IMDEA Networks Institute Digital Repository
Not a member yet
1915 research outputs found
Sort by
On the level of detail of synthetic highway traffic necessary to vehicular networking studies
Get PDFThe proper modeling of road traffic is paramount to the dependability of studies on vehicular networking solutions intended for highway environments. Yet, it is not clear which is the actual level of detail in the mobility representation that is sufficient and necessary to such studies. This uncertainty results into a variety of approaches being adopted in the literature, and ultimately undermines the reliability and reproducibility of research outcomes. We explore the space of possible mobility models and performance metrics, and pinpoint the level of detail needed for different types of vehicular networking studies.TRUEpu
Assessing peering evolution in Africa (Remote talk)
No full textThere is a lack of knowledge on how ASes have been behaving at African IXPs from 2005 up to now, the period those IXPs have been existing for, their richness in term of reachable ASes or prefixes, etc. To correct this situation, we collected and analyzed PCH data from 2005-2015 which involve 12 African IXPs in Africa, including JINX and KIXP (known to host RouteViews Collector). The analysis of such data reveals some interesting metrics per IXP, namely the time difference between prefix allocation (AFRINIC DB) and appearance on the Internet (RIS data/RIPE stats), the time difference between prefix allocation (AFRINIC DB) and appearance at an IXP (PCH data), the number of prefixes visible at each IXP, Origin ASN and Prefix growth stats per year for each IXP, and etc. We presented the results during this talk.TRUEpu
The Importance of Being Earnest in Crowdsourcing Systems
Get PDFThis paper presents the first systematic investigation of the potential performance gains for crowdsourcing systems deriving from available information at the requester about individual worker earnestness (reputation). In particular, we first formalize the optimal task assignment problem when workers' reputation estimates are available, as the maximization of a monotone (submodular) function subject to Matroid constraints. Then, being the optimal problem NP-hard, we propose a simple but efficient greedy heuristic task allocation algorithm. We also propose a simple "maximum a posteriori" decision rule. Finally, we test and compare different solutions, showing that system performance can greatly benefit from information about workers' reputation. Our main findings are that: i) even largely inaccurate estimates of workers' reputation can be effectively exploited in the task assignment to greatly improve system performance; ii) the performance of the maximum a-posteriori decision rule quickly degrades as worker reputation estimates become inaccurate; iii) when workers' reputation estimates are significantly inaccurate, the best performance may be obtained by combining our proposed task assignment algorithm with the decision rule introduced in [1,2].TRUEpu
BASEL (Buffering Architecture SpEcification Language)
Get PDFBuffering architectures and policies for their efficient management constitute one of the core ingredients of a network architecture.
In this work we introduce a new specification language, BASEL, that allows to express virtual buffering architectures
and management policies representing a variety of economic models. BASEL does not require the user to implement policies in a high-level language; rather, the entire buffering architecture and its policy are reduced to several comparators and simple functions. We show examples of buffering architectures in BASEL and demonstrate empirically the impact of various settings
on performance
Xhaul: toward an integrated fronthaul/backhaul architecture in 5G networks
Get PDFThe Xhaul architecture presented in this article is aimed at developing a 5G integrated backhaul and fronthaul transport network enabling flexible and software-defined reconfiguration of all networking elements in a multi-tenant and service-oriented unified management environment. The Xhaul transport network vision consists of high-capacity switches and heterogeneous transmission links (e.g., fiber or wireless optics, high-capacity copper, mmWave) interconnecting remote radio heads, 5G points of attachment (5GPoAs, e.g., macro- and small cells), centralized- processing units (mini data centers), and points of presence of the core networks of one or multiple service provider(s). This transport network shall flexibly interconnect distributed 5G radio access and core network functions, hosted on network centralized nodes, through the implementation of a control infrastructure using a unified, abstract network model for control plane integration (Xhaul Control Infrastructure, XCI); and a unified data plane encompassing innovative high-capacity transmission technologies and novel deterministic-latency switch architectures (Xhaul packet Forwarding Element, XFE). Standardization is expected to play a major role in a future 5G integrated front haul/backhaul architecture for multi-vendor interoperability reasons. To this end, we review the major relevant activities in the current standardization landscape and the potential impact on the Xhaul architecture.pu
