Showing papers by "INESC-ID published in 2017"

PHYLOViZ 2.0: providing scalable data integration and visualization for multiple phylogenetic inference methods

Abstract: Summary: High Throughput Sequencing provides a cost effective means of generating high resolution data for hundreds or even thousands of strains, and is rapidly superseding methodologies based on a few genomic loci. The wealth of genomic data deposited on public databases such as Sequence Read Archive/European Nucleotide Archive provides a powerful resource for evolutionary analysis and epidemiological surveillance. However, many of the analysis tools currently available do not scale well to these large datasets, nor provide the means to fully integrate ancillary data. Here we present PHYLOViZ 2.0, an extension of PHYLOViZ tool, a platform independent Java tool that allows phylogenetic inference and data visualization for large datasets of sequence based typing methods, including Single Nucleotide Polymorphism (SNP) and whole genome/core genome Multilocus Sequence Typing (wg/cgMLST) analysis. PHYLOViZ 2.0 incorporates new data analysis algorithms and new visualization modules, as well as the capability of saving projects for subsequent work or for dissemination of results. Availability and Implementation: http://www.phyloviz.net/ (licensed under GPLv3). Contact: cvaz@inesc-id.pt Supplementary information: Supplementary data are available at Bioinformatics online.

GrapeTree: Visualization of core genomic relationships among 100,000 bacterial pathogens

Abstract: Current methods struggle to reconstruct and visualise the genomic relationships of ≥100,000 bacterial genomes. GrapeTree facilitates the analyses of allelic profiles from 10,000’s of core genomes within a web browser window. GrapeTree implements a novel minimum spanning tree algorithm to reconstruct genetic relationships despite missing data together with a static “GrapeTree Layout” algorithm to render interactive visualisations of large trees. GrapeTree is a stand-along package for investigating Newick trees plus associated metadata and is also integrated into EnteroBase to facilitate cutting edge navigation of genomic relationships among >160,000 genomes from bacterial pathogens. The GrapeTree package was released under the GPL v3.0 Licence.

VRRRRoom: Virtual Reality for Radiologists in the Reading Room

Abstract: Reading room conditions such as illumination, ambient light, human factors and display luminance, play an important role on how radiologists analyze and interpret images. Indeed, serious diagnostic errors can appear when observing images through everyday monitors. Typically, these occur whenever professionals are ill-positioned with respect to the display or visualize images under improper light and luminance conditions. In this work, we show that virtual reality can assist radiodiagnostics by considerably diminishing or cancel out the effects of unsuitable ambient conditions. Our approach combines immersive head-mounted displays with interactive surfaces to support professional radiologists in analyzing medical images and formulating diagnostics. We evaluated our prototype with two senior medical doctors and four seasoned radiology fellows. Results indicate that our approach constitutes a viable, flexible, portable and cost-efficient option to traditional radiology reading rooms.

Pushing the Limits of Translation Quality Estimation

Abstract: Translation quality estimation is a task of growing importance in NLP, due to its potential to reduce post-editing human effort in disruptive ways. However, this potential is currently limited by the relatively low accuracy of existing systems. In this paper, we achieve remarkable improvements by exploiting synergies between the related tasks of word-level quality estimation and automatic post-editing. First, we stack a new, carefully engineered, neural model into a rich feature-based word-level quality estimation system. Then, we use the output of an automatic post-editing system as an extra feature, obtaining striking results on WMT16: a word-level F 1 MULT score of 57.47% (an absolute gain of +7.95% over the current state of the art), and a Pearson correlation score of 65.56% for sentence-level HTER prediction (an absolute gain of +13.36%).

Representation and analysis of enterprise models with semantic techniques: an application to ArchiMate, e3value and business model canvas

Abstract: Enterprise models assist the governance and transformation of organizations through the specification, communication and analysis of strategy, goals, processes, information, along with the underlying application and technological infrastructure. Such models cross-cut different concerns and are often conceptualized using domain-specific modelling languages. This paper explores the application of graph-based semantic techniques to specify, integrate and analyse multiple, heterogeneous enterprise models. In particular, the proposal described in this paper (1) specifies enterprise models as ontological schemas, (2) uses transformation mapping functions to integrate the ontological schemas and (3) analyses the integrated schemas with graph querying and logical inference. The proposal is evaluated through a scenario that integrates three distinct enterprise modelling languages: the business model canvas, e3value, and the business layer of the ArchiMate language. The results show, on the one hand, that the graph-based approach is able to handle the specification, integration and analysis of enterprise models represented with different modelling languages and, on the other, that the integration challenge resides in defining appropriate mapping functions between the schemas.

