Deployment Of Network Services Research Articles

Cloud Network Function Virtualization (NFV)

Cloud Network Function Virtualization (NFV) is a groundbreaking technology that revolutionizes the way network services are delivered. By virtualizing network functions, NFV enables faster deployment, scalability, and management of network services. This article explores the benefits and use cases of NFV, including virtualized load balancers, firewalls, and intrusion detection devices. We also discuss the evolution of NFV to Cloud-Native Network Functions (CNF) and the challenges and solutions associated with NFV adoption. Discover how Cloud NFV is transforming network architecture and enabling organizations to achieve greater agility, flexibility, and cost savings.

Containers-Based Network Services Deployment: A Practical Approach

In recent years, virtualizing network servicesand functions has enabled optimizing hardware resources onresource-constrained devices, such as CPU, memory, and storage.Traditional virtualization is achieved through virtual machinesusing a layer known as a hypervisor. While this form of virtualization offers advantages such as scalability and portability, it has disadvantages in terms of performance compared to non-virtualized deployments. In this context, alternative virtualization technologies like containers allow virtualization on the same physical infrastructure, improving overall performance, portability, and service scalability. This paper implements the deployment of network services on the Raspberry Pi development platform, which has limited resources. This is achieved through a multi-container virtualization solution using the Docker Compose tool, based on Docker containerization technology. Finally, a performance analysis of the implemented virtualization solution is conducted in terms of resource utilization by each service.

Technological Aspects of IBN Technology for Slicing at 5G Multiservice Network

Purpose: The 5G network offers a variety of services, each of which requires specific parameters such as performance, throughput, reliability and latency. In this regard, network virtualization is a robust solution known as network slicing that provides support for different types of services and provides resources that match the needs of each service. The purpose of this review is to examine the implementation of a slicing platform within a 5G multiservice intent-based network that allows network operators to deploy network services in a flexible and customizable manner. Results: The considered system architecture provides the ability to monitor virtual and physical resources using OpenStack Neutron and the FlexRAN controller application. In addition, the implementation of the platform for intent-based network segmentation under consideration gives operators the ability to deploy network services in a flexible and customizable manner.

Dynamic Joint VNF Forwarding Graph Composition and Embedding: A Deep Reinforcement Learning Framework

Network Function Virtualization (NFV) is a network service deployment technology that reduces capital and operational costs while yielding flexibility and scalability for service operators. As such, an ordered chain of Virtual Network Functions (VNFs), known as a VNF Forwarding Graph (VNF-FG), should be composed and embedded into the underlying substrate network. In the literature, the composition and embedding stages of VNF-FGs are usually targeted separately, which may result in undesired solutions. In this paper, we propose our joint VNF-FG composition and embedding solution, which considers the variations of service demands while also accounting for dynamic network conditions. Specifically, our proposed solution relies on deep reinforcement learning empowered by two components for estimating dynamic parameters: network resource utilization and service demand analyzers. Moreover, to efficiently explore the problem’s large discrete action space, we utilize a specialized branching Q-network and enhance it with an action filtering mechanism. We evaluated our proposed method against joint and disjoint composition and embedding heuristics as well as versus other deep learning-based methods. Our results show that the proposed method can achieve up to a 95% improvement of embedding cost compared to our benchmarks.

Service Deployment With Priority Queueing for Traffic Processing and Transmission in Network Function Virtualization

Network function virtualization enables service providers to flexibly deploy network services. The existing works mainly focus on function placement and capacity allocation without exploring traffic scheduling for service deployment, by adopting a first-in-first-out policy. It introduces inefficiency considering that different services vary in the delay requirements. This paper proposes a service deployment model with priority queuing for both traffic processing and transmission to minimize the deployment cost with satisfying the service delay constraints. We analyze the problem including proving its NP-hardness and the convexity of a subproblem. Based on the analysis, we develop a reinforcement learning-based approach to address the problem in a decomposition manner with a polynomial-time complexity in each episode. Several specific designs are introduced to fit the learning-based approach to the considered deployment problem with priority queueing. The numerical results show that the introduced approach achieves an objective value comparable to the optimal one obtained by brute force search with a computation time 103 times shorter. Compared to a conventional model with the first-in-first-out policy, the proposed model reduces the deployment cost by adopting a more flexible queueing policy.

