Nelson Fonseca

State University of Campinas (UNICAMP) | UNICAMP · Institute of Computing

Self-adaptation of service function chains (SFCs) has been considered as an important attribute to ensure the resource-efficiency and reliability of network function virtualization (NFV) systems. In this work, we leverage the idea of integrating segment routing over IPv6 (SRv6) and in-band network telemetry (INT) seamlessly to realize SRv6-INT and...

Characterizing and monitoring patient activities through time series data is critical for identifying lifestyle patterns that may impact health outcomes. Sedentary behavior is a significant concern due to its association with various health risks. This study introduces a lightweight supervised classifier for healthcare applications based on ordinal...

Accurate modeling of realistic network topologies is essential for evaluating novel Internet solutions. Numerous investigations have used topologies generated by graph generators employing scale-free-based models. Although scale-free networks accurately encode node degree distribution, they overlook crucial graph properties, such as betweenness, cl...

Passive Optical Network (PON) technology offers a cost-effective alternative to support Beyond 5G Mobile Network Fronthauling (MFH). However, MFH dimensioning for such networks is challenging, given its high bandwidth and strict latency requirements. The Functional Split of the Radio Access Network (RAN) has been introduced to provide flexibility i...

Federated Learning is a promising technique for providing distributed learning without clients disclosing their private data. In Hierarchical Federated Learning, edge servers partially aggregate the parameters of their connected clients’ models, improving scalability and reducing computational overhead on the central server. To speed up the converg...

The Zero-touch Network & Service Management (ZSM) paradigm, a direct response to the increasing complexity of communication networks, is a problem-solving approach. In this paper, taking advantage of recent advances in generative Artificial Intelligence, we introduce the Network ConFiguration Generator (LLM-NetCFG) that employs Large Language Model...

Data plane programmable (PDP) switches are becoming increasingly popular for network function virtualization (NFV), for their programmability and high packet processing performance. However, the inherent limitations of PDP switches, such as limited memory space, make it challenging to implement certain types of network functions (NFs) (i.e., the st...

This paper summarizes the thesis ”MEC Placement Problem in Protected 5G Networks”, which tackles the question of where to position Multiaccess edge computing nodes (MECs) in 5G networks to achieve stringent 5G requirements at a minimum cost. The thesis formulated the MEC placement problem with traditional protection schemes to provide redundancy in...

Este artigo apresenta a avaliação de desempenho de um mecanismo escalável de acesso aleatório para atender as demandas de escalabilidade no acesso aleatório aos recursos de rádio em redes móveis. O mecanismo permite acessar a rede utilizando uma palavra-código formada pela agregação de n preâmbulos ortogonais transmitidos em slots de acesso aleatór...

The zero-touch network and service management (ZSM) paradigm, a direct response to the increasing complexity of communication networks, is a problem-solving approach. In this article, taking advantage of recent advances in generative artificial intelligence, we introduce the network configuration generator (LLM-NetCFG) that architects ZSM configura...

This paper introduces a novel deep-learning based generator of synthetic graphs that represent intra-Autonomous System (AS) in the Internet, named Deep-generative graphs for the Internet (DGGI). It also presents a novel massive dataset of real intra-AS graphs extracted from the project Internet Topology Data Kit (ITDK), called Internet Graphs (IGra...

This paper summarizes the thesis “The Fog Node Location Problem”, which attempted to answer the question of how fog nodes should be located to process end-user demands that are variable in time and space. The problem was studied from different perspectives, optimizing the number of served users, deployment costs, energy consumption, and latency. Th...

A key component of the Internet of things (IoT) ecosystem is wide-area network connectivity, for which cellular network technologies are a promising option through their support of massive machine-type communication (mMTC). However, numerous devices transmitting sporadically small data packets in a highly synchronized way can generate overload on t...

Federated learning has been explored as a promising solution for training machine learning models at the network edge, without sharing private user data. With limited resources at the edge, new solutions must be developed to leverage the software and hardware resources as the existing solutions did not focus on resource management for network edge,...

