these resources among several SPs. We model the arrival and service dynamics of users belonging to different SPs using Erlang queuing model and show that under perfect information, the interaction between the NO and SPs can be formulated as a sub-modular maximization problem under multiple Knapsack constraints. To solve the problem, we use an approximation algorithm, guaranteeing a bounded gap with respect to the optimal theoretical solution. Our numerical results show that the proposed algorithm outperforms baseline proportional allocation in terms of the number of sessions accommodated at the edge for each SP.

We consider in this work Edge Computing (EC) in a multi-tenant environment: the resource owner, i.e., the Network Operator (NO), virtualizes the resources and lets third party Service Providers (SPs - tenants) run their services, which can be diverse and with heterogeneous requirements. Due to confidentiality guarantees, the NO cannot observe the nature of the traffic of SPs, which is encrypted. This makes resource allocation decisions challenging, since they must be taken based solely on observed monitoring information.We focus on one specific resource, i.e., cache space, deployed in some edge node, e.g., a base station. We study the decision of the NO about how to partition cache among several SPs in order to minimize the upstream traffic. Our goal is to optimize cache allocation using purely data-driven, model-free Reinforcement Learning (RL). Differently from most applications of RL, in which the decision policy is learned offline on a simulator, we assume no previous knowledge is available to build such a simulator. We thus apply RL in an online fashion, i.e., the policy is learned by directly perturbing the actual system and monitoring how its performance changes. Since perturbations generate spurious traffic, we also limit them. We show in simulation that our method rapidly converges toward the theoretical optimum, we study its fairness, its sensitivity to several scenario characteristics and compare it with a method from the state-of-the-art. Our code to reproduce the results is available as open source.1

We address the problem of a Network Operator (NO) owning limited resources at the network edge. The NO wishes to enable advanced services, by virtualizing and allocating such resources among multiple tenants, i.e., third-party Service Providers, co-existing at the edge, with different Quality of Service (QoS) constraints. The NO applies a resource allocation policy with the objective of minimizing energy consumption via switching off non-used resources while guaranteeing tenants QoS requirements. We propose a resource allocation policy based on Markov Decision Processes (MDP). In simulation we show that our policy is able to reduce energy consumption, by turning off unused resources, while meeting heterogeneous SP requirements. Our code is available as open source.

In this paper, we consider resource allocation for vehicular safety traffic. In 5G, this traffic is carried by using the Ultra-Reliable and Low-Latency Communications (URLLC) service as it needs stringent Quality of Service (QoS) requirements in terms of latency and reliability. Since URLLC services may require specific numerology and/or channel access and retransmission strategies, network slicing has been proposed as a solution for QoS requirements and its coexistence with other services such as enhanced Mobile Broad-Band (eMBB). In order to accommodate URLLC traffic, one can opt for static resource reservation, however this is not optimal as it does not follow the real URLLC traffic present in the cell and can impact negatively eMBB traffic. Reactive, on-demand resource reservation is not feasible either as it requires reconfiguration which introduces extra delay that makes it prohibitive to meet URLLC delay requirements. This paper proposes proactive resource reservation schemes that anticipate slice demand. Resource reservation is computed per gNodeB based on the expected traffic and radio conditions. We show how field measurements and trajectory predictions can be used to achieve URLLC objectives with low impact on eMBB performance.

The analysis of the techno-economic interactions among the mobile operators and tenants (slices) in the context of 5G network slicing has lately received growing interest by the research community. In particular the problems of admission control, resource allocation/scheduling and pricing for network slicing are not trivial as they affect the profits of the involved stakeholders such as the operator (slice provider) and the network slice owners (tenants or service providers) and therefore the viability of the slice market. Since such problems are entwined, we propose here a novel optimization framework that jointly addresses admission control, resource allocation and pricing. In the proposed framework, the operator owns a fixed amount of resources whereas each slice is characterized by a stochastic demand and utility function reflecting its Quality of Service (QoS) requirements. We have considered two types of slices, one with low traffic profile and deterministic QoS, representing Ultra Reliable Low Latency Communications (URLLC), and one with higher traffic profile and statistical QoS, representing enhanced Mobile BroadBand (eMBB). We devise several models based on whether we allow statistical multiplexing and whether we make use of different prices for different types of slices. We also evaluate the impact of the description of traffic in an aggregate way compared to a detailed one. The outcomes of all variants are analyzed through some numerical experiments which enable us to assess the main flavors of each model.

Critical services in wireless networks opt for performance objectives depending on their application. For in-stance, the defined class of services Ultra-Reliable Low-Latency Communication (URLLC) in 5G networks targets machine automation applications and aims to achieve a high reliability during small latency. A second set of critical services, used generally in monitoring applications, where the receiver is interested in getting as fresh information as possible, aims to minimize the age of the received packets. In this paper, we study the transport of these two sets of services over unlicensed spectrum. We propose a transmission scheme which adapts the transmission power depending on the packet status, in order to achieve the target performance objective while preserving energy. We show that different objectives lead to opposite policies with either increasing or decreasing power levels with respect to delay.

Ultra-Reliable Low Latency Communications (URLLC) is a key service in fifth generation (5G) networks, that requires stringent Quality of Service (QoS) in terms of latency and reliability. As URLLC services may require specific numerology and/or specific channel access and re-transmission strategies, network slicing has been proposed as a solution for multiplexing them with other services such as enhanced Mobile Broadband (eMBB). Once the URLLC slice is configured and resources are dimensioned and allocated to it, URLLC performance targets should be attained thanks to the 5G New Radio (NR) low latency and high reliability features. However, in vehicular services such as safety message exchange, URLLC slice resource dimensioning cannot be static due to the varying number of vehicles in the cell. We show in this paper how the delay for slice reconfiguration alters the URLLC performance and propose a proactive resource reservation scheme that anticipates slice needs and allows ensuring URLLC targets. In order to reduce the impact of this proactive reservation on eMBB performance, we make use of vehicle trajectory prediction and show that limiting anticipated reservation to fewer cells allows reaching the target URLLC QoS with a limited degradation of the network capacity.

