Kubernetes (K8s) has become the cornerstone solution for orchestrating containerized applications across cloud-edge computing environments. Container Network Interface (CNI) plugins are crucial in this ecosystem, enabling network connectivity among K8s components, supporting advanced features and facilitating robust network management across diverse environments, such as centralized cloud or decentralized edge locations. In this paper, we extend our previous work, focusing on the various features, either built-in or non-native, provided within each networking solution and across diverse resource-intense and lightweight K8s distributions. Along these lines, tunneling options, security and encryption mechanisms, as well as common service mesh and observability tools have been evaluated with regards to their performance and resource utilization.

Kubernetes (K8s) serves as a mature orchestration system for the seamless deployment and management of containerized applications spanning across cloud and edge environments. In this context, optimizing Kubernetes networking to achieve high-performance connectivity and minimal resource utilization is crucial for its applicability and effectiveness at the edge. This paper contributes to this effort, by conducting a qualitative and quantitative performance evaluation of diverse Container Network Interface (CNI) plugins within different K8s environments, incorporating lightweight implementations designed for the Edge. Our experimental assessment was conducted in two distinct (intra- and inter-host) scenarios, revealing interesting insights and tradeoffs for both researchers and practitioners. For example, the deployment of plugins across lightweight distributions does not necessarily lead to resource utilization improvements, e.g., in terms of CPU/memory or throughput, while, in contrast, their impact on lifecycle metrics such as pod readiness times, is significant.

Smart cities are progressively emerging and require city-wide networks able to maintain low delay and increased reliability. However, wireless communications in the urban environment are occasionally disrupted, leading to volatile network conditions and impeding traditional networks to cope with these demands. To this end, Named Data Networking (NDN) and Delay/Disruption Tolerant Networking (DTN) architectures, as well as the combined NDN-over-DTN (NoD) scheme have been introduced and are actively investigated for combating distinct communication challenges in wireless city-wide deployments. In this work we investigate the performance and reliability trade-offs of each approach in the smart city context, and accordingly propose an experimental system that employs interchangeably NDN, DTN and NoD driven by a novel Change Point (CP) based network reliability detector. The real-world experiments showed that the proposed multi-protocol adaptive system, called ALTER, leads to lower latency and increased reliability of the urban network, compared to non-adaptive NDN and NoD solutions.

Internet of Things (IoT) evolution transformed modern cities into smart cities. As smart cities are becoming part of our everyday lives, it becomes essential to address challenging communication issues that are present, such as the intermittent connectivity, the link asymmetry and the indirect line-of-sight. In this work we propose an adaptive solution that employs interchangeably Named-Data Networking (NDN), Delay/Disruption-Tolerant Networking (DTN) and NDN-over-DTN (NoD) to compensate reliability issues and enable the seamless smart city network operation using the Software-Defined Networking (SDN) technology. The implemented system has been deployed and validated in the CityLab test-bed, a real smart city network located in Antwerp, Belgium. The evaluation results show that our adaptive protocol solution outperformed the single NoD scheme in terms of Content Retrieval Delay, Interest Satisfaction Ratio and Cache Hit Ratio.

We propose an SDN-based Non-IP multi-protocol platform (REWIRE) that mixes and matches, on-demand, multi-ple Non-IP protocol strategies with real rapidly-detected network conditions and loT data communication patterns. The REWIRE solution supports centralized monitoring of Wireless Mesh Net-works (WMN) and manages alternative Non-IP protocols stacks (i.e., NDN, DTN & NoD) in an adaptable manner based on change-point analysis & clustering mechanisms. To this end, our proposal grafts flexibility and adaptability capabilities to the WMN providing the communication backbone in real Smart-City environments. Our platform was implemented in real WMNs environments over Fed4FIRE+ test-beds considering an loT sce-nario using traffic patterns based on real sensor measurements.