Optimal Communication Structures for Big Data Aggregation
Get PDFAggregation of computed sets of results fundamentally underlies the distillation of information in many of today’s
big data applications. To this end there are many systems which have been introduced which allow users to obtain aggregate results by aggregating along communication structures such as trees, but they do not focus on optimizing performance by optimizing the underlying structure to perform the aggregation. We consider two cases of the problem – aggregation of (1) single blocks of data, and of (2) streaming input. For each case we determine which metric of “fast” completion is the most relevant and mathematically model resulting systems based on aggregation trees to optimize that metric. Our assumptions and model are laid out in depth. From our model we determine how to create a provably ideal aggregation tree (i.e., with optimal fanin) using only limited information about the aggregation function being applied. Experiments in the Amazon Elastic Compute Cloud (EC2) confirm the validity of our models in practice.TRUEpu
Efficient Networking in Millimeter Wave Bands
No full textState-of-the-art wireless communication already operates close to Shannon capacity and one of the most promising options to further increase data rates is to increase the communication bandwidth. Very high bandwidth channels are only available in the extremely high frequency part of the radio spectrum, the millimeter wave band (mm-wave). Upcoming communication technologies, such as IEEE 802.11ad, are already starting to exploit this part of the radio spectrum to achieve data rates of several GBit/s. However, communication at such high frequencies also suffers from high attenuation and signal absorption, often restricting communication to line-of-sight (LOS) scenarios and requiring the use of highly directional antennas. This in turn requires a radical rethinking of wireless network design. On the one hand side, such channels experience little interference, allowing for a high degree of spatial reuse and potentially simpler MAC and interference management mechanisms. On the other hand, such an environment is extremely dynamic and channels may appear and disappear over very short time intervals, in particular for mobile devices. This talk will highlight some of the challenges of and possible approaches for networking in the mm-wave band.TRUEpu
Prediction-based Optimization in Mobile Networks
No full textA highly interesting trend in mobile network optimization is to exploit knowledge of future network capacity to allow mobile terminals to prefetch data when signal quality is high and to refrain from communication when signal quality is low. While this approach offers remarkable benefits, it relies on the availability of a reliable forecast of system conditions.
This talk moves from the analysis of prediction techniques for mobile networks to proposing different optimization schemes for resource allocation and admission control. In particular, the talk will discuss a model for throughput prediction in mobile networks and its use in a single user resource allocation mechanism under imperfect prediction. Then, the problem will be extended to multi-user, multi-quality scenarios and the talk will address both optimal solutions and heuristic approaches for resource allocation and admission control.TRUEpu
Online Parallel Scheduling of Non-uniform Tasks: Trading Failures for Energy
Get PDFDOI information: 10.1016/j.tcs.2015.01.027Consider a system in which tasks of different execution times arrive continuously and have to be executed by a set of machines that are prone to crashes and restarts. In this paper we model and study the impact of parallelism and failures on the competitiveness of such an online system. In a fault-free environment, a simple Longest-in-System scheduling policy, enhanced by a redundancy-avoidance mechanism, guarantees optimality in a long-term execution. In the presence of failures though, scheduling becomes a much more challenging task. In particular, no parallel deterministic algorithm can be competitive against an off-line optimal solution, even with one single machine and tasks of only two different execution times. We find that when additional energy is provided to the system in the form of processing speedup, the situation changes. Specifically, we identify thresholds on the speedup under which such competitiveness cannot be achieved by any deterministic algorithm, and above which competitive algorithms exist. Finally, we propose algorithms that achieve small bounded competitive ratios when the speedup is over the threshold.pu