We consider in this paper a wireless network powered by a smart grid. The base stations are endowed with backup battery in order to prevent possible power failures in the grid. The battery can also be used for arbitrage purposes wherein the operator can purchase electricity from the grid with a certain price and store it in order to use it later when prices go up. Our aim is to design, on a daily basis, an optimal energy storage strategy, taking into account the electricity price fluctuations in order to reduce the operator's daily energy acquisition cost while satisfying the users traffic requirements and a maximal number of cycles to be performed per day in order to extend the battery lifetime. We do so using a Markov Decision Process (MDP) formulation. With respect to the battery type (Lithium, Lead-acid), cycle and calendar life, we set, on a long-term basis, an optimal battery cycling strategy using a simple exhaustive search algorithm. The aim of this long-term policy is to optimize the operator's return on investment taking into account the battery performance degradation, the increase of user traffic and the seasonality of the electricity prices.

In this paper, we exploit the use of unlicensed spectrum in industrial scenarios for UltraReliable Low-Latency Communication (URLLC). We consider a set of automated machines communicating with a central base station and competing to the radio resources among themselves and with enhanced Mobile Broadband (eMBB) traffic. Observing the impact of Listen-Before-Talk (LBT) on the stringent reliability and delay requirements, we propose a progressive bilevel power transmission policy for URLLC, where a high power transmission can be regarded as a preemption of eMBB packets with lower power. We formulate the objective as a stochastic optimization problem under constraints and solve it in the simple Aloha-like case, and in the more complex LBT case. We then use this model to build an online (re-)transmission algorithm which enables the URLLC stations to determine in a distributed manner the optimal policy using observations of the environment parameters.

We study in this paper the coexistence of Ultra Reliable Low Latency Communications (URLLC) and enhanced Mobile Broadband (eMBB) services in unlicensed spectrum for the uplink transmission in a 5G smart-factory scenario. We first model the medium access for both services coexisting in unlicensed spectrum and evaluate their performance metrics: high reliability and stringent delay budget for URLLC and throughput for eMBB. The results show that URLLC requirements cannot be met even for low eMBB traffic load. In order to cope with this, we explore a preemptive approach where URLLC packets are transmitted with high power when their delay approaches the delay constraint, increasing their chance of being successfully received. This approach enhances URLLC performance, with a little impact on that of eMBB. We finally show that with a good calibration of some system parameters can lead to an optimal performance for both services.

We focus in this paper on the transport of critical services in unlicensed spectrum, where stringent constraints on latency and reliability are to be met, in the context of Ultra-Reliable Low Latency Communication (URLLC). Since contention-based medium access performs poorly in the case of high traffic load, we propose a new transmission scheme where the transmitter can increase its transmission power when the delay of the packet approaches the delay constraint, increasing by that its chance of being decoded even in case of collision with other lower-power packets. We are however interested in minimizing the usage of high power transmissions, mainly to conserve energy for battery-powered devices and to limit the range of interference. Therefore, we define a transmission policy that makes use of a delay threshold after which the high-power transmission starts, and propose a new online-learning approach based on Multi-Armed Bandit (MAB) in order to identify the policy which achieves minimum energy consumption while guaranteeing reliability. However, we observe that the MAB converges slowly to the optimal policy because the loss event is rare in the load regime of interest. We then propose a model-aided learning approach where a simple analytical model helps estimating the longterm reliability resulting from an action and thus its reward. Our results show a significant enhancement of the convergence towards the optimal policy.

We consider in this paper the transport of Ultra-Reliable Low-Latency Communication (URLLC) uplink traffic over unlicensed spectrum. We specifically consider Licensed Assisted Access (LAA) system and study the feasibility of the strict reliability and delay requirements by the means of an exact formulation which incorporates the use of a timer tracking the lifetime of each packet and identifying delay budget violation. When the delay constraint is too tight and traffic density is high, unlicensed spectrum alone is not sufficient, hence we propose the use of 5G licensed spectrum to compensate for the lacking resources. We incorporate this usage of licensed spectrum for delayed packets in the model and show how to calculate performance metrics for a given amount of additional licensed resources using grant-free allocation. We validate our analytical models with simulations and dimension the joint unlicensed/licensed system resources so as to meet both reliability and delay constraints.

In this paper, we propose a truthful combinatorial auction for the joint radio and processing resource allocation problem in the context of a Cloud-based Radio Access Network (C-RAN). We formulate the auction as an Integer Linear Program (ILP), taking into accurate account interference constraints while leveraging radio resource reuse to generate an optimal revenue for the RAN operator. Then, we propose Truthful Greedy Approach (TGA), an effective and truthful heuristic that guarantees a close-to-optimum revenue compared to the one obtained with the ILP formulation. Extensive simulations, conducted in representative network scenarios, compare and evaluate our auction with state-of-the-art approaches from the literature, showing its effectiveness.

We study in this paper optimal control strategy for Advanced Sleep Modes (ASM) in 5G networks. ASM correspond to different levels of sleep modes ranging from deactivation of some components of the base station for several micro-seconds to switching off of almost all of them for one second or more. ASMs are made possible in 5G networks thanks to the definition of so-called lean carrier radio access which allows for configurable signaling periodicities. We model such a system using Markov Decision Processes (MDP) and find optimal sleep policy in terms of a trade-off between saved power consumption versus additional incurred delay for user traffic which has to wait for the network components to be woken-up and serve it. Eventually, for the system not to oscillate between sleep levels, we add a switching component in the cost function and show its impact on the energy reduction versus delay trade-off.

Advanced Sleep Modes (ASMs) are defined as a progressive shutdown of the Base Station (BS) depending on the activation and the deactivation times of the different components. This transition duration defines different levels of sleep modes that can be implemented in future 5G networks. We propose in this paper a management strategy based on Q-learning approach which will enable to find the best combination and durations of ASM levels depending on the traffic load and the network operator's policy regarding energy reduction versus latency. Our results show that even in delay-sensitive scenarios, high energy gains can be achieved in low and moderate traffic loads, respectively 55% and 10%, without inducing an extra latency. Starting from a certain traffic load (approximately 30%), ASM should not be implemented in case of stringent latency constraint.

We investigate in this paper the assessment of the energy efficiency of the service categories provided by a network equipment. We consider five service categories, namely, Streaming, Web, Download, other data services and Voice. We consider two scenarios: a case when the network equipment has no sleep mode feature, and another one when it is put into sleep mode during idle periods. We share the responsibility of the service categories in the power consumption of the network equipment, which represents the share of the power consumption imputed to each service category, and then compute the energy efficiency as the ratio of the traffic throughput to the imputed power consumption. The power consumption of a network equipment is composed of two components: a variable component that is consumed proportionally to its load, and a fixed component consumed irrespective of the traffic. We share the former among the service categories proportionally to their traffic proportions. The latter is shared according to the scenario. In the first one, it is shared with our Shapley-based model introduced in a previous work [1]. In the second scenario, we introduce a new sharing model, also based on Shapley value, which takes into account the sleep mode feature. Our results, applied to a real dataset extracted from an operational network in Europe, show that the former sharing model is better than the sleep-mode-oriented model in terms of fairness, although it does not consider the sleep mode feature of the equipment.