Research on Reliable Deployment Algorithm for Service Function Chain Based on Deep Reinforcement Learning

This paper investigates the reliable deployment algorithm for Service Function Chains (SFC) based on deep reinforcement learning. SFC, as a chained function composition for complex network services, plays a crucial role in improving network efficiency and stability. To address the issue of existing SFC deployment algorithms that overlook the reliability of network functions and links, this paper proposes a deep reinforcement learning-based algorithm that utilizes a virtual network function and virtual link reliability mapping model for optimization. By learning the mapping between system states and actions, the algorithm can optimize the deployment strategy of SFC, thereby enhancing its reliability and performance. Experimental results demonstrate that the proposed algorithm can significantly improve the reliability of SFC and have practical implications for network service deployment.

SD-SRF: An Intelligent Service Deployment Scheme for Serverless-operated Cloud-Edge Computing in 6G Networks

With the development of serverless computing, developers can implement and deploy business applications as a combination of stateless functions. Although originally proposed for cloud computing, serverless computing is gradually applied to cloud-edge systems for service deployment to provide users with high-quality, low-latency services. However, optimized service deployment in 6G networks is a very challenging issue because of the vast number of deployable devices in the network, and its permutations are highly exponential. In this paper, we propose optimized service deployment schemes for online and offline, respectively, to minimize the overall latency at a lower cost. (1) First, a SD-SRF algorithm based on the greedy algorithm is proposed to optimize the service deployment for a multi-layer edge network, which consists of two phases: SR and SF. (a) Services are deployed in the nearest ancestor devices in the routing tree of all such service requests with the least cost of deployment. (b) When the deployment cost of moving some service replicas to devices in the lower layer is less than the benefit, the service will fall. (2) However, the offline algorithm relies on the availability of prior information heavily, such as request arrival pattern and number, which is difficult to obtain. Therefore, this paper proposes a PPO-MSD algorithm for optimal deployment online, where a Markov decision process (MDP) is modeled. Extensive simulation results show that PPO-MSD outperforms existing algorithms in terms of overall delay and utility, and its performance is close to the optimal ones obtained by SD-SRF, with the SD-SRF and PPO-MSD algorithms reducing the delay on average by 32.40% and 9.91%.

Look-Ahead VNF-FG Embedding Framework for Latency-Sensitive Network Services

Dynamic and zero-touch management is expected to be the key feature of next-generation 6G networks. Network Function Virtualization (NFV) is one of the key technologies for realizing such management through software-based networks. Despite great benefits offered by NFV, deploying network services (NSs) in NFV ecosystems remains a challenge, especially for latency-sensitive NSs, as they demand stringent latency requirements and fast service provisioning. Specifically, service graphs should be embedded into an infrastructure such that these requirements are satisfied while optimizing network operator’s objectives. To cope with the scalability of optimization-based approaches, heuristic methods are known as promising alternatives to find a satisfactory solution within an acceptable execution time. However, existing VNF embedding heuristics still suffer from the so-called causality issue, which may degrade the embedding solution quality. The causality issue means that embedding decisions cannot be optimally determined before all neighboring dependencies are known. To this end, we introduce our h-horizon sequential look-ahead greedy embedding framework, which provides efficient embedding and re-embedding strategies to alleviate the impact of the causality issue. The simulation results indicate that our proposed algorithm significantly improves embedding cost, compared to the existing heuristic algorithms while being much more scalable than an optimization-based approach.

A Link-Layer Virtual Networking Solution for Cloud-Native Network Function Virtualisation Ecosystems: L2S-M

Microservices have become promising candidates for the deployment of network and vertical functions in the fifth generation of mobile networks. However, microservice platforms like Kubernetes use a flat networking approach towards the connectivity of virtualised workloads, which prevents the deployment of network functions on isolated network segments (for example, the components of an IP Telephony system or a content distribution network). This paper presents L2S-M, a solution that enables the connectivity of Kubernetes microservices over isolated link-layer virtual networks, regardless of the compute nodes where workloads are actually deployed. L2S-M uses software-defined networking (SDN) to fulfil this purpose. Furthermore, the L2S-M design is flexible to support the connectivity of Kubernetes workloads across different Kubernetes clusters. We validate the functional behaviour of our solution in a moderately complex Smart Campus scenario, where L2S-M is used to deploy a content distribution network, showing its potential for the deployment of network services in distributed and heterogeneous environments.

