Network Functions Virtualization (NFV) virtualizes network functions typically provided by hardware, such as firewalls and routers, using Virtualized Network Functions (VNFs). Service Function Chains (SFCs) are sequences of VNFs that process user data streams, which can span different autonomous and competing infrastructure providers. Blockchain technology has been used several times for resource allocation solutions for VNFs, such as in BRAIN, using reverse auctions. This paper presents BRAIN*, an enhanced version of BRAIN, capable of managing SFC auctions in a single smart contract using consortiums. In addition, it proposes an interface for auction contracts and performs experiments to validate the proposed solution. Experimental results shows that BRAIN* con efficiently handles a high number of simultaneous auctions without incurring additional costs. The results also show that the proposed solution is more effective in scenarios with a high number of providers participating in the auctions, returning lower prices to users and maintain a block rate below 10%.

Dimensioning networks whose structure is composed of a rare and clustered population is not a trivial task. Several areas of computer science, like mobile and wireless systems, in which the research success and the optimization of resources in the network may depend on the size and population distribution, can be related to this problem domain. In these circumstances, it is usual to overlay a grid over the region in which the population is contained, select cells from this grid, analyze the existence of the population elements, and, in some cases, also add neighboring cells containing the variable of interest. This methodology is presented by Adaptive Cluster Sampling - ACS. However, ACS considers the collection of all interest elements within the cell, which is not realistic for all cases, so in this paper, it is called the Optimum Method - OM. To solve this fact, a technique called the 2-Layer and 2-Estimator Method - 2L2EM was proposed, which implements the Multiple Capture and Recapture Method - MCRM in layer 1 to obtain the total population estimates within the cell to be used as input to ACS in layer 2, where the total population in the network is estimated. However, for being able to apply the proposed technique, it was necessary to implement a stopping criterion for the number of recaptures to the MCRM to obtain efficient estimates without exceeding the number of recaptures. The application to real data reveals that 2L2EM provides relevant estimates in relation to OM and significant advantages over MCRM.

Recommender systems play a crucial role in helping users decide what to watch or purchase by suggesting relevant items.These systems can enhance the media experience by considering user preferences, inferring behavior, and personalized recommendations.However, users often do not understand why a particular item was recommended to them.Explainable recommender systems aim to clarify the reasoning behind recommendations, increasing user trust and confidence.Despite advancements, gaps remain in the literature, particularly in evaluating these systems.This thesis will explore and propose new metrics to better assess explanation methods, investigating why and how current explanations fall short in evaluations.Additionally, we aim to examine whether explanations can reveal if recommender systems create filter bubbles and explore ways to mitigate this issue based on user preferences.

As mobile networks become more complex to handle increasing data traffic and a broader range of services, operators must balance the trade-offs between static and dynamic configurations. While traditional static configurations across the entire network are simpler to manage, dynamic adjustments, though more complex to operate, are better suited to adapting to evolving demands. To explore this balance, in this paper, we use real data from a mobile network to evaluate the potential gains in throughput gains, measured by downlink traffic, when dynamically adjusting configurations at both spatial and temporal scales. Our findings show that combining these dynamic adjustments leads to significant performance improvements, with traffic volume gains exceeding 30% when configurations are tailored at the cell level and to the hour scale.

Generative Adversarial Networks (GANs) have emerged as tools for creating synthetic data that mimics real datasets. Time series GANs extend GANs concept by attempting to replicate the temporal dependencies and patterns of time series, while preserving the privacy of the real data. However, measuring the fidelity and utility of a synthetic time series remains a challenge. Thus, this paper discusses the pros and cons of quantitative metrics to assess synthetic time series generated by GANs based on sensitive data of two application domains. We first review how metrics based on probability distributions, such as Kullback-Leibler Divergence, Jensen-Shannon Distance, Wasserstein Distance, and Maximum Mean Discrepancy, have been used to evaluate synthetic data. Meanwhile, to assess the utility of synthetic data the Testing on Synthetic, Training on Real (TSTR) score is used. To assess these metrics, we compare three time series GANs (RGAN, TimeGAN, Doppelganger) to generate synthetic datasets of two network domain applications: i) user content requests for a Brazilian streaming provider; and ii) devices connected to mobile base stations in a large Brazilian city. We compare the fidelity of synthetic datasets and assess their utility in predicting content requests and the number of users connected to a given mobile base station.