We focus in this paper on energy management strategies for mobile networks equipped with battery storage capacity and powered by a smart grid. We consider the case where the network operator signs an incentive based demand response contract with the energy provider, in which case the former should stop using the grid if the latter tells it to do so, typically in periods of high energy consumption. The operator gets in this case a premium, e.g. in the form of a discount on its energy bill. However, if the operator fails to comply with the energy reduction requests, it would pay a penalty. We formulate the problem using Markov Decision Process (MDP) and apply dynamic programming algorithm to devise an optimal offline policy which minimizes the network operator energy expenditure, by indicating to it when to buy energy and when to operate its network on the smart grid or on its own energy storage. We also show how the operator should manage its battery in order to respond to the energy provider requests to stop consuming from the grid at high electricity load.

Advanced Sleep Modes (ASMs) correspond to a gradual deactivation of the Base Station (BS)'s components in order to reduce its Energy Consumption (EC). Different levels of Sleep Modes (SMs) can be considered according to the transition time (deactivation and activation durations) of each component. We propose in this paper a management solution for ASMs based on Q-learning approach. The target is to find the optimal durations for each SM level according to the requirements of the network operator in terms of EC reduction and delay constraints. The proposed solution shows that even with a high constraint on the delay, we can achieve high energy savings in a low load scenario (up to 57% of EC reduction) without inducing any impact on the delay. When the delay constraint is relaxed, we can achieve up to almost 90% of energy savings.

A network is a common resource typically shared by several services, and so, it is important to determine the share of each service in the total energy consumed by the network. The network energy consumption is composed of a variable component which is consumed to serve traffic, and a fixed component consumed irrespective of traffic. The share of a service in the variable energy consumption equals its traffic proportion as this energy component is load-dependent. In this paper, we focus on the fixed energy consumption and share the responsibility of the services transported by a end-to-end network in its fixed energy consumption. To do so, we use our model based on the cooperative game concept Shapley value introduced in a previous work [1] and apply it first to traffic, then to so-called useful outputs. According to ETSI, the useful output of an equipment is defined to be its maximum capacity and is expressed as the number of Erlangs, packets/s, subscribers, or simultaneously attached users. We consider seven network blocks composing the end-to-end path, namely, the radio access, the fixed access, the aggregation, the mobile core, the registers, the IP core and the service platforms. We consider five service categories: Streaming, Web, Download, other data services and Voice. We apply our model on a real data set extracted from an operational network in Europe, and analyze the end-to-end energy consumption assigned to the service categories.

We focus in this paper on energy management strategies for mobile network, equipped with battery storage capacity as well as local energy production capability, and powered by a smart grid. At each time instant, the mobile network operator has to decide whether to operate its network based on its own energy resources or the smart grid ones, with a possibility to sell energy to the smart grid as well. We formulate our problem using Markov Decision Process (MDP) and derive an optimal policy, which minimizes the telecommunication operator energy bill, using dynamic programming algorithm. We show the optimality of our solution by numerical comparison with the case of being exclusively powered by the grid. Our numerical applications allow to further understand when the operator has an incentive to buy energy, whether it is beneficial for him to act as aggregator (energy seller) as well as the size of the battery to deploy.

We consider in this paper a heterogeneous wireless network where the cell sites, known as Remote Radio Heads, are connected to a central processing unit composed of a pool of Base Band Units, so-called VirtualRadio Access Network (V-RAN). We evaluate its performance, in terms of user flow throughput, system stability and fairness under two resource allocation strategies: Dominant Resource Fairness (DRF) and a heuristic based on Proportional Fairness (PF), which we apply locally on each access network; both allocations being applied jointly on radio and processing resources. We consider the case when the V-RAN is restrictive in terms of processing resources as compared to the radio ones, and compare it to the case when the V-RAN is not restrictive. Our results show that locally-applied PF achieves higher throughput and system stability than DRF, which achieves better fairness.

The concept of Advanced Sleep Modes (ASMs) corresponds to a gradual deactivation of the Base Station (BS)'s components in order to decrease the Energy Consumption (EC). We propose in this paper a strategy to implement the ASMs and to manage users whose service requests occur whilst the BS is sleeping. As the BS has to wake up periodically to send signaling bursts, we increase this periodicity in order to extend the sleep durations and hence maximize the energy gains. We study several scenarios with various traffic profiles and different periodicities of signaling bursts and quantify the outcomes of this solution, both in terms of EC reduction and quality perceived by the users (mainly throughput and latency). Our results show that high energy gains (up to 90%) can be obtained using ASMs, albeit some degradation in throughput (ranging from 3% to 19% in low loads depending on the signaling periodicity) and latency (less than 1ms for low periodicities and around 5ms for larger ones).

Energy efficiency of a network, defined as the number of bits transmitted per unit of consumed energy, increases with the traffic load for a constant network capacity. This comes from the fact that energy is composed of two components: a fixed one, consumed by the network regardless of the traffic load, and a variable one, which depends on the traffic load. And so, when traffic load increases, the fixed component gets amortized. However, a network upgrade, namely adding more equipment in the network to fit traffic increase, comes typically with a higher increase in capacity than traffic, at least for a while after the upgrade, as traffic previsions are based on relatively long term projections. Thus, the power consumption of the network would increase faster than the traffic, and energy efficiency would then decrease. We investigate in this work the conditions under which a network upgrade does not deteriorate its energy efficiency. We consider two ways of upgrading a network: either by adding equipment with the same technology or by deploying equipment with another technology, typically more recent and more efficient. We discuss in both cases the number of equipment to be added so that to preserve the network's energy performance.

Today, mobile networks are witnessing an exponential growth of traffic volumes, linked to new services, especially for smart cities and smart-grid. The European Celtic-Plus SooGREEN project, started mid 2015, is targeting to reduce the energy consumption of the services in different mobile architectures in interaction with smart-grid. So GREEN is focused on the services energy consumption modelling and measurement, the dynamic optimization of the mobile access network and of the content delivery, the design of an Energy Efficient Virtualized and Centralized Radio Access Network (RAN), and the bi-directional interaction of the mobile network with the smart-grid. This paper presents insight into the project after its first year, and discusses research trends in green communication networks for the future.