With the advent of 5G networks, telecommunications infrastructure providers (InP) have faced numerous challenges as they attempt to meet the stringent quality of service requirements. The placement of applications at the edge of the mobile network in Multi-access Edge Computing (MEC) and slicing techniques have provided powerful tools to enable net...

This letter analyzes the performance of the TCP CUBIC and TCP BBR protocols in the presence of background traffic. The analysis is performed via emulation using actual TCP implementations and considering high capacity end-to-end data connections and different connection durations (e.g., mouse and elephant flows). The results indicate that the BBR p...

Reliability is essential in Smart Farming supported by the IoT-Fog-Cloud continuum. Smart Farms’ unprotection may cause significant economic losses and low yields of production. This paper introduces an optimization model for providing reliability and, consequently, service continuity to the IoT-Fog-Cloud continuum-based smart farms. The proposed m...

Reinforcement Learning (RL) comes with the promise of automating network management. However, due to its trial-and-error learning approach, model-based RL (MBRL) is not applicable in some network management scenarios. This paper explores the potential of using Automated Planning (AP) to achieve this MBRL in the functional areas of network managemen...

The great interest in flix-like services has amplified multimedia traffic over the Internet. Recently released traffic forecasting predicts that video-related traffic will be responsible for the majority of Internet traffic by 2022. Such traffic will come in a wide range of duration and in the two modes of live and on-demand. Additionally, it is ex...

As a promising technology, Flex Ethernet (FlexE) helps to realize deterministic and ultra-low latency in metro and transport networks. Meanwhile, previous studies have confirmed the advantages of the symbiosis of FlexE and elastic optical network (EON) (i.e., a FlexE-over-EON) on resource utilization and cost-effectiveness. In this paper, we consid...

Network slicing is a promising technology for providing customized logical and virtualized networks for the fifth-generation (5G) use-cases (enhanced mobile broadband, ultrareliable low-latency communications, and massive machine-type communications), which pose distinct quality of service (QoS) requirements. Admission control and resource allocati...

The standardisation of 5G is reaching its end, and the networks have started being deployed. Thus, 6G architecture is under study and design, to define the characteristics and the guidelines for its standardisation. In parallel, communications based on quantum-mechanical principles, named quantum communications, are under design and standardisation...

Federated Learning (FL) is a distributed machine learning type of processing that preserves the privacy of user data, sharing only the parameters of ML models with a common server. The processing of FL requires specific latency and bandwidth demands that must be fulfilled by the operation of the communication network. This paper introduces two Dyna...

Resources of Passive Optical Networks (PONs) can be shared by a diverse spectrum of customers such as residential users, business subscribers, mobile network operators, and service providers. A PON customer can rent one or several Optical Network Units (ONUs) from an Infrastructure Service Provider. The ONUs of a customer (multi-ONU customer) can b...

Access Networks based on Passive Optical Network (PON) technology can offer transport services to broadband service providers, such as mobile backhauling/fronthauling for Mobile Network Operators and multi-ONU Service Level Agreements for Virtual Network Operators. Besides that, Infrastructure Service Provider (InP) also provides broadband access s...

Traditional routing protocols employ limited information to make routing decisions, which leads to slow adaptation to traffic variability and restricted support to the quality of service requirements of applications. To address these shortcomings, in previous work, we proposed RSIR, a routing solution based on Reinforcement Learning (RL) in Softwar...

This paper introduces a new solution to improve network performance by decreasing spectrum fragmentation, crosstalk interference, blocking of virtual networks, cost, and link load imbalance. These problems degrade the performance of Elastic Optical Networks with Space-Division Multiplexing. The proposed solution, called Cognitive control loop (CO-O...

This paper provides a theoretical model for the computation of the probability density function of multi-hop broadcast latency when using probabilistic broadcasting schemes in Wireless Sensor Networks (WSN). The probability density functions of multi-hop latency are derived by considering the per-hop random delays employed in protocols and the geom...

In fog computing, processing, network, and storage resources are placed close to end users to ensure low latency in comparison to the latency experienced when accessing the cloud. One limitation of this solution, however, is that fog nodes are usually fixed, whereas demands are variable over time at all locations, which may result in either under-...