Predicting progression of mild cognitive impairment to dementia using neuropsychological data: a supervised learning approach using time windows

Abstract: Predicting progression from a stage of Mild Cognitive Impairment to dementia is a major pursuit in current research. It is broadly accepted that cognition declines with a continuum between MCI and dementia. As such, cohorts of MCI patients are usually heterogeneous, containing patients at different stages of the neurodegenerative process. This hampers the prognostic task. Nevertheless, when learning prognostic models, most studies use the entire cohort of MCI patients regardless of their disease stages. In this paper, we propose a Time Windows approach to predict conversion to dementia, learning with patients stratified using time windows, thus fine-tuning the prognosis regarding the time to conversion. In the proposed Time Windows approach, we grouped patients based on the clinical information of whether they converted (converter MCI) or remained MCI (stable MCI) within a specific time window. We tested time windows of 2, 3, 4 and 5 years. We developed a prognostic model for each time window using clinical and neuropsychological data and compared this approach with the commonly used in the literature, where all patients are used to learn the models, named as First Last approach. This enables to move from the traditional question “Will a MCI patient convert to dementia somewhere in the future” to the question “Will a MCI patient convert to dementia in a specific time window”. The proposed Time Windows approach outperformed the First Last approach. The results showed that we can predict conversion to dementia as early as 5 years before the event with an AUC of 0.88 in the cross-validation set and 0.76 in an independent validation set. Prognostic models using time windows have higher performance when predicting progression from MCI to dementia, when compared to the prognostic approach commonly used in the literature. Furthermore, the proposed Time Windows approach is more relevant from a clinical point of view, predicting conversion within a temporal interval rather than sometime in the future and allowing clinicians to timely adjust treatments and clinical appointments.

MISNIS: An intelligent platform for twitter topic mining

Abstract: Twitter has become a major tool for spreading news, for dissemination of positions and ideas, and for the commenting and analysis of current world events. However, with more than 500 million tweets flowing per day, it is necessary to find efficient ways of collecting, storing, managing, mining and visualizing all this information. This is especially relevant if one considers that Twitter has no ways of indexing tweet contents, and that the only available categorization “mechanism” is the #hashtag, which is totally dependent of a user's will to use it. This paper presents an intelligent platform and framework, named MISNIS - Intelligent Mining of Public Social Networks’ Influence in Society - that facilitates these issues and allows a non-technical user to easily mine a given topic from a very large tweet's corpus and obtain relevant contents and indicators such as user influence or sentiment analysis. When compared to other existent similar platforms, MISNIS is an expert system that includes specifically developed intelligent techniques that: (1) Circumvent the Twitter API restrictions that limit access to 1% of all flowing tweets. The platform has been able to collect more than 80% of all flowing portuguese language tweets in Portugal when online; (2) Intelligently retrieve most tweets related to a given topic even when the tweets do not contain the topic #hashtag or user indicated keywords. A 40% increase in the number of retrieved relevant tweets has been reported in real world case studies. The platform is currently focused on Portuguese language tweets posted in Portugal. However, most developed technologies are language independent (e.g. intelligent retrieval, sentiment analysis, etc.), and technically MISNIS can be easily expanded to cover other languages and locations.

The PathoYeastract database: an information system for the analysis of gene and genomic transcription regulation in pathogenic yeasts.

Abstract: We present the PATHOgenic YEAst Search for Transcriptional Regulators And Consensus Tracking (PathoYeastract - http://pathoyeastract.org) database, a tool for the analysis and prediction of transcription regulatory associations at the gene and genomic levels in the pathogenic yeasts Candida albicans and C. glabrata Upon data retrieval from hundreds of publications, followed by curation, the database currently includes 28 000 unique documented regulatory associations between transcription factors (TF) and target genes and 107 DNA binding sites, considering 134 TFs in both species. Following the structure used for the YEASTRACT database, PathoYeastract makes available bioinformatics tools that enable the user to exploit the existing information to predict the TFs involved in the regulation of a gene or genome-wide transcriptional response, while ranking those TFs in order of their relative importance. Each search can be filtered based on the selection of specific environmental conditions, experimental evidence or positive/negative regulatory effect. Promoter analysis tools and interactive visualization tools for the representation of TF regulatory networks are also provided. The PathoYeastract database further provides simple tools for the prediction of gene and genomic regulation based on orthologous regulatory associations described for other yeast species, a comparative genomics setup for the study of cross-species evolution of regulatory networks.