We investigate in this paper the assessment of the energy efficiency of a wireless access network per service category. We consider five categories of service, two categories with high traffic: streaming and web browsing, and three other with lower traffic: download, voice and other minor data services. We introduce two scenarios, one where some services are mandatory, it is typically the case of Voice which is mandatory to be provided due to legal constraints, and another one where there is no mandatory service. We used our Shapley-based sharing model introduced in previous works to share the total energy consumption of the access network among the service categories, and then derive the energy efficiency of each service category as the ratio of its traffic volume (measured in the network) and its energy consumption assessed with our Shapley-based model. We applied the models on a real dataset extracted from an operational network in Europe, and analyze the energy efficiency of each considered service category.

This paper studies backhaul bandwidth aggregation in the context of a wireless local area network composed of two different types of access points: those with spare backhaul capacity (which we term providers) and those in shortage of it (beneficiaries); the aim is to transfer excess capacity from providers to beneficiaries. We model the system as a matching game with many-to-one setting wherein several providers can be matched to one beneficiary and adopt the so-called deferred acceptance algorithm to reach an optimal and stable solution. We consider two flavors, when the beneficiaries are limited in their resource demands and when they are not, and two scenarios, when resources are abundant and when they are scarce. Our results show that the many-to-one setting outperforms the one-to-one case in terms of overall throughput gain, resource usage, and individual beneficiaries satisfaction by up to 50%, whether resources are scarce or abundant. As of the limited versus nonlimited case, the former ensures more fair sharing of spectral resources and higher satisfaction percentage between beneficiaries.

We conduct in this work a comparative study between a greedy versus limited proposals for resource aggregation in the backhaul of wireless local area networks. We consider a setting with several access points, some with excess backhaul capacity (providers), and others in shortage of it (beneficiaries) and investigate two approaches for the distribution of resources between them: a greedy one wherein each beneficiary seeks to maximize the resource it could obtain from the provider and a limited one in which the operator limits the amount of acquired resources to some limit, so as to ensure fairness between beneficiaries. We model the system using a one-to-one matching game, and compare the two approaches in three scenarios: a symmetric one with equal number of providers and beneficiaries, and asymmetric ones with higher number of either providers or beneficiaries. Our results show that the greedy scheme outperforms the limited one when the number of providers is larger than that of beneficiaries, the limited one performs best in the opposite case and ensures fairness between beneficiary access points. In all cases, both approaches outperform the random allocation scheme.

In this paper, the concept of moving small cells in mobile networks is presented and evaluated taking into account the dynamics of the system. We consider a small cell moving according to a Manhattan mobility model which is the case when the small cell is deployed on the top of a bus following a predefined trajectory in areas which are generally crowded. Taking into account the distribution of user locations, we study the dynamic level considering a queuing model composed of multi-class Processor Sharing queues. Macro and small cells are assumed to be operating in the same bandwidth. Consequently, they are coupled due to the mutual interferences generated by each cell to the other. Our results show that deploying moving small cells could be an efficient solution to offload traffic hotspots.

With the exponential growth of data traffic in modern mobile networks and the emergence of variety of connected devices, congestion mitigation in cellular networks has become a major concern. Therefore, in this paper, we propose and extensively describe a new drone-based mechanism allowing to relay traffic and improve network performances. Scenarios related to the appearance, disappearance and mobility of the traffic hotspot are thoroughly specified and developed. The proposed solution is efficient in terms of capacity improvement with a good reactivity to the appearance of traffic hotspots. Moreover, the deployed drones are very energy-efficient in the way they operate in the proposed mechanism. This encourages more and more operators to deploy such solutions leveraging their investments.

We investigate in this paper the sharing of energy consumption among service categories in the access of a wireless network. We focus on the fixed part of the energy consumption of the network, which is known to be significantly larger than the load-dependent variable part, and propose its sharing among the service categories based on coalition game concept, the Shapley value. We consider five service categories, two large players: streaming and web browsing, and three smaller ones: download, voice and other minor services, and compare our proposal with two other sharing strategies: uniform and proportional which follows the same volume proportions. Our results, applied on a real dataset extracted from an operational network in Europe, show that our proposal is more fair both towards small services in that it reduces their shares in comparison to the uniform approach, and towards larger services as it reduces their shares in comparison with the proportional one. Indeed, our Shapley-based model accommodates both short term network behavior, in which the fixed energy component is independent of the traffic load, and longer term behavior, in which it varies with the load and infrastructure. Uniform sharing accounts only for the short term, and the proportional one only for the longer term.

We study in this work user impatience by adopting a realistic realization: a user gets impatient when its throughput falls below a certain threshold for a given patience duration. We model such a behavior in a cellular network, transporting elastic flows, for a realistic user configuration with random arrivals, following a Poisson process, and departure after a finite duration, either upon transfer completion or because of impatience. We derive several performance metrics, pertaining to the probability of abandoning the transfer as well as transport efficiency in terms of traffic efficiently transported until service completion versus wasted resources because of abandonment. We further implement a control mechanism using two approaches: a preventive one, based on blocking of users when their number exceeds a certain threshold, and a reactive one, based on dropping of users when the network experiences congestion, and show how the system performance gets enhanced in both cases.

This paper presents an analytical performance evaluation framework for 4G/WiFi multihoming techniques. Several multihoming flavors have been compared, with a focus on user centric allocation strategies. In these strategies the user splits his packets between two radio interfaces following a policy that depends on the amount of information he gets. First, the simplest policy is considered in which the user sends an amount of traffic on each interface proportional to the latter's peak rate. Second, the optimal selfish policy is investigated in which the user has complete information about the traffic intensities. This latter policy has been proved to achieves a global optimum. The numerical results show that an optimized user centric policy achieves good gains for both the system capacity and the user throughput compared to a peak maximization strategy. The results also show that multihoming provides large performance gains for multi-homed users while degrading neither the capacity of the system nor the performance of single-homed users.

We study in this work the quality of experience of live mobile TV users in the presence of a playout buffer at the receiver side. We specifically consider an LTE network delivering both unicast and broadcast services, and study two cases: i. lossless channels where interruptions in mobile TV delivery are due to the absence of resources at the base station because they are totally consumed by higher priority real-time voice services, and ii. lossy channels, where we propose to exploit the existence of feedback channels to request, if possible, retransmission of the lost video frame. We derive in both cases several QoS metrics pertaining to mobile TV performance, such as interruption frequency and duration as well as video frame loss probability. Our results show the trends of these metrics as a function of several system parameters and yield design rules for network planning as well as for the playout buffer so as to enhance efficiently the live TV watching experience.