This article introduces four multipath routing, core, and spectrum assignment algorithms, all of which employ image processing techniques for fast identification of allocable slots, since this is less computationally complex and adequate for real-time operation. Numerical results show that multipath routing can decrease the probability of blocking...

Federated Learning (FL) is a distributed machine learning (ML) type of processing that preserves the privacy of user data, sharing only the parameters of ML models with a common server. The processing of FL requires specific latency and bandwidth demands that need to be fulfilled by the operation of the communication network. This paper introduces...

Passive optical network (PON) technology offers an attractive cost-efficient alternative to support 5G and Beyond 5G mobile network fronthauling (MFH). However, MFH for such networks is challenging given its high bandwidth and strict latency requirements. To reduce these requirements, radio access network functional splitting has been introduced in...

Traditional routing protocols employ limited information to make routing decisions, leading to slow adaptation to traffic variability and restricted support to applications quality of service requirements. This paper introduces the work developed in the MSc. thesis entitled "Routing based on Reinforcement Learning for Software-Defined Networking",...

This paper summarizes the research in the master thesis entitled "Admission Control and Resource Allocation in 5G Network Slicing". We propose two solutions, SARA and DSARA, based on Reinforcement Learning algorithms to learn the admission policy that optimizes the profit of providers. Resource allocation considers the QoS requirements. Results sho...

Federated learning (FL) is a framework to train machine learning models using decentralized data, especially unbalanced and non-iid. Adaptive methods can be used to accelerate convergence, reducing the number of rounds of local computation and communication to a centralized server. This paper proposes an adaptive controller to adapt the number of e...

Federated learning has been explored as a promising solution for training at the edge, where end devices collaborate to train models without sharing data with other entities. Since the execution of these learning models occurs at the edge, where resources are limited, new solutions must be developed. In this paper, we describe the recent work on re...

Code-expanded Random Access (CeRA) is a promising technique for supporting massive machine-type communications in cellular networks. However, its potentiality is limited by code ambiguity, which results from the inference of a larger number of codewords than those actually transmitted. In this letter, we propose a random access (RA) scheme to allev...

Code-expanded Random Access (CeRA) is a promising technique for supporting mMTC in cellular IoT networks. However, its potentiality is limited by code ambiguity, which results from the inference of a larger number of codewords than those actually transmitted. In this letter, we propose a random access scheme to alleviate this problem by allowing de...

The lack of a Management Plane specification in the software-defined networking (SDN) is a fundamental problem for the integrated management of networks that follow this paradigm. Several solutions have addressed this problem, but they do not provide data models intended to support Fault, Configuration, Accounting, Performance, and Security (FCAPS)...

Traditional routing protocols employ limited information to make routing decisions which leads to slow adaptation to traffic variability and restricted support to the quality of service requirements of the applications. To address these shortcomings, in previous work, we proposed RSIR, a routing solution based on Reinforcement Learning (RL) in Softw...

Traditional routing protocols employ limited information to make routing decisions which leads to slow adaptation to traffic variability and restricted support to the quality of service requirements of the applications. To address these shortcomings, in previous work, we proposed RSIR, a routing solution based on Reinforcement Learning (RL) in Softw...

Network Slicing is a promising technology for providing customized logical and virtualized networks for the industry’s vertical segments.This paper proposes SARA and DSARA for the performance of admission control and resource allocation for network slice requests of eMBB, URLLC, and MIoT type in the 5G core network. SARA introduced a Q-learning bas...

Network Slicing is a promising technology for providing customized logical and virtualized networks for the industry’s vertical segments.This paper proposes SARA and DSARA for the performance of admission control and resource allocation for network slice requests of eMBB, URLLC, and MIoT type in the 5G core network. SARA introduced a Q-learning bas...

One of the significant limitations in the evaluation of hybrid clouds is the difficulty in validating new solutions by their deployment in real systems since replication of tests on large environments is usually highly expensive. To overcome these limitations, simulators have become popular for conducting preliminary tests. However, there is still...

Traditional routing protocols employ limited information to make routing decisions, which can lead to a slow adaptation to traffic variability, as well as restricted support to the Quality of Service (QoS) requirements of applications. This paper introduces a novel approach for routing in Software-Defined Networking (SDN), called Reinforcement Lear...

