It has been shown that deploying device-to-device (D2D) networks in urban environments requires equipping a considerable proportion of crossroads with relays. This represents a necessary economic investment for an operator. In this work, we tackle the problem of the economic feasibility of such relay assisted D2D networks. First, we propose a stochastic model taking into account a positive surface for streets and crossroads, thus allowing for a more realistic estimation of the minimal number of needed relays. Secondly, we introduce a cost model for the deployment of relays, allowing one to study operators' D2D deployment strategies. We investigate the example of an uberizing neo-operator willing to set up a network entirely relying on D2D and show that a return on the initial investment in relays is possible in a realistic period of time, even if the network is funded by a very low revenue per D2D user. Our results bring quantitative arguments to the discussion on possible uberization scenarios of telecommunications networks.

Heterogeneous Network (HetNet) is among the most important challenges for the upcoming cellular wireless networks. In this paper, we propose a method to optimize the Radio Access Technologies (RATs) selection and resource allocation in multitechnology wireless networks during a time period. We optimized on a realistic topology of Base Stations (BS) with overlaps of the cellular coverage and dynamic users traffic (arrival and departure). The optimization takes into account the requested services, different users' contract, and user satisfaction. Furthermore, in the proposed approach, we add constraints to ban the session drops and handovers for static users. For each instance, we formulate the problem as a linear optimization problem and we optimize it successively. The aim of the optimization is to jointly maximize the overall user satisfaction and the number of users connected. Compared with a legacy approach, numerical results show that our solution outperforms in terms of user satisfaction.

In this work, we use percolation theory to study the feasibility of large-scale connectivity of relay-augmented device-to-device (D2D) networks in an urban scenario featuring a haphazard system of streets and canyon shadowing allowing only for line-of-sight (LOS) communications in a finite range. We use a homogeneous Poisson-Voronoi tessellation (PVT) model of streets with homogeneous Poisson users (devices) on its edges and independent Bernoulli relays on the vertices. Using this model, we demonstrate the existence of a minimal threshold for relays below which large-scale connectivity of the network is not possible, regardless of all other network parameters. Through simulations, we estimate this threshold to 71.3%. Moreover, if the mean street length is not larger than some threshold (predicted to 74.3% of the communication range; which might be the case in a typical urban scenario) then any (whatever small) density of users can be compensated by equipping more crossroads with relays. Above this latter threshold, good connectivity requires some minimal density of users, compensated by the relays in a way we make explicit. The existence of the above regimes brings interesting qualitative arguments to the discussion on the possible D2D deployment scenarios.

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.

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 study fundamental characteristics for the connectivity of multi-hop D2D networks. Devices are randomly distributed on street systems and are able to communicate with each other whenever their separation is smaller than some connectivity threshold. We model the street systems as Poisson-Voronoi or Poisson-Delaunay tessellations with varying street lengths. We interpret the existence of adequate D2D connectivity as percolation of the underlying random graph. We derive and compare approximations for the critical device-intensity for percolation, the percolation probability and the graph distance. Our results show that for urban areas, the Poisson Boolean Model gives a very good approximation, while for rural areas, the percolation probability stays far from 1 even far above the percolation threshold.

Video traffic constitutes the majority of traffic in bytes that mobile and fixed line operators deliver to their customer. This type of traffic is both resource consuming and QoE sensitive. Either because of content quality or QoE, a large fraction of users often abandon viewing prematurely. These abandonment phenomena lead to a huge waste of network resources and device batteries. Several strategies have been devised to account for all those dimensions. Dominant approaches are fast-caching where the server pushes traffic as fast as possible to the client in order to limit starvation, and ON-OFF strategy where the client forces the server to pause the transfer regularly in order to mitigate the wasted bytes and energy due to users abandonment. In this work, we focus on fast-caching and on ON-OFF strategies. We develop an analytical model which takes into account users dynamics and allows us to quantify the loss in bytes due to users' viewing abandonment. Furthermore, we formulate a multi-objective optimization problem to find ON and OFF period durations that feature a good trade-off between loss due to abandonment and starvation probability.

In this paper, we tackle the problem of access point selection and resource allocation in heterogeneous networks including 3G/4G and Wi-Fi. We study the user satisfaction optimization in terms of throughput and blocking rate. More precisely, the aim is to associate users with the optimal Radio Access Technology (RAT) and to allocate them the optimal number of Resource Units (RUs) based on their requested services and contracts. Starting with a non-linear problem, we re-formulate it into a Mixed Integer Linear Programming (MILP) and assess it on a realistic network configuration. We also compare it with a legacy strategy of selection and allocation. We show that our approach improves the user satisfaction while maintaining a reasonable blocking rate.