Network Function Virtualization in Elastic Optical Networks

In Network Function Virtualization (NFV), traditional purpose-built proprietary middle-boxes are replaced with virtual network function (VNF) instances that are deployed over generic hardware; network traffic flows are steered through a sequence of VNF instances (i.e., that forms a service function chain) after resolving the VNF Forward Graph Embedding (FGE) problem. NFV eases the network management and fosters the agile deployment of network services, thus enabling a foremost solution to address issues of traditional enterprise networks. With the everlasting Internet traffic growth, in this work, we extend the provision of NFV upon Elastic Optical Networks (EONs) that deliver abundant elastic and fine-granular bandwidth for connections between VNFs. We systematically examine the integrated NFV-EON architecture, and investigate one key enabler, namely the FGE-EON problem, where one has to coordinate the allocation of computing resources (as in a typical FGE problem) and the assignment of optical bandwidth (i.e., spectrum allocation in EONs). The contributions of this work are multi-fold. The FGE-EON problem is proven to be NP-Complete for the first time. An Integer Linear Programming (ILP) model is developed to address FGE-EON problem of small instances. Despite its complexity, with novel modeling techniques, the proposed ILP model maintains a compact structure even after accounting for VNF sharing and VNF co-location together for the first time. For large scale problems, we design a heuristic algorithm based on layered graph that is shown to be near-optimal in our evaluations.

On the Game-Theoretic Analysis of Dynamic VNF Service Chaining in Edge-Cloud EONs

Network function virtualization (NFV) enables flexible, cost-effective, and timely deployment of new network services in elastic optical networks (EONs). Nowadays, the integration of cloud and edge computing becomes a prevailing trend, which has promoted the idea of edge-cloud EON (EC-EON) ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i.e.</i> , a multi-domain EON that consists of a cloud domain and one or more edge domains). In this work, we study the dynamic provisioning of virtual network function based service chain (VNF-SC) requests in EC-EONs, and leverage game theory to explain why and how the cloud and edge domain managers (DMs) should collaborate when protecting their own interests. We first formulate the non-cooperative interactions between cloud and edge DMs as a two-stage Stackelberg game, prove the existence of Nash Equilibrium in the game, and propose an algorithm based on backward induction for the non-cooperative service provisioning. Next, we address the cooperative provisioning scheme where the DMs can reach an agreement for mutual benefit, and model it with Nash bargaining. Finally, we conduct extensive simulations to compare the non-cooperative and cooperative provisioning schemes with the traditional centralized approach in EC-EONs in-depth. Simulation results confirm the effectiveness of our non-cooperative and cooperative provisioning schemes on protecting the interests of DMs, and suggest that the cooperative provisioning scheme can outperform the traditional centralized approach in terms of blocking probability, when inter- and intra-domain VNF-SC requests both exist in an EC-EON.

Service Deployment Model Based on Virtual Network Function Resizing

Network function virtualization enables to deploy network services more flexibly with virtual network functions (VNFs). A service provider needs to allocate the required VNFs to hosts, with satisfying different requirements on service delay. It is challenging for service providers to deploy services as efficiently as possible. The previous work has addressed this challenge by studying that multiple services share a VNF instance whose computing capacity is fixed; different VNF instances do not share their computing resources. More efficiency is expected with sharing resources among different VNF instances. This paper proposes a service deployment model for network function virtualization to minimize the deployment cost with capacity sharing and priority queuing. In the proposed model, VNF instances on the same host share the computing capacity in the host by VNF resizing during runtime. In addition, the priority queuing system is applied to each host. We formulate the proposed model as an optimization problem to minimize the service deployment cost. We develop a solution strategy named FlexSize to solve it heuristically in practical time. We evaluate the proposed model with a baseline which does not share the computing capacity among VNF instances in the host. The numerical results show that the proposed model reduces the service deployment cost compared with the baseline.