Programmable wireless environments (PWEs) utilize internetworked intelligent metasurfaces to transform wireless propagation into a software-controlled resource. In this article, the interplay is explored between the user devices, the metasurfaces, and the PWE control system from the theory to the end-to-end implementation. This article first discusses the metasurface hardware and software, covering the complete workflow from the user device initialization to its final service via the PWE. Furthermore, to be compatible with the 5G and 6G wireless systems, the software-defined networking (SDN) paradigm is extended to achieve scalable internetworking and central control in PWE deployments with multiple metasurfaces and multihop communication. Subsequently, the set of SDN foundations is exploited in order to abstract the physics behind PWEs and a theoretical framework is established to describe and manipulate them in an algorithmic form. This can lead to smart radio environments that are readily accessible from various engineering disciplines, facilitating their integration into existing networks, wireless systems, and applications. This article is concluded by outlining strategies for the optimal placement of metasurfaces within a PWE-controlled space, open challenges in PWE security, specialized SDN integration issues, and theoretical problems toward the graph-driven modeling of PWEs.

Edge cloud computing is a crucial enabler for applications or network services with stringent communication requirements in the 5G networks and beyond era. Although edge compute resources contribute to a significant portion of end-to-end performance (e.g., delay), the involved performance trade-offs in relation to the utilized virtualization technologies and application functions are not yet thoroughly investigated. In this context, we study the performance dynamics of alternative edge cloud technologies, including different unikernel flavors, container builds, and implementations of exemplary Web services. Our experiments do not target identifying a single best option, rather than the involved performance trade-offs with respect to both service and infrastructure viewpoints as well as realistic edge cloud tasks, including on service operation, cloud resource or service elasticity, dynamic resource allocation and removal. We provide insights gained from a relevant extensive comparative evaluation based on a novel edge cloud experimentation infrastructure. Our experimental results inspired a conceptual edge cloud orchestration platform and its basic design guidelines, i.e., mixing and matching heterogeneous edge cloud virtual entities with particular operational or performance requirements.

The emerging Internet of Things (IoT) in conjunction with Autonomous Vehicles (AVs) have increased the complexity of network administration and control. In this context, recent proposals bring together Software-Defined Networks (SDNs) with AVs practices, i.e., introducing innovative network control strategies bespoke for smart city infrastructures. However, an experimentation environment that combines the realism of an AVs ecosystem with the SDN network paradigm is a necessity, facilitating the investigation of particular research issues, including efficient data management and routing, as well as security. In this demo, we introduce a relevant facility that builds-up on novel open environments for hands-on experimentation, namely: (i) CARLA, an open realistic urban-driving simulator; (ii) Cooja, a state-of-the-art emulator for Wireless Sensor Networks (WSNs); and (iii) SD-MIoT, a novel SDN solution for mobile IoT. Our experimentation exercise currently focuses on maintaining connectivity of AVs to the fixed infrastructure, e.g., Road Side Units (RSUs), with SDN strategies. We demonstrate the capabilities of the proposed experimentation environment with proof-of-concept results on packet delivery ratio and control overhead, quantifying the efficiency of the considered SDN approach to maintain AV connectivity, as well as with a video visualizing the outcome of our experiments.

The Internet of Things (IoT) brings new potentials in humans' everyday life activities through enabling applications with stringent communication demands, including mobile IoT applications. State-of-the-art routing protocols for the IoT, such as IPv6 routing protocol for low-power and lossy networks (RPL), have not been originally designed for such challenging applications. Recent proposals blend the software-defined networking (SDN) paradigm with IoT, enabling better-informed and bespoke routing control, matching the application requirements. In this article, we propose SD-MIoT, an open-source SDN solution for mobile IoT environments that consists of a modular SDN controller and an OpenFlow-like protocol. SD-MIoT detects passively in real-time, the network's mobile nodes through mobility detector (MODE), an intelligent algorithm that utilizes the connectivity graph of the SDN controller. It incorporates novel mobility-aware topology discovery mechanisms, routing policies, and flow-rule establishment methods, all of them balancing control overhead with routing robustness, according to nodes' mobility behavior detected by MODE. We provide extensive evaluation results over two realistic scenarios, further confirming the suitability of SD-MIoT for mobile IoT environments and demonstrating reliable operation in terms of successful data packet delivery with low control overhead.