Fog computing extends cloud services to the edge of the network. In such scenario, it is necessary to decide where applications should be executed so that their quality of service requirements can be supported. Thus, a cloud-fog system requires an efficient task scheduler to decide the locality where applications should run. This paper presents two...

In fog computing, processing, network, and storage resources are placed close to the end users to assure a low latency in comparison to the latency experienced when accessing services in the cloud. One limitation of this solution, however, is that fog nodes are usually fixed, whereas demands are variable over time at all locations, resulting in und...

In fog computing, processing, network, and storage resources are placed close to the end users to assure a low latency in comparison to the latency experienced when accessing services in the cloud. One limitation of this solution, however, is that fog nodes are usually fixed, whereas demands are variable over time at all locations, resulting in und...

This paper studies the cross-layer network planning that tries to combine flexible Ethernet (FlexE) and elastic optical networks (EONs), for FlexE-over-EONs. We focus our investigation on the most challenging setting, i.e., the FlexE-over-EONs based on the FlexE-aware architecture, and consider both single-hop and multi-hop scenarios for the cross-...

Passive optical network (PON) is a cost-efficient access network technology to deliver broadband services. Moreover, service providers employ PONs to offer novel services. New business scenarios are, thus, envisioned in which customers owning multiple optical network units (ONUs) are connected to a single PON (multi-ONU customers). This paper propo...

This paper presents the work developed in the MSc. thesis entitled ”Resource Allocation in Space-Division Multiplexing-Based Elastic Optical Networks” which explores the resource allocation problem in Space-Division Multiplexing-based Elastic Optical Networks (SDM-EON). The proposed algorithms jointly address spectrum fragmentation, modulation form...

O fluxo contínuo de aplicações e serviços digitais emergentes tem impulsionado um enorme crescimento de tráfego no núcleo da Internet e consequentemente o estudo de tecnologias para aumentar a capacidade dessas redes. A multiplexação por divisão espacial é uma solução promissora para lidar com o esgotamento esperado da capacidade dessas redes. Adic...

Passive Optical Networks (PONs) belonging to Infrastructure Providers are usually shared by diverse customers such as mobile network operators, virtual service providers, and multi-site enterprises. In this scenario, PON customers may own single or multiple Optical Network Units (ONUs). Moreover, some ONUs belonging to a customer can be underutiliz...

Although network function virtualization (NFV) has been proven to be beneficial in terms of equipment cost, service delivery flexibility, and time-to-market, most of the studies in this area only addressed homogeneous NFV platforms (e.g., with virtual machines (VMs) only). In this work, we argue that by leveraging heterogeneous NFV platforms such a...

In Elastic Optical Networks with Space Division Multiplexing, the dynamic allocation and deallocation of frequency slots can generate spectrum fragmentation, which increases the blocking of requests for lightpath establishment. In this article, we introduce a reactive algorithm and a proactive one that can jointly reduce spectrum fragmentation. We...

Infrastructure service providers (InPs) can employ bandwidth sharing to offer new services and business models to their customers. In this paper, we introduce a dynamic bandwidth allocation (DBA) algorithm which allows cooperation among Ethernet PON (EPON) customers so that they can share unused bandwidth among themselves without affecting their gu...

New applications and the number of mobile devices connected to the Internet have increased considerably. Fog computing has been arising to allow the deployment of latency-sensitive applications which cannot be processed in the cloud. In fog computing, small facilities, known as fog nodes, provide processing, storage, and networking resources to use...

The support of stringent requirements such as ultra-low latency and ultra-reliability of the forthcoming 5G services poses several challenges to telecommunications infrastructure providers. Network Function Virtualization, multi-access edge computing (MEC), and network slicing capabilities can help the support of such requirements. However, a trade...

Currently, Internet applications running on mobile devices generate a massive amount of data that can be transmitted to a Cloud for processing. However, one fundamental limitation of a Cloud is the connectivity with end devices. Fog Computing overcomes this limitation and supports the requirements of time-sensitive applications by distributing comp...