Geometric Detection Algorithms for Cavities on Protein Surfaces in Molecular Graphics: A Survey

Abstract: Detecting and analyzing protein cavities provides significant information about active sites for biological processes (eg, protein-protein or protein-ligand binding) in molecular graphics and modeling Using the three-dimensional structure of a given protein (ie, atom types and their locations in 3D) as retrieved from a PDB (Protein Data Bank) file, it is now computationally viable to determine a description of these cavities Such cavities correspond to pockets, clefts, invaginations, voids, tunnels, channels, and grooves on the surface of a given protein In this work, we survey the literature on protein cavity computation and classify algorithmic approaches into three categories: evolution-based, energy-based, and geometry-based Our survey focuses on geometric algorithms, whose taxonomy is extended to include not only sphere-, grid-, and tessellation-based methods, but also surface-based, hybrid geometric, consensus, and time-varying methods Finally, we detail those techniques that have been customized for GPU (Graphics Processing Unit) computing

Novel eddy current probes for pipes: Application in austenitic round-in-square profiles of ITER

Abstract: Novel eddy current probes were developed to detect sub-millimetre defects with any orientation on the inner surface of pipes. Five different probes were designed, produced and experimentally validated. These probes include arrays of planar trapezoidal coils in a flexible substrate used alone or together with different winded drive coils. Numerical simulations with Finite Element Method were used to predict the probe response to defects with any orientation. Experimental results in austenitic steel jackets used in ITER revealed that the new probes have an improved reliability compared to conventional toroidal bobbin probes, allowing a higher sensitivity to circumferential defects.

Bioinspired Ciliary Force Sensor for Robotic Platforms

Abstract: The detection of small forces is of great interest in any robotic application that involves interaction with the environment (e.g., objects manipulation, physical human-robot interaction, minimally invasive surgery), since it allows the robot to detect the contacts early on and to act accordingly. In this letter, we present a sensor design inspired by the ciliary structure frequently found in nature, consisting of an array of permanently magnetized cylinders (cilia) patterned over a giant magnetoresistance sensor (GMR). When these cylinders are deformed in shape due to applied forces, the stray magnetic field variation will change the GMR sensor resistivity, thus enabling the electrical measurement of the applied force. In this letter, we present two 3 mm × 3 mm prototypes composed of an array of five cilia with 1 mm of height and 120 and 200 μm of diameter for each prototype. A minimum force of 333 μN was measured. A simulation model for determining the magnetized cylinders average stray magnetic field is also presented.

Function-based modulation control for modular multilevel converters under varying loading and parameters conditions

Abstract: This study presents a new function-based modulation control technique for modular multilevel converters (MMCs). The main contribution of this study is the formulation of two new modulation functions for the required switching signals of the MMC's upper and lower sub-modules, respectively. The output and circulating current equations of the converter are employed to attain the arm's currents which are utilised for the proposed modulation functions, which have two important features: (i) it is much less complex compared to the existing control methods of MMC; and (ii) the proposed controller can be regulated properly to deal with parameter variations in a bid to ensure stable and accurate performance. In this controller, the MMC output current magnitude and phase angle required for special active and reactive power sharing can be easily applied to the modulation functions. Also, the equivalent capacitors of upper and lower sub-modules are discussed based on the proposed modulation functions. Finally, simulations are performed in Matlab/Simulink environment to evaluate the performance of the proposed control technique in both the dynamic conditions of load as well as varying parameters.

A new look at the fractionalization of the logistic equation

Abstract: The fractional version of the logistic equation will be studied in this paper. Motivated by unsuccessful previous papers, we showed how to obtain the correct solution. The algorithm is very simple. Its numerical implementation will be studied and exemplified using a Pade approximation.