The microservices design paradigm enables applications, usually based on containers, exploiting the flexibility of cloud computing and bringing unique scalability, fault-tolerance and resource-allocation benefits. A number of orchestration facilities, including Kubernetes, target the efficient deployment and operation of containers and are mainly focusing on the maintenance of server resource allocation under predefined thresholds, i.e., through scaling up or down containers to mitigate dynamic changes in the workload. In this work, we highlight the technical capabilities and cost-saving impact of microservices in contrast to traditional monolithic applications, based on a techno-economic analysis. We also investigate the service performance vs resource allocation trade-off, uncovering interesting dynamics when elasticity is driven from service quality metrics. This approach allows the Service Providers (SPs) to balance their profit margins with the customer satisfaction, i.e., reducing the infrastructure cost while keeping the service performance at an acceptable level.

Dynamic end-to-end cloud network slices following a Slice-as-a-Service (SaaS) paradigm offer promising opportunities in support of diverse vertical industries attached to different administrative domains. This paper presents a distributed Marketplace model to dynamically create cloud network slices involving different slice resource providers. Portrayed through the lens of three reference use case scenarios, we discuss potential attributes of having a Marketplace and the outcomes feeding the use case requirements within a Slice Marketplace scope. Effective selection, negotiation, and proper monitoring and enforcement of service level agreements between multiple administrative and technological domains are among the prime challenges towards the realization of a distributed Marketplace of federated network and cloud resource providers.

The 5G networks enable new touristic services with challenging communication requirements, such as augmented reality (AR) applications, and allow the visitors to enjoy a touristic experience that involves both the physical and virtual space. Here, we propose a novel multi- user travel itinerary planning framework based on an optimal problem formulation that considers both individual trip itinerary (e.g., tourist's preferences, time or cost) and touristic service constraints (e.g., nearby edge cloud resources and application requirements). The main idea is to maximize the itinerary score of individual visitors, while also optimizing the resource allocation at the edge. We consider two services, video streaming and AR, and evaluate our framework using data from Flickr. Results demonstrate gains up to 100% in the resource allocation and user experience in comparison with a state-of-the-art solution adapted to this scenario.

Cloud network slicing can be defined as the process that enables isolated end-to-end and on-demand networking abstractions, which: (a) contain both cloud and network resources, and (b) are independently controlled, managed and orchestrated. This paper contributes to the vision of the NECOS project and relevant platform, that aim to address the limitations of current cloud computing infrastructures to accomplish the challenging requirements of the slicing approach. The NECOS platform implements the Slice-as-a-Service model, enabling the dynamic creation of end-to-end (E2E) slices from a set of constituent slice parts contributed from multiple domains. A challenging issue is to define the facility that implements dynamic slice resource discovery, aligned to the requirements of the slice owner or tenant, over different infrastructure providers. Here, we propose a Marketplace-based approach implementing relevant federated interactions for the resource discovery and we detail its architecture, workflows, and information model. We also present its initial implementation details and provide both quantitative and qualitative experimental results validating its main operation.

Video content is responsible for more than 70% of the global IP traffic. Consequently, it is important for content delivery infrastructures to rapidly detect and respond to changes in content popularity dynamics. For flexible and highly adaptive solutions, the capability for a quick response should be driven from early (real-time) and low-complexity content popularity detection schemes. In this paper, we focus on the early and low-complexity detection of video content popularity, which we address as a statistical change point (CP) detection problem. Our proposed methodology estimates in real-time the existence, the number, the magnitude and the direction of changes in the average number of video visits by combining: (i) off-line and on-line CP schemes; (ii) an improved measurements window segmentation heuristic for the detection of multiple CPs; and (iii) a variation of the moving average convergence divergence (MACD) indicator to detect the direction of changes. We evaluated the proposed framework using a large database of real youtube video visits. The proposed algorithm is shown to accurately identify CPs and the direction of change in the off-line phase. Finally, a few illustrative examples of two variations of the on-line algorithm are also included.