With the popularization of video streaming services, it has become increasingly important to discover which videos will be popular to prepare the network infrastructure. Popularity prediction has been studied with several Machine Learning models and with different features captured from videos. This article presents a multimodal model that uses different classifiers fed by different features, building a robust and flexible model that surpasses models used in practice. To build this model, we used data from Globoplay, the largest streaming service in Latin America. On the other hand, predicting content popularity from a catalog of available media can identify videos that demand more resources from the network infrastructure, allowing service providers to adopt preventive measures to maintain transmission quality. Notably, we analyze if visual features extracted from thumbnails add value to this task. We experiment with the proposed approaches on a set of videos from GloboPlay. Our model gives the best accuracy found, in addition to the advantages of robustness and flexibility, adapting better to practical cases of popularity prediction.

Incentive mechanisms are increasingly needed in opportunistic networks that contain nodes with selfish behavior. For this, credit-based mechanisms need a virtual bank (central entity) to promote the incentive. However, the existence of this central entity in an opportunistic network may be challenging. Therefore, we propose a credit incentive mechanism in this paper. The main contribution is that the mechanism does not use a virtual bank to distribute credits (reward for forwarding messages) but rather a decentralized approach. In addition, we propose mathematical modeling to represent the collection and distribution of credits to avoid Edge Insertion attacks in some instances. Finally, the proposed mechanism was evaluated through simulation using real mobility traces and different routing protocols and compared its performance with the RELICS incentive mechanism. The obtained results show that the proposed mechanism is promising in diminishing the occurrence of selfish nodes.

The number of fixed broadband customers has continued to increase over the past decade in Brazil. In spite of its demand, customers face numerous issues with broadband services. Thus, in this paper, we partner with TIM, one of the largest fixed broadband service providers in Brazil, to analyze customers’ Quality of Service (QoS) parameters and predict customers’ download rates. We consider 3.4 million logs collected from 5% of the total customers for 31 days starting from May 22nd, 2021. We build a framework using Error-Correcting Output Codes (ECOC) and H2O’s Automatic Machine Learning (AutoML) that accurately predicts the quality of service, particularly the download rate, achieved by the customers using features related to customer location, internet plan, and equipment. Our experiments demonstrate that our model achieves around 83% accuracy on average on our dataset. Our framework can be used by TIM to improve its fixed broadband services.

With the advancement in technology, many organizations use Trouble Ticket Systems (TTS) to record and manage problems, facilitating the process of assigning it to the right technical team. However, in large organizations, which receive a huge number of complaints, the task of allocating or classifying a problem becomes a challenge. In this paper, we propose a solution for automatically classifying customer complaints related to fixed broadband service for TIM, large Brazilian telecommunications company. We consider the fixed broadband customer complaints reported for a period of 31 days starting from December 28, 2020. We propose a custom textual preprocessing technique and use several machine learning classifiers to accurately classify a complaint in one of the six problem classes. Our results demonstrate that the proposed technique generated an increase in accuracy of 8.4% when compared to techniques commonly used in textual preprocessing. Our results also demonstrate that the Extra Tree Classifier achieves the best performance among all models with an accuracy of around 89%. Our work can assist TIM to improve their complaint resolution process.

Predicting network metrics with high accuracy is a challenging task. Network Latency prediction allows Network Operators (NO), Internet Service Providers (ISP), and Over The Top (OTT) Service Providers to optimize their performance in almost real-time. Moreover, applications can make decisions based on latency predictions, improving the quality of the provided services. In recent years, many works proposed machine learning based techniques to predict network metrics, especially using Recurrent Neural Network (RNN) techniques, such as Long Short-Term Memory (LSTM) Networks. Nevertheless, despite the good results achieved, the computational cost of training and keeping a model updated makes adopting those techniques unfeasible in some scenarios. In this work, we propose the usage of Data Stream Learning techniques to predict RoundTrip Time (RTT) and One-Way Delay (OWD) metrics using realworld data. The experiment results demonstrated a prediction performance similar to LSTM networks using only a fraction of the computational resources used by LSTM.

In this paper, our goal is to model the aggregate mobility of individuals in a city by analyzing cellular network connections, and then leverage the designed mobility model to model and predict the number of COVID-19 infections in future. We analyze cellular network connections from 973 antennas for all users in the city of Rio de Janeiro from April 5, 2020 to July 2, 2020. We design a Markovian model that captures the mobility across municipalities. We then combine the transition probabilities of the Markov chain with the number of COVID-19 cases in a municipality during a particular week in the design of our mobility-aware COVID-19 case prediction models to predict the number of cases for the following week. Our experiments demonstrate that our mobility-aware models significantly out-perform a baseline mobility-agnostic linear regression model in terms of metrics such as Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE).