A SURVEY ON MOBILITY-AWARE AND DELAY-SENSITIVE SERVICE PROVISIONING IN MOBILE EDGE-CLOUD NETWORKS

Mobile edge computing (MEC) has emerged as a viable technique for pushing the cloud frontier to the network edge, enabling network services to be provisioned in close proximity to mobile consumers. Serving customers at the edge of the cloud can minimize service latency, lower operational costs, and increase network resource availability. Along with MEC, network function virtualization (NFV) is a viable strategy for implementing various network service functions as software in cloudlets (servers or clusters of servers). Giving mobile consumers virtualized network services can improve their service experience, simplify network service deployment, and make network resource management easier. However, mobile users roam arbitrarily inside networks, and different users often request various services with varying resource needs and latency requirements. As a result, it is a significant issue to provide dependable and smooth virtualized network services for mobile users in a MEC network while satisfying their specific latency needs, subject to network resource limits. This research focuses on the supply of virtualized network function services for mobile users in MEC while taking user mobility and service latency requirements into mind. We begin by posing two novel optimization issues for user service request admissions, with the goals of maximizing accumulative network utility and accumulative network throughput over a certain time horizon. Then, for the utility maximization issue, we design a constant approximation approach. We also create an online method for the problem of accumulative throughput maximization. Finally, we use experimental simulations to assess the performance of the suggested methods. The results of the experiments show that the proposed algorithms are promising.

Intent-driven autonomous network and service management in future cellular networks: A structured literature review

Intent-driven networks are an essential stepping stone in the evolution of network and service management towards a truly autonomous paradigm. User centric intents provide an abstracted means of impacting the design, provisioning, deployment and assurance of network infrastructure and services with the help of service level agreements and minimum network capability exposure. The concept of Intent Based Networking (IBN) poses several challenges in terms of the contextual definition of intents, role of different stakeholders, and a generalized architecture. In this review, we provide a comprehensive analysis of the state-of-the-art in IBN including the intent description models, intent lifecycle management, significance of IBN and a generalized architectural framework along with challenges and prospects for IBN in future cellular networks. An analytical study is performed on the data collected from relevant studies primarily focusing on the inter-working of IBN with softwarized networking based on NFV/SDN infrastructures. Critical functions required in the IBN management and service model design are explored with different abstract modeling techniques and a converged architectural framework is proposed. The key findings include: (1) benefits and role of IBN in autonomous networking, (2) improvements needed to integrate intents as fundamental policies for service modeling and network management, (3) need for appropriate representation models for intents in domain agnostic abstract manner, and (4) need to include learning as a fundamental function in autonomous networks. These observations provide the basis for in-depth investigation and standardization efforts for IBN as a fundamental network management paradigm in beyond 5G cellular networks.

A Proposal for Formulating a Spectrum Usage Fee for 5G Private Networks in Indonesian Industrial Areas

The Indonesian spectrum usage fees—the so-called Biaya Hak Pengguna Frekuensi Izin Pita Frekuensi Radio (BHP IPFR)—are currently calculated using a formula determined by the three following main parameters: the frequency band, the country’s economic parameter, and the nationwide population. As spectrum usage fees are proportional to the width of the bandwidth, the current formula would result in an extremely high price when applied to 5G-mmWave private networks, with the cost burden being a direct consequence for the service operator. In this paper, we propose the formulation of a new spectrum usage fee for 5G-mmWave private network implementation in Indonesian industrial areas. To do so, we evaluate the current formula, adopt the framework offered by the ITU-R SM.2012-5 (06/2016), and use an industrial reference index—the Indonesia Industry Readiness Index 4.0 (INDI 4.0) score. We test the proposal by applying the new formula to calculate the 5G-mmWave private network spectrum usage fee for the Jakarta industrial area. The result shows that the new formula gives a lower spectrum usage fee than the current formula, which benefits 5G-mmWave private network service operators. Such savings can be regarded as a government subsidy for the service operators to use in various ways in the industry, providing further economic benefits. Using the input–output model, we prove that despite the proposed new formula brings a lower spectrum usage fee, resulting in a loss in state income, it will lead to a much greater positive impact on the national economic output. Applying the new formula will eventually have a multiplier effect on various sectors and encourage digital economic growth and national digital transformation, especially for vertical industries in Indonesia. This study may serve as a guideline or initial reference for Indonesian policymakers and service operators for applying the CAPEX and OPEX cost of using the new spectrum for 5G-mmWave private network service implementation and estimating the economic multiplier for 5G-mmWave private network service deployment in industrial areas. It can also be used as a benchmark case for other countries to apply spectrum usage fees for private networks in industrial areas.