We develop in this paper a trend detection algorithm, designed to find trendy topics being disseminated in a social network. We assume that the broadcasts of messages in the social network is governed by a self-exciting point process, namely a Hawkes process, which takes into consideration the real broadcasting times of messages and the interaction between users and topics. We formally define trendiness and derive trend indices for each topic being disseminated in the social network. These indices take into consideration the time between the detection and the message broadcasts, the distance between the real broadcast intensity and the maximum expected broadcast intensity, and the social network topology. The proposed trend detection algorithm is simple and uses stochastic control techniques in order to calculate the trend indices. It is also fast and aggregates all the information of the broadcasts into a simple one-dimensional process, thus reducing its complexity and the quantity of data necessary to the detection.

Heterogeneous Networks (HetNets), long been considered in operators' roadmaps for macrocells' network improvements, still continue to attract interest for 5G network deployments. Understanding the efficiency of small cell deployment in the presence of traffic hotspots can further draw operators' attention to this feature. In this context, we evaluate the impact of imperfect small cell positioning on the network performances. We show that the latter is mainly impacted by the position of the hotspot within the cell: in case the hotspot is near the macrocell, even a perfect positioning of the small cell will not yield improved performance due to the interference coming from the macrocell. In the case where the hotspot is located far enough from the macrocell, even a large error in small cell positioning would still be beneficial in offloading traffic from the congested macrocell.

The performance of Delay Tolerant Mobile Networks (DTMNs) depends on the mobility pattern through the nodes encounter process which constitutes opportunities for exchange of information between nodes. However, the exchange of information itself depends also on the routing/replication protocol and the network resources as the node buffer capacity and which determines the effective network performance. We focus in this work on those two concepts: encounter and exchange, and evaluate them along with resulting network performance for different mobility patterns and routing protocols as well as under the limited buffer space constraint. We also propose the use of additional, mobile, dedicated relays so as to enhance the performance of the DTMN in cases of ill-behaved mobility schemes.

The performance of Delay Tolerant Mobile Networks (DTMNs) depends on the mobility pattern through the nodes encounter process which constitutes opportunities for exchange of information between nodes. The effective exchange of information depends also on the routing/replication protocol and on several network resources such as buffer capacity, radio propagation conditions and energy. We focus in this work on those two concepts: encounter and exchange, and evaluate them along with resulting network performance for different synthetic mobility patterns and routing protocols, as well as under several constraints on the radio conditions and buffer space. We also propose the use of additional, mobile, dedicated relays so as to enhance the performance of the DTMN in cases of ill-behaved mobility schemes.

In this paper, we propose a new scheme for aggregating access points backhaul bandwidth resources in wireless local area networks. We consider a setting with several access points, some with excess backhaul capacity, and others in shortage of it, and propose a symmetric approach wherein the distribution of resources is managed between providing and beneficiary access points, without involving the direct participation of the users. We then model the proposal using one-to-one matching game theory and make use of the so-called deferred access (DA) algorithm to reach an optimal, stable solution for the problem. In comparison with random distribution of available backhaul resources, the DA scheme offers significantly better performance in terms of overall throughput, reaching up to 40% when compared to random matching.

Cooperation in multi-access wireless networks have become an effective strategy for answering the rising data rate requests. The interworking is performed either by traffic steering, load balancing or yet radio resource aggregation, the so-called multi-homing. While most studies focused mainly on the first two techniques, we focus in this work on multi- homing, and target principally LTE and WiFi access networks. We specifically consider the co-existence of several classes of users in the system: single-homed users and multi- homed users, and study two network-centric, resource allocation strategies: proportional fairness applied to the global network as a whole or to each wireless access network in isolation of the other. We evaluate the system performance and show that both strategies improve the achievable throughput and stability conditions by applying a load balancing between accesses, thus increasing the overall network capacity.

We consider a delay tolerant network under two message forwarding schemes-a non-replicative direct delivery scheme and a replicative epidemic routing scheme. Our objective is to track the degree of spread of a message in the network. Such estimation can be used for on-line control of message dissemination. With a homogeneous mobility model with pairwise i.i.d. exponential inter-meeting times, we rigorously derive the system dynamic and measurement equations for optimal tracking by a Kalman filter. Moreover, we provide a framework for tracking a large class of processes that can be modeled as density-dependent Markov chains. We also apply the same filter with a heterogeneous mobility, where the aggregate inter-meeting times exhibit a power law with exponential tail as in real-world mobility traces, and show that the performance of the filter is comparable to that with homogeneous mobility. Through customized simulations, we demonstrate the trade-offs and provide several insightful observations on how the number of observers impacts the filter performance.

In recent years, there has been an increasing awareness to traffic localization techniques driven by the emergence of heterogeneous networks (HetNet) with small cells deployment and the green networks. The localization of hotspot data traffic with a very high accuracy is indeed of great interest to know where the small cells should be deployed and how can be managed for sleep mode concept. In this paper, we propose a new traffic localization technique based on the combination of different key performance indicators (KPI) extracted from the operation and maintenance center (OMC). The proposed localization algorithm is composed with five main steps; each one corresponds to the determination of traffic weight per area using only one KPI. These KPIs are Timing Advance (TA), Angle of Arrival (AoA), Neighbor cell level, the load of each cell and the Harmonic mean throughput (HMT) versus the Arithmetic mean throughput (AMT). The five KPIs are finally combined by a function taking as variables the values computed from the five steps. By mixing such KPIs, we show that it is possible to lessen significantly the errors of localization in a high precision attaining small cell dimensions.

We propose in this paper, some practical algorithms to deliver Mobile TV in a hybrid network where coexist both LTE and DVB-NGH technologies and two types of demanded video qualities: a basic single definition and another high definition. Those algorithms aim to manage the continuous reception of Mobile TV by using both infrastructures and taking decisions of TV scheduling on those operators. Two decisions policies are explored: one based on instantaneous actions and a second based on predictive actions. Results shows that the latter, which uses the Kalman filter, decreases the ping-pong between the HD reception delivered by LTE and the SD one offered by DVB-NGH, if compared to the former policy.