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.

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.

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.

The endeavor of this work is to study the impact of content popularity in a large-scale Peer-to-Peer network, namely KAD. Based on an extensive measurement campaign, we pinpoint several deficiencies of KAD in handling popular content and provide a series of improvements to address such shortcomings. Our work reveals that keywords, which are associated with content, may become popular for two distinct reasons. First, we show that some keywords are intrinsically popular because they are common to many disparate contents: in such case we ameliorate KAD by introducing a simple mechanism that identifies stopwords. Then, we focus on keyword popularity that directly relates to popular content. We design and evaluate an adaptive load balancing mechanism that is backward compatible with the original implementation of KAD. Our scheme features the following properties: 1) it drives the process that selects the location of peers responsible to store references to objects, based on object popularity; 2) it solves problems related to saturated peers that would otherwise inflict a significant drop in the diversity of references to objects, and 3) if coupled with a load-aware content search procedure, it allows for a more fair and efficient usage of peer resources.

Enterprise networks have a complexity that sometimes rival the one of the larger Internet. Still, enterprise traffic has received little attention so far from the research community. Most studies rely on port numbers to identify applications.

Accurate identification of network traffic according to application type is a key issue for most companies, including ISPs. For example, some companies might want to ban p2p traffic from their network while some ISPs might want to offer additional services based on the application. To classify applications on the fly, most companies rely on deep packet inspection (DPI) solutions. While DPI tools can be accurate, they require constant updates of their signatures database. Recently, several statistical traffic classification methods have been proposed. In this paper, we investigate the use of these methods for an ADSL provider managing many Points of Presence (PoPs). We demonstrate that statistical methods can offer performance similar to the ones of DPI tools when the classifier is trained for a specific site. It can also complement existing DPI techniques to mine traffic that the DPI solution failed to identify. However, we also demonstrate that, even if a statistical classifier is very accurate on one site, the resulting model cannot be applied directly to other locations. We show that this problem stems from the statistical classifier learning site specific information.

Most tools designed to estimate the capacity of an Internet path require access on both end hosts of the path, which makes them difficult to deploy and use. In this paper we present a single-sided technique for measuring the capacity without the active cooperation of the destination host, focusing particularly on ADSL links. Compared to current methods used on broadband hosts, our approach generates two orders of magnitude less traffic and is much less intrusive. Our tool, DSLprobe, exploits the typical characteristics of ADSL, namely its bandwidth asymmetry and the relatively low absolute bandwidth, in order to measure both downlink and uplink capacities and to mitigate the impact of cross-traffic. To further improve the accuracy, we study different ways to detect and filter cross-traffic packets and we show how to recognize and overcome limited uplink capacities. We validate our tool both on controlled hosts and on a wide variety of Internet hosts. Finally, we present a case study of two large ADSL providers.

Distributed storage systems provide data availability by means of redundancy. To assure a given level of availability in case of node failures, new redundant fragments need to be introduced.

Distributed hash tables (DHTs) have been actively studied in literature and many different proposals have been made on how to organize peers in a DHT. However, very few DHT shave been implemented in real systems and deployed on alarge scale. One exception is KAD , a DHT based on Kademlia, which is part of eDonkey2000, a peer-to-peer file sharing system with several million simultaneous users. We have been crawling KAD continuously for about six months and obtained information about the total number of peers online and their geographical distribution.

While the applications using the Internet have changed over time, TCP is still the dominating transport protocol that carries over 90% of the total traffic. Throughput is the key performance metric for long TCP connections. The achieved throughput results from the aggregate effects of the network path, the parameters of the TCP end points, and the application on top of TCP. Finding out which of these factors is limiting the throughput of a TCP connection -- referred to as TCP root cause analysis -- is important for end users that want to understand the origins of their problems, ISPs that need to troubleshoot their network, and application designers that need to know how to interpret the performance of the application. In this paper, we revisit TCP root cause analysis by first demonstrating the weaknesses of a previously proposed flight-based approach. We next discuss in detail the different possible limitations and highlight the need to account for the application behavior during the analysis process. The main contribution of this paper is a new approach based on the analysis of time series extracted from packet traces. These time series allow for a quantitative assessment of the different causes with respect to the resulting throughput. We demonstrate the interest of our approach on a large BitTorrent dataset.

International audience

International audience

International audience