In this paper, our goal is to analyze and compare cellular network usage data from pre-lockdown, during lock-down, and post-lockdown phases surrounding the COVID-19 pandemic to understand and model human mobility patterns during the pandemic. To this end, we collect and analyze cellular network connections from 1400 antennas for all users in the city of Rio de Janeiro and its suburbs from March 1, 2020 to July 1, 2020. Our analysis reveals that the total number of cellular connections decreases to 78% during the lockdown phase and then increases to 85% of the pre-COVID era as the lockdown eases. We observe that user mobility starts increasing around 3 weeks before the end of lockdown, with the trend continuing into the post-lockdown period. We also design an interactive tool that showcases mobility patterns in different granularities and can help government officials take informed actions to control the spread of the disease.

In this paper, our goal is to leverage cellular network traffic data to model and forecast the number of COVID-19 infections in the future. To this end, we partner with one of the main cellular network providers in Brazil, TIM Brazil, and collect and analyze cellular network connections from 973 antennas for all users in the city of Rio de Janeiro and its suburbs. We develop a Markovian model that captures the mobility of individuals across municipalities of the city. The transition probabilities of the Markov chain are determined by analyzing user-level mobility events between antennas from the cellular network connectivity logs. We combine the aggregate mobility characteristics across municipalities as evidenced from the transition probabilities with the number of reported COVID-19 cases in a municipality during a particular week to design mobility-aware COVID-19 case prediction models that predict the number of cases for the following week. Our experiments demonstrate that our mobility-aware models significantly outperform a baseline mobility-agnostic linear regression model in terms of metrics such as Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE).

In this paper, our goal is to analyze and compare cellular network usage data from Rio de Janeiro from pre-lockdown, during lockdown, and post-lockdown phases surrounding the COVID-19 pandemic to understand and model human mobility patterns during the pandemic, and to evaluate the effect of lockdowns on mobility. Our analysis reveals that human mobility increases significantly even before lockdown restrictions are eased, with the trend continuing in the post-lockdown period. We also observe that the day of week has a significant impact on mobility of individuals, with the overall mobility on Fridays increasing over time possibly due to people self-relaxing restrictions and engaging in social activities on Friday evenings. We also design an interactive tool that showcases mobility patterns in different granularities and can potentially help people and government officials understand the mobility of individuals and the number of COVID-19 cases in a particular neighborhood.

In recent years many changes have taken place, such as increasingly powerful smartphones and cellular network allowing broadband access, and is expected in the coming years an increase in live streaming data traffic for mobile devices. On the other hand, it is common to find in the literature criticism of the traditional client-server model, that the Internet was not designed to support multimedia applications, or even that mobile network isn't appropriated to streaming, despite the fact that video streaming works and is a very popular Internet application. This work proposes, from the log files of a large CDN, discuss the influence of impact factors in the quality of the users' transmissions using mobile devices in popular live video transmissions. Using association rule, a data mining technique, this paper aims to analyze popular live streaming sessions to understand what factors may impact the broadcasts.

Media streaming is one of the key applications on the Internet, with mobile devices becoming each day more powerful and popular. Although mobile users have experienced a large improvement in wireless access networks in the last years, live-streaming videos still face several challenges, especially for large-scale popular events. Therefore, understanding live streaming for mobile users becomes imperative nowadays. This paper targets at the characterization of mobile user behaviors, The main goal of this paper is to understand the behavior of mobile users when watching large popular live events in Brazil, such as 2016 former Brazilian president Impeachment and 2016 Opening Ceremony for Rio Olympic Games. We focus our analysis on comprehending the influence of interruptions on the session duration and new attempts to rejoin the video transmission. The results show that a significant number of mobile users experiences low transmission rates and the type of event imply different user behaviors.

The Internet of Things (IoT) is increasingly a reality today. Nevertheless, some key challenges still need to be given particular attention so that IoT solutions further support the growing demand for connected devices and the services offered. Due to the potential relevance and sensitivity of services, IoT solutions should address the security and privacy concerns surrounding these devices and the data they collect, generate, and process. Recently, the Blockchain technology has gained much attention in IoT solutions. Its primary usage scenarios are in the financial domain, where Blockchain creates a promising applications world and can be leveraged to solve security and privacy issues. However, this emerging technology has a great potential in the most diverse technological areas and can significantly help achieve the Internet of Things view in different aspects, increasing the capacity of decentralization, facilitating interactions, enabling new transaction models, and allowing autonomous coordination of the devices. The paper goal is to provide the concepts about the structure and operation of Blockchain and, mainly, analyze how the use of this technology can be used to provide security and privacy in IoT. Finally, we present the stalker, which is a selfish miner variant that has the objective of preventing a node to publish its blocks on the main chain.