This paper tackles the issue of reliable transport in delay-tolerant mobile ad hoc networks (DTNs) that are operated by some opportunistic routing algorithm. We propose a reliable transport mechanism that relies on acknowledgements (ACKs) and coding at the source. The various versions of the problem depending on buffer management policies are formulated and a fluid model based on mean-field approximation is derived for the designed reliable transport mechanism. This model allows both the mean file completion time and the energy consumption to be expressed up to the delivery of the last ACK at the source. The accuracy of this model is assessed through numerical simulations and a detailed investigation of the impact of the system parameters on the performance is conducted. We eventually present a joint optimization of the mean completion delay with or without an energy constraint, to identify the optimal set of parameters to use.

The LTE/DVB convergence to offer users a seamless Mobile TV service is at the heart of the emerging DVB-NGH standard. Most of the works addressing this issue focused solely on the impact of such convergence in terms of capacity gains brought by LTE and coverage gains brought by DVB. To the best of our knowledge, the impact of such convergence on the Mobile TV users' QoS has not been addressed yet and is the focus of our present work. We specifically evaluate the users' QoS in terms of quantifiable metrics, notably the probability of service degradation due to non availability of LTE bearers, its frequency and duration, and show how these metrics evolve with varying traffic load and in different network configurations.

We study in this work user impatience and quantify its impact on the performance of mobile networks, notably LTE networks, in the presence of data flows experiencing heterogeneous radio conditions. We consider a dynamic user setting where users come to the system at different time instants and leave it after a finite duration, either after completion of their data transfers or earlier, at the expiry of some patience duration. We derive closed-form analytical expressions for the system stationary probability distribution, in general and for some particular distributions for the impatience duration and file size, and obtain several performance metrics such as system load and mean transfer times, taking into account the heterogeneity of users locations in the cell. Our numerical results quantify these metrics, as a function of the impatience rate, as well as the traffic region where the impatience impact becomes preponderant.

Our aim in this work is to develop profit sharing strategies for coooperative 4G Long Term Evolution (LTE) and Digital Video Broadcasting - Next Generation Handheld (DVB-NGH) systems offering mobile TV service. We first study the capacity of the hybrid system for offering such a service and derive the associated cost for each network. We then focus on the sharing of profit between LTE and DVB operators as well as TV channels providers, taking into account the subscription revenues as well as the infrastructure and operation costs. We consider two cases: one in which both LTE and DVB networks are managed by a single operator and one in which the operators are separate. In both cases, we derive closed-form expressions for each player profit share using coalition game concept Shapley value. We further obtain the players optimal strategies, at the Nash equilibrium, where each player tries to maximize its own profit.

We study, in this work, optimal sleep/wake up schemes for the base stations of network-operated femto cells deployed within macro cells for the purpose of offloading part of its traffic. Our aim is to minimize the energy consumption of the overall heterogeneous network while preserving the Quality of Service (QoS) experienced by users. We model such a system at the flow level, considering a dynamic user configuration, and derive, using Markov Decision Processes (MDPs), optimal sleep/wake up schemes based on the information on traffic load and user localization in the cell, in the cases where this information is complete, partial or delayed. Our results quantify the energy consumption and QoS perceived by the users in each of these cases and identify the tradeoffs between those two quantities. We also illustrate numerically the optimal policies in different traffic scenarios.

In this paper, we investigate network sleep mode for reducing energy consumption of radio access networks. We propose an offline-optimized controller that associates to each traffic an activation/deactivation policy that maximizes a multiple objective function of the Quality of Service (QoS) and the energy consumption. We focus on practical implementation issues that may affect the QoS and the stability of the system. We namely consider the activation time issue that results in the degradation of the throughput and the ping-pong effect that results in unnecessary ON/OFF oscillations. We illustrate our results numerically in beyond 3G networks.

We propose a new scheme for reliable transport, both for unicast and multicast flows, in Delay Tolerant Networks (DTNs). Reliability is ensured through the use of Global Selective ACKnowledgements (G-SACKs) which contain detailed (and potentially global) information about the receipt of packets at all the destinations. The motivation for using G-SACKs comes from the observation that one should take the maximum advantage of the contact opportunities which occur quite infrequently in DTNs. We also propose sharing of “packet header space” with G-SACK information and allow for random linear coding at the relay nodes. Our results from extensive simulations of the proposed scheme quantify the gains due to each new feature.

This paper addresses the problem of power allocation of the downlink in cognitive radio networks (CRNs). In a CRN the channel availability for the secondary users (SUs) is a random process determined by primary user (PU) behavior. In an overlay CRN where the PU has the priority and CRN may use only the idle channels, CRN loses the symbols that are being transmitted during that time slot when the PU takes the channel and if the error correction code cannot handle the length of symbol loss. Further, in the next time slot, the CRN has to select again a free channel to continue the communication or in the event that all channels are occupied by PUs and other SUs it enters an indefinite waiting process. Thus it is intuitive that the time averaged data rate of the CRN does not only depend on the channel characteristics (governed by Shannon's law) but also on the channels' PU activity level. CRNs, with the spectrum sensing capabilities inherent to them, have the ability to characterize those activity statistics. In this paper, we propose two schemes named ENPA (equivalent noise of PU activity) and EBPA (effective bandwidth of PU activity), which take into account the PU activity in CRN downlink power allocation. These schemes are shown to achieve higher average rates than the water-filling algorithm which is optimal for a linear filter channel.

In this paper, we show the impact of femtocells deployment on the capacity of LTE-Advanced networks. We analyze the Erlang-like capacity of a network composed of macro networks only and study the impact of introducing femtocells. We study the ability of open and closed access femtocells to offload traffic from the macro network and draw the capacity gains that are expected from this offload. Knowing that femtocells are not managed by the operator, but installed by clients, we take into account the random nature of the resulting deployment in the capacity analysis. Our results quantify the gains thus achieved, in terms of user QoS and cell load, as well as energy efficiency, as compared to a pure macro network.

Relays are typically used to convey information to users who are unable to decode the direct signal from the base station successfully. In the case where the users are able to successfully decode the direct signal without the help of relays, the use of the latter can only decrease their performance because they will consume, uselessly, additional resources. If, however, hierarchical modulation can also be jointly used, the overall performance of a relay-based system can be enhanced even if the users are able to decode the direct signal successfully, without the need for relays. And this is the rationale behind this paper where we study the joint use of hierarchical modulation and relays, and model such a setting for a dynamic user configuration at the flow level. We validate our model against simulations and quantify the gains thus achieved in comparison with a system without relays, with and without hierarchical modulation.