Although the forthcoming 5G era will significantly reshape today's information society, there are still challenges to support the stringent quality-of-service (QoS) requirements from emerging applications. This motivates the exploration of new network architectures to overcome the inherent limitations of conventional IP-based technologies. Therefor...

This paper summarizes the dissertation ”Energy-aware load balancing in distributed data centers”, which proposed two new algorithms for minimizing energy consumption in cloud data centers. Both algorithms consider hierarchical data center network topologies and requests for the allocation of groups of virtual machines (VMs). The Topology-aware Virt...

Fog computing brings the advantages and power ofcloud computing to the edge of the network. Software-DefinedNetworking (SDN) has been considered as a feasible solution to cope with the complexity of the orchestration of fog devices. Nevertheless, the use of an SDN controller introduces delays into the transport of packet flows in the fog layer, whi...

Passive optical networks (PONs) provide high access capacity to multiple services and applications. However, the deployment of PONs involves significant costs. On the other hand, infrastructure provider (InP) can lease their PONs to several customers. In this paper, we introduce a novel dynamic bandwidth allocation (DBA) algorithm for 10G-EPON call...

LTE-Advanced (LTE-A) is expected to be one of the main platforms for Massive Machine-Type Communications (mMTC) system. The provision of Quality of Service (QoS) in future cellular networks strongly depends on efficient allocation of resources in the downlink control channel. However, the design of the downlink control channel is not flexible enoug...

Smart grids are expected to provide various benefits to society by integrating advances in power engineering with recent developments in the field of information and communications technology. One of the advantages is the support to efficient demand‐side management (DSM), for example, changes in consumer demands for energy based on using incentives...

The processing of big data generated by the Industrial Internet of Things (IIoT) calls for the support of processing at the edge of the network as well as at the cloud data centers. The Equal-Cost Multi-Path (ECMP), which is the default routing technique in the cloud data centers, can degrade the network performance when handling mouse and elephant...

Energy efficiency has become an essential component in the design of communication protocols, especially for battery-enabled user equipment (UE) devices such as smartphones. In this paper, an energy-efficient UE packet scheduling scheme for LTE/LTE-Advanced cellular networks is proposed. It achieves energy savings by promoting the simultaneous occu...

In this paper, we summarize the contributions of the Doctoral Thesis entitled Protection in Elastic Optical Networks with Spatial Multiplexing. This thesis proposes solutions for protection of elastic optical networks with spatial multiplexing, aiming at reducing the blocking of requests and the improvement of resource utilization. For this, we tak...

The introduction of space division multiplexing in optical networks brings new challenges for network protection since the lightpath can span high capacity and transmit data at different rates. In addition, these networks suffer from the fragmentation of the spectrum that hinders contiguity and continuity constraints and, therefore, increases the b...

This paper presents heuristics based on FIPP p-cycle as a solution for the protection problem in elastic optical networks. The proposed algorithms employ failure independent path protecting p-cycle (FIPP). The algorithms differ by the protection of either one or two simultaneous failure as well as by the employment of traffic grooming and spectrum...

In the fog computing paradigm, fog nodes are placed on the network edge to meet end-user demands with low latency, providing the possibility of new applications. Although the role of the cloud remains unchanged, a new network infrastructure for fog nodes must be created. The design of such an infrastructure must consider user mobility, which causes...

A significant problem affecting the overall performance of Data Center Networks (DCNs) based on Software-Defined Networking (SDN) is the delay introduced to latency-sensitive small flows (i.e., mice) by the SDN controller. Current research works face this problem by compiling and installing the path for mice and elephants dynamically and individual...

The collision of multiple MSG 3 transmissions due to the selection of the same preamble sequence in the Long Term Evolution (LTE) Random Access procedure is an important problem which can impact on the performance of cellular Internet of Things (IoT) networks. In this paper, we propose a standard-compatible probabilistic retransmission approach to...

This letter introduces a strategy to reduce broadcast latency in multi-hop ad-hoc networks; this overall reduction is the result of the accumulation of a series of per-node delay reductions. Current broadcasting protocols employ uniformly distributed random delays on a per-node basis, while the proposed strategy consists of using a truncated-expone...