Many BitTorrent (BT) clients are using these BT networks as a video-on-demand service, taking advantage of the popularity and the large collection of media available. However, transforming the swarms into an on-demand media service can cause serious damage to the overall network performance. In this paper, we propose a methodology, using the concepts of Entropy, and present a Spy BitTorrent client that is able to identify peers streaming in a swarm. Large scale monitoring, for real swarms, were performed to detect the presence of streamers.

Many BitTorrent clients are using these networks as a video-on-demand service, taking advantage of the popularity and the enormity of the collection of media available. However, it is a common sense that transforming the swarms into an on-demand media service causes serious damage to the overall network performance, given that streamer clients modify the method of downloading determined in the protocol. In this paper, we perform controlled experiments to analyze the possible impact in the swarms' performance in the presence of streamers. Surprisingly we show that despite the presence of streamers harms the overall system performance, the degradation is not so sharp and, surprisingly, in some cases/scenarios, it has even decreased the download time.

Named Data Networking (NDN) is one of the promising proposals of Future Internet Architectures (FIAs). Similarly to most of the other FIA proposals, NDN promises better performance and resilience against current Internet attacks. However, NDN's resilience has not been largely analyzed yet, in special the flooding attacks that exploit the content request and distribution data structure in NDN routers (called Pending Interest Table - PIT). This paper focuses on analyzing this type of denial of service (DoS) attack. It proposes an analytical model that helps to understand the conditions that make the architecture more or less susceptible to this threat. Evaluation shows that the model is useful to analyze the circumstances in which the PIT is more vulnerable to flooding attack. An extension of the model is used to formulate an optimization function which maximizes the system throughput, minimizing the effects of a DoS attack.

Content-Centric Networking is an architecture proposal for the future Internet that brings fundamental changes in the way the network operates. Contents are identified and requested based on their names and for security reasons they must be digitally signed by their publishers. Even though this new architecture was designed to be safe, one potential security threat is that malicious publishers may create polluted versions of legitimate contents, reducing their availability and degrading network resources. Because of the non-negligible overhead of checking a large number of signatures, it is not feasible to make it a mandatory task for every router, especially in the network core. In this paper, we propose CCNCheck: a mechanism in which CCN routers probabilistically check the content signatures. We evaluate the mechanism against simulations and found evidences that using CCNCheck increases the fraction of recovered contents and decreases the wastage of network resources.

BitTorrent, the immensely popular file swarming system, suffers a fundamental problem: content unavailability. Although swarming scales well to tolerate flash crowds for popular content, it is less useful for unpopular content as peers arriving after the initial rush find it unavailable. In this paper, we present a model to quantify content availability in swarming systems. We use the model to analyze the availability and the performance implications of bundling, a strategy commonly adopted by many BitTorrent publishers today. We find that even a limited amount of bundling exponentially reduces content unavailability. For swarms with highly unavailable publishers, the availability gain of bundling can result in a net decrease in average download time. We empirically confirm the model's conclusions through experiments on PlanetLab using the Mainline BitTorrent client.

BitTorrent, the immensely popular file swarming system, suffers a fundamental problem: unavailability. Although swarming scales well to tolerate flash crowds for popular content, it is less useful for unpopular or rare files as peers arriving after the initial rush find the content unavailable.

The deployment of wireless LANs (WLANs) has been steadily increasing over the years and estimating the actual bit rate of a WLAN device is important for management and applications such as those that can adapt their transmission rates according to the network characteristics. We propose a simple and accurate active measurement technique to infer the bit rate of an IEEE802.11 device. The proposed method is based both on a recently proposed technique to infer the type of access network and on the packet pair approach, but adapted to take into account the overhead caused by the IEEE802.11 control and the existence of concurrent WLAN traffic. Furthermore, the technique does not require the WLAN to be the bottleneck link in the path from the measuring point to the end computer. Results from simulation and from measurements show that the approach is accurate to infer WLAN access point rates both in scenarios where the WLAN devices can adapt their rates as well as WLANs with fixed transmission rates.