In this paper, we show the impact of deploying picocells on the capacity and energy efficiency of LTE-A networks. We analyze the Erlang-like capacity of a network composed of macro networks only and study the impact of introducing a number of picocells per site. Knowing that the capacity is not the only factor that will drive the evolution of the network, we also consider the energy efficiency as a Key Performance Indicator (KPI). Our results show that, in some scenarios, introducing picocells is a good network densification method as they achieve a higher network capacity with good energy efficiency.

Hierarchical modulation is a means to better use the overall resources of a given system by superposing, in terms of modulation, a user with better radio conditions on one with inferior radio conditions so as to advantageously transfer some resources from the latter to the former. This in turn impacts the overall performance of the system, which we consider in this work at the flow level, for a realistic dynamic setting where users come to the system and leave it after a finite duration corresponding, for instance, to the completion of a file transfer. We quantify, in this work, analytically and via simulations, the gain thus achieved and propose a new scheme in which users with bad radio conditions are also superposed on ones with better radio conditions so as to enhance the system performance even further.

Delay Tolerant networks (DTNs) are one type of wireless networks where the number of nodes per unit area is small and hence the connectivity between the nodes is intermittent. In this case, the performance in terms of transport of information from source to destination relies on the mobility of the nodes which would cause their encounters and hence the relay of information from one node to another as well as on the routing protocol that is deployed. There exists several mobility patterns, each yielding a different performance of the network. In this work, we show first that the Manhattan mobility pattern performs worse than other widely-used ones, such as Random Way Point. In the second part of this work, our aim is to propose and evaluate a new proposal, based on the deployment of fixed relays, so as to enhance the performance of Manhattan.

Troubleshooting of wireless networks is a challenging network management task. We have developed, in a previous work, a new troubleshooting methodology, which we named Statistical Learning Automated Healing (SLAH). This methodology uses statistical learning, in particular logistic regression, to extract the functional relationships between the noisy Key Performance Indicators (KPIs) and Radio Resource Management (RRM) parameters. These relationships are then processed by an optimization engine so as to calculate the optimized RRM parameters which improve the KPIs of a degraded cell. The process is iterative and converges to the optimum RRM parameter value in few iterations, which makes it suitable for wireless networks. The present work focuses on the adaptation of SLAH for troubleshooting the mobility parameter, namely the handover margin, in 3G Long Term Evolution (LTE) networks. The simulation results, which we obtain for a practical use case, show the advantage of this new, automated troubleshooting methodology.

Data communication in Vehicular Ad hoc NETworks (VANETs) tends to be an important challenge mainly due to intermittent connectivity and MAC layer collisions. Such issues hinder reliable and in-order reception of data packets and hence lead to poor performance of some applications like file transferring. In this paper we propose a file exchange scheme based on Fountain coding aiming at improving throughput of the network. In the proposed approach there is no need of in-order reception of packets, thus, vehicles can make benefit of opportunistic delivery of data. As a result we use UDP in contrast to classical use of TCP while reliability is still preserved. Results of extensive simulations show that the proposed approach increases the number of successful file exchanges as well as the average throughput of each vehicle. It also facilitate resumable downloads. Therefore it can be a good candidate for comfort applications in VANETs which mainly rely on file (or any sort of bulk data) exchange.

In this paper, we develop new energy-efficient, radio resource management schemes for green wireless networks. Our goal is to optimize energy consumption at the network scale while preserving the Quality of Service (QoS) perceived by users. We specifically propose two new sleep mechanisms for base station where a number of resources in system can be shut down for some traffic scenarios in one of two ways: a dynamic way where resources are activated/deactivated in real-time as a function of the instantaneous load of the system, and a semi-static one where resources are kept unchanged during longer time intervals, in the order of one hour. We apply the proposed schemes to 2G and HSPA (High Speed Packet Access) systems and show that the dynamic one achieves larger energy reductions while the semi-static one has an acceptable performance with low complexity.

We investigate in this paper the impact of mobility on the capacity of an OFDMA-based mobile WiMAX network in the presence of two types of flows: streaming and elastic, under a dynamic configuration where users come to the system following a Poisson process and leave it after a finite duration. This duration is independent on the amount of resources for streaming flow, such as voice, and proportional to the amount of resources they receive for elastic ones, such as data. Our analysis makes it possible to quantify several performance measures such as blocking/dropping probabilities as well as mean transfer time. Our results show that mobility decreases the blocking rate, introduced some dropping and increases the overall cell throughput.

In this paper, we propose and analyse a new scheme to handle ranging requests in IEEE 802.16e networks based on differentiated backoff parameters and a novel hierarchical code partitioning principle. Our analysis is of the form of a fixed-point equation which characterizes the system operating points and which allows us to compute the attempt rate of each QoS class as well as the distribution of the number of ranging bandwidth requests received successfully by the base station as well as their average delay. Both analysis and simulation confirm that our scheme yields high throughput and low collision probability for real-time connections without altering the collision probability for non-real-time connections.

In this paper, we study the joint design of uplink and downlink resources in OFDMA-based systems. We first analyze the interactions between uplink and downlink, due essentially to the usage of time division duplexing (TDD) and the TCP ACKs. This analysis is based on analytical models for interference and throughputs that we develop for both directions. We next study the capacity of the system using a Markovian analysis and matrix geometric solutions. Our performance analysis study allows us to determine the optimal proportion of resources that has to be allocated to the uplink so as to minimize blocking. In addition to that, we show that an admission control policy reserving some uplink resources to TCP ACKs may improve the overall capacity of the system.

OFDMA-based IEEE802.16 implements Adaptive Modulation and Coding (AMC) which results in a different bit rate for each user depending on its position in the cell as well as the radio condition it experiences; users away from the base station experience lower throughput. We hence propose in this paper a new QoS solution that compensates the degradation in modulation by a higher number of sub-carriers so as to maintain the bit rate of streaming flows at a constant level throughout the whole coverage area. This proposal is modeled analytically in a dynamic user configuration, where users of both types, streaming and elastic, remain in the system for a finite duration. And this in multi-cell environment and for different frequency reuse schemes.

We study in this paper the capacity of the downlink of OFDMA-based IEEE802.16 WiMAX system in the presence of two types of traffic, streaming and elastic. We focus in particular on the impact of Adaptive Modulation and Coding (AMC) as well as inter-cell interference resulting from different frequency reuse schemes. Several performance measures, namely blocking rates, mean transfer time and the mean number of collisions between two OFDMA WiMAX cells, are then derived and quantified. We show that reuse partitioning results in a lower blocking probability for streaming flows in the inner region but a much higher one for elastic flows in the outer region; the overall cell throughput increases meaning that the decrease in the number of collisions improves the overall throughput.