A CMOS Front-End With Integrated Magnetoresistive Sensors for Biomolecular Recognition Detection Applications

Abstract: The development of giant magnetoresistive (GMR) sensors has demonstrated significant advantages in nanomedicine, particularly for ultrasensitive point-of-care diagnostics. To this end, the detection system is required to be compact, portable, and low power consuming at the same time that a maximum signal to noise ratio is maintained. This paper reports a CMOS front-end with integrated magnetoresistive sensors for biomolecular recognition detection applications. Based on the characterization of the GMR sensor's signal and noise, CMOS building blocks (i.e., current source, multiplexers, and preamplifier) were designed targeting a negligible noise when compared with the GMR sensor's noise and a low power consumption. The CMOS front-end was fabricated using AMS $\mathrm{{\text{0.35}}\;\mu{\text{m}}}$ technology and the magnetoresistive sensors were post-fabricated on top of the CMOS chip with high yield ( ${\text{97.9}}\%$ ). Due to its low circuit noise (16 $\mathrm{{\text{nV}}/\sqrt{{\text{Hz}}}}$ ) and overall equivalent magnetic noise ( $\mathrm{{\text{11.5}}\;{\text{nT}}/\sqrt{{\text{Hz}}}}$ ), the full system was able to detect 250 nm magnetic nanoparticles with a circuit imposed signal-to-noise ratio degradation of only $-$ 1.4 dB. Furthermore, the low power consumption (6.5 mW) and small dimensions ( $\mathrm{{\text{7.59: mm}}^{2}}$ ) of the presented solution guarantees the portability of the detection system allowing its usage at the point-of-care.

A Social Robot as a Card Game Player.

Abstract: This paper describes a social robotic game player that is able to successfully play a team card game called Sueca. The question we will address in this paper is: how can we build a social robot player that is able to balance its ability to play the card game with natural and social behaviours towards its partner and its opponents. The first challenge we faced concerned the development of a competent artificial player for a hidden information game, whose time constraint is the average human decision time. To accomplish this requirement, the Perfect Information Monte-Carlo (PIMC) algorithm was used. Further, we have performed an analysis of this algorithm’s possible parametrizations for games trees that cannot be fully explored in a reasonable amount of time with a MinMax search. Additionally, given the nature of the Sueca game, such robotic player must master the social interactions both as a partner and as an opponent. To do that, an emotional agent framework (FAtiMA) was used to build the emotional and social behaviours of the robot. At each moment, the robot not only plays competitively but also appraises the situation and responds emotionally in a natural manner. To test the approach, we conducted a user study and compared the levels of trust participants attributed to the robots and to human partners. Results have shown that the robot team exhibited a winning rate of 60%. Concerning the social aspects, the results also showed that human players increased their trust in the robot as their game partners (similar to the way to the trust levels change towards human partners). As interactive entertainment expands, computer games are progressively moving from the virtual world back to the physical world. Augmented reality games, haptic interfaces in gaming, touch tables, etc, are some of the types of interactivity placing human players in physically situated entertainment experiences. In parallel with this move into the physical world, artificial partners and opponents can also be created to exist in such physical world. To do that, the area of entertainment robots offers challenging opportunities as it explores the role of a robot as a game player. In general, social robots can contribute with new and broad ways of creating socially engaging interactions with humans in entertainment contexts. The challenges of these human-robot interactions may vary from game to game. Some games, when played in the physical world not only hold complex social Copyright c © 2017, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. behaviours, but they also hinder performance aspects related with the competitive nature of the game itself. Furthermore, in many types of games, current advances in Artificial Intelligence (AI) over the past few years has shown that strong and powerful algorithms combined with significant amounts of data, are able to defeat human world champions of these games (see for example, the game of Go). These results raise our expectations and people are starting to consider such artificial agents as fierce competitors. Yet, when we consider multi-player games, where the social environment becomes more relevant, and when the games are played in the physical world, how will people perceive a social robotic player compared to human standards? Will people be willing to trust a social robot to be his partner in a team game? To address these questions we created a social autonomous robotic partner for the Sueca card game, with a twofold goal of both playing competitively and interacting socially with the other players. The development of such robotic game player introduces some challenging aspects, in particular finding the best balance between social responses and computations related with the game, in order for the socially intelligent agent to produce natural and human-like behaviours. Another important challenge of creating an intelligent agent in this social context is the time constraint on the computation of a hidden information card game. State-of-the-art approaches, for instance PIMC, promise good results on the Sueca domain according to the game properties. However, the full computation of multiple perfect information games is not time-efficient and will hinder natural interaction in a game with human players. Therefore, this paper also explores how the algorithm’s parametrizations affect the game results in order to choose the best performance-time configuration. Finally, by using an expressive robot that is able to express emotions, provide spoken feedback, and respond socially, the game experience can be created balancing these social and game competencies. In this case, we built the social competencies by using an emotional agent framework (FatiMA) which allows for emotional appraisal to occur and fire social and emotional behaviours. At each moment, the robot not only plays competitively, but also appraises the situation and responds emotionally to the game situations. Proceedings, The Thirteenth AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment (AIIDE-17)