Being part of the Internet of Things (IoT), the Wireless Sensor Networks (WSNs) inherit characteristics such as large-scale deployment, dynamicity, heterogeneity and mobility. These aspects mandate elasticity in many network functions and especially in routing. The IPv6 Routing Protocol for Low-Power and Lossy Networks (RPL) is the state-of-the-art routing protocol for resource-constrained devices in environments with signal issues, but it was not designed to support mobility. However, mobility is fundamental in critical WSNs applications, e.g., it improves sensing coverage and brings back connectivity. In this paper, we propose a centralized routing control approach, which exploits the global view of the network inspired by the Software-Defined Networks (SDNs), in order to provide dynamic end-to-end routing service in heterogeneous environments. We briefly present our novel Cross-Layer Control of Data Flows (CORAL) framework, which-in the context of the current work- has been enhanced with management and control features to provide elasticity in the RPL's functionality. In practice we developed and propose dynamic and individual parameters' configuration and adaptation, live monitoring of performance through visualization and Ansible-based protocol's deployment. Our results show that dynamic RPL configuration, which tackles mobility issues on-the-fly, along with individual configuration which handles them separately, can efficiently tune the protocol's performance trade-offs, e.g., between packet delivery ratio (PDR) and routing overhead, bringing up to 35% improvement in PDR and offloading the control overhead from the mobile nodes.

In Ethernet Passive Optical Networks (EPONs), the equipment placed at the customer premises, i.e., the Optical Network Units (ONUs), has been shown to be responsible for almost 65% of the total EPON power consumption. Sleep mechanisms, implemented at ONUs' side, can contribute to the EPONs energy efficiency. However, the trade-off between energy saving and Quality of Service (QoS) requirements should be carefully tuned, especially when the downstream transmission is considered, in order to achieve the desirable results. In this paper, we exploit formal methods in order to build a holistic model representing both the state machine of the ONU in its details as well as the ONU's communication with the Optical Line Terminal (OLT) under EPONs specifications. The quantitative results of our analysis highlight the impact of the non-active periods on the aforementioned trade-off and shows how they can be configured in line with the network parameters.

New flexible network environments emerged in the last years, such as the Software-Defined Networks, the Network Function Virtualization and the various network slicing technologies. Their main goal is to satisfy and to flexibly cope with the changing requirements of both users and service / infrastructure providers with respect to the constraints of physical resources, the frequently changing network conditions, and with strategies that resolve performance bottlenecks by tuning the various performance trade-offs.

Novel evolutions in the networking world have been proposed under the umbrella of 5G initiatives. The NFV concept is related with the building blocks for virtual networks that are characterized as highly dynamic network environments. SDN performs logically-centralized network control that is decoupled from the data plane, enabling holistic network management. Furthermore, there is a recent trend towards lightweight virtualized network devices and servers bringing significant advantages in terms of adaptability and responsiveness to the network and service environment dynamics. The above paradigms can be combined together, resulting in unprecedented flexibility in Software Defined Infrastructures (SDI) operations. Such unified environments require new efficient and distributed management facilities that are characterized by scalability, reliability, and adaptability to the dynamic conditions in terms of resource availability and changing service and infrastructure requirements. To assist the evaluation of these components, we developed a distributed facility for testing, evaluating, and experimenting with the management of these SDI environments — the Very Lightweight Software-Driven Network and Services Platform (VLSP). It exhibits the following properties: (i) it is a complete integrated management platform for SDI environments; and (ii) it is distributed and scalable, making it suitable for a wide range of topologies and network service deployments.

The Internet landscape is gradually adopting new communication paradigms characterized by flexibility and adaptability to the resource constraints and service requirements, including network function virtualization (NFV), software-defined networks, and various virtualization and network slicing technologies. These approaches need to be realized from multiple management and network entities exchanging information between each other. We propose a novel information exchange management as a service facility as an extension to ETSI's NFV management and orchestration framework, namely the virtual infrastructure information service (VIS). VIS is characterized by the following properties: 1) it exhibits the dynamic characteristics of such network paradigms; 2) it supports information flow establishment, operation, and optimization; and 3) it provides a logically centralized control of the established information flows with respect to the diverse demands of the entities exchanging information elements. Our proposal addresses the information exchange management requirements of NFV environments and is information-model agnostic. This paper includes an experimental analysis of its main functional and non-functional characteristics.

This paper addresses the problem of provisioning management/monitoring nodes within highly dynamic network environments, particularly virtual networks. In a network, where nodes and links may be spontaneously created and destroyed (perhaps rapidly) there is a need for stable and responsive management and monitoring, which does not create a large load (in terms of traffic or processing) for the system. A subset of nodes has to be chosen for management/monitoring, each of which will manage a subset of the nodes in the network. A new, simple, and locally optimal greedy algorithm called Pressure is provided for choice of node position to minimize traffic. This algorithm is combined with a system for predicting the lifespan of nodes, and a tunable parameter is also given so that a system operator could express a preference for elected nodes to be chosen to reduce traffic, to be “stable,” or some compromise between these positions. The combined algorithm called PressureTime is lightweight and could be run in a distributed manner. The resulting algorithms are tested both in simulation and in a testbed environment of virtual routers. They perform well, both at reducing traffic and at choosing long lifespan nodes.