Energy is a major concern in ad hoc networks as it dictates the lifetime of a terminal and consequently the whole network. In an ad hoc network, not all terminals are equally solicited : depending on their position, some serve as relay nodes more than others. In order to maximize the lifetime of the whole network, we propose in this work to give priority in accessing the medium to the ones with larger number of neighbors, as done in Reference [1] but for routing only. In doing so, not only do we save energy for hard working nodes, as energy is mostly used by the MAC access, but we also take advantage at most out of these central nodes before they move to other, less dense locations. As far as the MAC access is concerned, our results show up to 82% increase in the lifetime of the whole network.

We focus in this paper on modeling the integration between streaming and elastic flows in IEEE802.11-based systems. In such a CSMA/CA shared context, both types of flows do not behave in the same way. Streaming flows, characterized by a constant bit rate, are not able to get more than their need; they however suffer a degraded performance when their share is lower than their constant rate. We then model them as non-saturated sources. Data sources are on the contrary saturated ones and share the bandwidth in a fair manner, as given by a processor sharing model. In the presence of the two types of flows, a joint model is hence considered at the MAC layer; the results of which, in terms of QoS for streaming flows and service rates for data ones are then fed into a Markovian flow level analysis.

We focus in this paper on the cross-layer modeling of the integration of streaming and elastic flows in the downlink of UMTS/HSDPA systems. Streaming flows are transported over dedicated channels whereas elastic ones shall share the left over capacity according to opportunistic scheduling implemented at the MAC layer. We first study the system at the MAC layer and then make use of these results at the flow level where the system is studied in a dynamic setting. The quasi-stationary assumption makes it possible to obtain a product form expression for the steady-state probabilities of the system. Flow level dynamics are then used at the packet level to investigate TCP performance and quantify the probability of a spurious event, namely timeout. Several performance metrics are then derived and give hints into dimensioning issues of such a system.

In this work, we study the transport of real-time voice traffic in the IP-based UMTS terrestrial radio access network (UTRAN) where a stringent-delay bound, some 5 ms, is to be met. Our work is analytical and is validated through empirical results on a local experimental test-bed emulating the UTRAN. It shows the trade-offs between performance, in terms of delay and cost, in terms of link utilization, and quantifies optimal values for the timer used in the packetization of frame protocol (FP) frames into larger IP packets as well as the number of FP frames per IP packet for a given output link capacity.

In this paper, we analytically compare IP versus AAL2/ATM as a transport technology in the UMTS terrestrial radio access network (UTRAN), where stringent delay bounds are to be met for both real-time and non real-time traffic. We particularly select voice and Web traffic so as to illustrate our investigation. We show that, for the case of voice traffic. AAL2/ATM has a small advantage over IP, in terms of delay; IP is however more efficient than AAL2/ATM in terms of link utilization, especially for heavy loaded systems. As for data traffic, IP performs best.

In this paper, we focus on the capacity of a multicell UMTS system. We first determine an upper bound on the other cell interference and obtain novel expressions for the SIR and powers for both the uplink and the downlink. The former is an asynchronous CDMA system, often using one of three types of receivers : matched filter, minimum mean-square error (MMSE) and decorrelator. In the latter, the SIR depends on the distance between the user and the base station. We show that the inter-cell interference alters the capacity of the system for all kinds of receivers, though the effective bandwidth expressions remain identical to the ones obtained in the single-cell case. As an application, we show how to use this analytical model to develop a new efficient connection admission control (CAC) algorithm.

In this work, we focus on modeling of TCP in an end-to-end path where both wired and wireless, mobile sections of the network are present. The former is governed by loss and is subject to TCP retransmission. The latter is governed by error and path loss and is under data link layer control error detection/retransmission mechanism as well as UMTS-oriented error correction and power control. We first investigate the end-to-end performance of TCP in the outer-loop, with bit error ratio (BER) as a QoS metric. Second, we consider the inner-loop control case, and relate TCP performance to the signal-to-interference ratio (SIR) and the despread bit energy to interference density ratio E/sub b//N/sub 0/. Third, we review the power control algorithm and propose a novel one basing it on end-to-end TCP performance rather than UMTS channel measurements.

In this paper, we develop an efficient call admission control (CAC) algorithm for UMTS systems. We first introduce the expressions that we developed for signal-to-interference (SIR) for both uplink and downlink, to obtain a novel CAC algorithm that takes into account, in addition to SIR constraints, the effects of mobility, coverage as well as the wired capacity in the LMTS terrestrial radio access network (UTRAN). for the uplink, and the maximal transmission power of the base station, for the downlink. As of its implementation, we investigate the measurement-based approach as a means to predict future, both handoff and new, call arrivals and thus manage different priority levels. Compared to classical CAC algorithms, our CAC mechanism achieves better performance in terms of outage probability and QoS management.

The focus of this paper is the performance of the AAL2/ATM transport technology for the Universal Terrestrial Radio Access Network (UTRAN) in UMTS/IMT2000. We particularly consider the case of voice and data streams, with a service differentiation between them conducted at the AAL2 level by the use of the weighted round robin (WRR) scheme. We obtain closed-form expressions for bounds on the delay violation probabilities for both voice and data frames, taking into account the impact of WRR as well as the traffic patterns inherent to UMTS.

The focus of this paper is the performance of ATM/AAL2 as a transport technology for Universal Terrestrial Radio Access Network (UTRAN) in UMTS/IMT-2000 setting. In this work, we propose an analytical model for the transparent multiplexing and transport through UTRAN of several voice channels onto a single AAL2/ATM VC. We investigate the queue building up at the AAL2 multiplexer in order to derive some characteristics of the sojourn time of voice frames within the UTRAN. Our results quantify the performance of AAL2/ATM-based transport of voice in terms of mean queue at the multiplexer, mean sojourn time and delay violation probability within the UTRAN.

Quality of service (QoS) is a major issue in the recent and current developments of telecommunication networks, notably in the ATM-based B-ISDN or the emerging QoS-capable IP networks. As in most cases where IP runs over ATM, translation of QoS between them, through the translation of their respective performance parameters (PPs), is central to the overall end-to-end performance and is the focus of this framework. We specifically set the end-to-end path which encompasses the major entities of the system, identify the layers of concern and their mechanisms, define inter-protocolar, inter-layer PPs and discuss relevant issues related to their translation between the protocols as well as between the layers, and relate them to the overall end-to-end QoS in IP and ATM.