Synchronous active proportional resonant-based control technique for high penetration of distributed generation units into power grids

Abstract: This paper deals with a synchronous active proportional resonant-based (SAPR) control technique for interfaced converters, enhancing the stable operation of the power grid under high penetration of distributed generation sources. By considering the grid specifications and load currents, both d and q axis of converter currents are obtained in terms of active and reactive power and also angular speed using small-signal linearization method. Then, swing equation is analyzed in detail to achieve the reference current components in the current control loop of the interfaced converter. By using the obtained swing equation and a non-ideal proportional resonant (PR) controller, a new control technique is proposed, which introduces the behavior of synchronous power generators based on power electronic converters in distributed generation (DG) technology. The effectiveness of the proposed control technique is verified through stringent simulation studies in MATLAB/SIMULINK.

Parallel dot-products for deep learning on FPGA

Abstract: Deep neural networks have recently shown great results in a vast set of image applications. The associated deep learning models are computationally very demanding and, therefore, several hardware solutions have been proposed to accelerate their computation. FPGAs have recently shown very good performances for these kind of applications and so it is considered a promising platform to accelerate the execution of deep learning algorithms. A common operation in these algorithms is multiply-accumulate (MACC) that is used to calculate dot-products. Since many dot products can be calculated in parallel, as long as memory bandwidth is available, it is very important to implement this operation very efficiently to increase the density of MACC units in an FPGA. In this paper, we propose an implementation of parallel MACC units in FPGA for dot-product operations with very high performance/area ratios using a mix of DSP blocks and LUTs. We consider fixed-point representations with 8 bits of size, but the method can be applied to other bit widths. The method allows us to achieve TOPs performances, even for low cost FPGAs.

Double synchronous controller for integration of large-scale renewable energy sources into a low-inertia power grid

Abstract: Renewable energy sources are normally connected to the power grid via power electronic converters. High penetration of these energy sources into the power grid leads to high instability in voltage and frequency. This issue is caused by neglecting the inherent characteristics of synchronous generators i.e., inertia, damping and proper active and reactive power sharing in the structure of the used control technique in the control loop of the interfaced converter between power grid and renewable energy sources. This paper presents a power-based control technique based on a double synchronous controller (DSC) for interfaced converter between the renewable energy sources and the power grid, including an active-reactive power based dynamic equation. Through the proposed DSC, a decoupled control method is performed in which both active and reactive power can be injected from renewable energy sources into the power grid by the interfaced power converter with the inherent features of synchronous power generators. By using the proposed control technique, a stable operation of the power grid can be guaranteed during the integration of large-scale renewable energy sources. Stringent simulation results performed in MATLAB/SIMULINK environment verify the proficiency of the proposed control technique.

Three-phase qZ-source inverter with fault tolerant capability

Abstract: The requirement of three-phase inverters with high reliability capability is considered important in several applications. Most of the adopted fault-tolerant solutions use classical three-phase voltage source inverters (VSI) with redundant controlled power devices or modified control strategies for acceptably degraded performance in the event of a failure yielding the loss of devices. This work proposes the use of a fault-tolerant converter based on the quasi-Z-source inverter (qZSI). This inverter presents important features such as boost capability and not requires dead time circuits. These features associated to the proposed topology results in a converter with high reliability and fault-tolerant capability. The proposed topology is designed to allow the reconfiguration of the converter after a power switch failure, using only two legs in the faulty mode. Additionally, this topology avoids extra inductors, capacitors and active power switches for the fault-tolerant condition. A current controller associated to a vectorial voltage modulator is used to control the proposed topology, which results in three-phase balanced AC currents with reduced total harmonic distortion even in faulty condition. The performance of this solution is confirmed through experimental results obtained from a laboratory prototype.