The current Internet does not enable easy introduction and deployment of new network technologies and services. This paper aims to progress the Future Internet (FI) by introduction of a service composition and execution environment that re-use existing components of access and core networks. This paper presents essential service-centric platforms and software systems that have been developed with the aim to create a flexible environment for an Autonomic Internet.

There has recently been an increasing research interest in network management infrastructures that autonomously adapt to the dynamics of the environment. In this paper, we present an information management platform, called the Information Management Overlay (IMO), which is an infrastructure that regulates information flow based on the properties and the state of the network environment. We discuss the design, implementation and optimization issues of this platform and we investigate experimentally how the IMO can adapt to different settings and optimization requirements.

We discuss a new packet service paradigm, called "Less Impact Better Servicerdquo (LIBS), which is realized through a novel queuing discipline, called "noncongestive queuingrdquo (NCQ). NCQ prioritizes small packets when conditions permit and utilizes service thresholds to confine the delay impact of prioritization on congestive applications. We show that LIBS and NCQ satisfy more users with diverse demands on delay and throughput. We obtained both analytical and simulation results, which are very promising. Diversity is simulated by using FTP, sensor, and VoIP traffic.

Future mobile services are expected to have different levels of QoS requirements, raising the need to enhance networks with the capacity to differentiate among different classes of traffic, which may use different QoS models. To support and shelter different QoS models, future mobile networks should make use of an inter-network mechanism allowing the establishment of bi-lateral service level specifications (SLSs), without making any assumption about the signalling supported by internal network devices. Hence, this paper aims to explore the best way to use bi-lateral signalling to build end-to-end chains of SLSs. Two methods are evaluated. In the first one, control messages are only propagated after the establishment of an SLS between a pair of networks; in the second one control messages flow from the provider of the SLS to each of its customers, passing a chain of transit-networks.

We investigate the energy-saving potential of transport protocols. We seek an answer to strategic issues of maximizing energy and bandwidth exploitation, without damaging the dynamics of multiple-flow equilibrium. We claim that (i) an energy-saving strategy of the transport level needs to be associated with some energy potential index which, unlike energy expenditure, is not device-specific and (ii) system-wise an energy-efficient system of flows is not always a better choice: we show that a less energy-efficient system may be more reliable in terms of packet multiplexing and, in turn, may reduce the probability that some flows may expend their energy with zero gain. We present results using a real testbed

We show that TCP timers, based solely on RTT estimations and measurements, cannot capture with precision the level of flow contention. We notice that increased contention may stabilize RTT variation, minimize the deviation and, in turn, shorten the timeout. We show that this behavior is undesirable indeed, since it leads to unfair resource utilization. We propose CA-RTO, an algorithm that incorporates a contention parameter and a randomization technique into the retransmission timeout. We report significant improvement in fairness, great reduction of retransmitted packets and slight improvements in application goodput.

We investigate the behavior of TCP(/spl alpha/,/spl beta/) protocols in the presence of wireless networks. We seek an answer to strategic issues of maximizing energy and bandwidth exploitation, without damaging the dynamics of multiple-flow equilibrium. Our perspective is novel indeed: What is the return of the effort that a protocol expends? Can we achieve more gains with less effort? We study first the design assumptions of TCP(/spl alpha/,/spl beta/) protocols and discuss the impact of equation-based modulation of /spl alpha/ and /spl beta/ on protocol efficiency. We introduce two new metrics to capture protocol behavior: The "extra energy expenditure" and the "unexploited available resource index". We confirm, by simulation experiments, that, in general, smoothness and responsiveness constitute a tradeoff; however, we show that this tradeoff does not graft its dynamics into a conservative/aggressive behavior, as is traditionally believed. We uncover patterns of unjustified tactics; our results suggest that an adaptive congestion control algorithm is needed to integrate the dynamics of heterogeneous networks into protocol behavior.