Hip-directed walking-in-place using a single depth camera

Abstract: Walking-in-place (WIP) is a locomotion technique that allows users to travel in virtual environments (VEs) without significantly changing their physical position on the floor. Hip-directed steering(HDS) is a novel physical technique for controlling direction changes in virtual travel using hip movements. We present an WIP-based navigation approach for controlling locomotion in VEs that combines the speed and direction in a scenario similar to a domestic setup in which people interact with a flat screen. Their physical motion data are captured by one depth camera properly aligned with the screen and oriented toward the user. We approach the characteristically noisy data generated by depth cameras via a user study to determine both the range of values and their robustness from the motion data associated with the joints relevant to WIP speed(knee, ankle and foot) and HDS (spine, hip and shoulder) to derive a reliable technique. Our WIP speed method is supported in a simple vocabulary of five different footstep types. Experimental results show that both the knee and hip provide the most robust data. We evaluated our techniques via usability tests exercising common locomotion tasks. The results show that users liked both the speed control and comfort afforded by our speed method. Regarding HDS, users reported that the angular-based method allowed them to travel faster and was both more controllable and easier to learn than the time-based method. Our work shows that a single depth camera can be used to combine locomotion and direction control in a simple and affordable setup.

Automatic Detection of Parkinson’s Disease: An Experimental Analysis of Common Speech Production Tasks Used for Diagnosis

Abstract: Parkinson’s disease (PD) is the second most common neurodegenerative disorder of mid-to-late life after Alzheimer’s disease. During the progression of the disease, most individuals with PD report impairments in speech due to deficits in phonation, articulation, prosody, and fluency. In the literature, several studies perform the automatic classification of speech of people with PD considering various types of acoustic information extracted from different speech tasks. Nevertheless, it is unclear which tasks are more important for an automatic classification of the disease. In this work, we compare the discriminant capabilities of eight verbal tasks designed to capture the major symptoms affecting speech. To this end, we introduce a new database of Portuguese speakers consisting of 65 healthy control and 75 PD subjects. For each task, an automatic classifier is built using feature sets and modeling approaches in compliance with the current state of the art. Experimental results permit to identify reading aloud prosodic sentences and story-telling tasks as the most useful for the automatic detection of PD.

Depression Detection Using Automatic Transcriptions of De-Identified Speech.

Abstract: Depression is a mood disorder that is usually addressed by outpatient treatments in order to favour patient’s inclusion in society. This leads to a need for novel automatic tools exploiting speech processing approaches that can help to monitor the emotional state of patients via telephone or the Internet. However, the transmission, processing and subsequent storage of such sensitive data raises several privacy concerns. Speech deidentification can be used to protect the patients’ identity. Nevertheless, these techniques modify the speech signal, eventually affecting the performance of depression detection approaches based on either speech characteristics or automatic transcriptions. This paper presents a study on the influence of speech de-identification when using transcription-based approaches for depression detection. To this effect, a system based on the global vectors method for natural language processing is proposed. In contrast to previous works, two main sources of nuisance have been considered: the de-identification process itself and the transcription errors introduced by the automatic recognition of the patients’ speech. Experimental validation on the DAIC-WOZ corpus reveals very promising results, obtaining only a slight performance degradation with respect to the use of manual transcriptions.

Modelling and implementation of virtual radio resources management for 5G Cloud RAN

Abstract: The virtualisation of Radio Access Networks (RANs) is one of the goals in designing 5G mobile networks. This paper aims at presenting a proof of concept for the virtualisation of radio resources using Open Air Interface (OAI), a software-based Long-Term Evolution (LTE) eNodeB physical emulator. OAI was extended to support multi-tenancy, representing diverse Virtual mobile Network Operators (VNOs) with different Service Level Agreements (SLAs). A comprehensive analytical model for managing the virtual radio resources has been proposed, with two key parts: estimation of available radio resources and their allocation to different VNOs. The estimation is performed by the model, and the allocation is managed by OAI scheduling. Various scenarios and use cases are studied in this virtual RAN environment, network performance being evaluated for different situations, by varying guaranteed levels, serving weights, and used services. Results show that the proposed approach offers almost the same capacity to guaranteed VNOs regardless of other existing VNOs, experiencing at worst a degradation of 32% of its initial allocated data rate, without violation of the guaranteed data rate. The data rate allocated to best effort VNOs may decrease up to 7% of its initial value, which is acceptable, to guarantee other more demanding SLAs.