GNN-Based Hierarchical Deep Reinforcement Learning for NFV-Oriented Online Resource Orchestration in Elastic Optical DCIs

Network function virtualization (NFV) in elastic optical datacenter interconnections (EO-DCIs) enables flexible and timely deployment of network services. However, as the service provisioning of virtual network function service chains (vNF-SCs) in an EO-DCI needs to orchestrate the allocations of IT resources in datacenters (DCs) and spectrum resources on fiber links dynamically, it is a complex and challenging problem. In this work, we model the problem as a Markov decision process (MDP), and propose a hierarchical deep reinforcement learning (DRL) model based on graph neural network (GNN), namely, HRLOrch, to tackle it. To ensure its universality and scalability, we design the policy neural network (NN) in HRLOrch based on a GNN. As the GNN-based policy NN can operate on the graph-structured network state of an EO-DCI directly, it can adapt to an arbitrary EO-DCI topology without any structural changes. Then, through analysis, we find that the EO-DCI is a sparse reward environment if we want to train a DRL model to minimize the blocking probability of vNF-SCs in it directly. To address this issue, we design a hierarchical DRL with lower-level and upper-level models to improve the convergence performance of training. Specifically, we make the lower-level DRL optimize the provisioning scheme of each vNF-SC to minimize its resource usage, while the upper-level one coordinates the provisioning of all the active vNF-SCs to minimize the overall blocking probability. Hence, the lower-level and upper-level DRL models operate cooperatively in the training to optimize the dynamic provisioning of vNF-SCs. Our simulations demonstrate the universality and scalability of HRLOrch, and confirm that it can outperform the existing algorithms for vNF-SC provisioning in an EO-DCI.

PPTMon: Real-Time and Fine-Grained Packet Processing Time Monitoring in Virtual Network Functions

By softwarizing the legacy network functions, Network Function Virtualization (NFV) allows rapid development and deployment of network services as well as simplicity and flexibility in network operations and management. Monitoring the performance characteristics of Virtual Network Functions (VNFs), particularly packet processing time, is important to ensure that VNFs are operating correctly with desired performance. This is especially crucial for low-latency network services. In this paper, we present Packet Processing Time Monitoring (PPTMon), a real-time, fine-grained, and end-to-end solution for VNF packet processing time monitoring. PPTMon can provide per-hop monitoring for a single VNF as well as end-to-end monitoring for multiple VNFs in a service function chain. PPTMon allows monitoring in both sampling and continuous fashions. Continuously monitoring every packet may greatly degrade the performance of the VNFs and generate a huge amount of monitoring data. PPTMon’s event-filtering algorithm effectively filters out non-important data and reduces the performance overhead. PPTMon processes packets in-stack by embedding timestamp information directly into the packets, thus further reducing the effect on the VNF performance. PPTMon is implemented on top of extended Berkeley Packet Filter (eBPF) – a Linux framework that allows high-speed packet processing. Our experiment results shows that PPTMon can monitor VNF packet processing time with high accuracy and low impact on performance.

Energy-delay tradeoff for virtual machine placement in virtualized multi-access edge computing: a two-sided matching approach

By decoupling network functions from the underlying physical machines (PMs) at the edge of the networks, the virtualized multi-access edge computing (MEC) enables deployment of new network services and elastic network scaling to reduce the maintenance costs in a more flexible, scalable and cost-effective manner. Although there are appealing performance gains to be achieved, the placement of virtual machines (VMs) on top of the sharing PMs to support computation-intensive applications for smart mobile devices becomes a major challenge, especially for an increasing network scale. In this paper, we attempt to deal with the VM placement problem in virtualized MEC system, which is targeted for finding a performance balance between energy consumption and computing/offloading delay. To capture such a tradeoff for VM placement, we formulate a weighted sum based cost minimization problem as a pure 0–1 integer linear programming problem, which is NP-complete and very difficult to solve with lower complexity. Based on the one-to-one mapping relation constraint, the VM placement problem is then converted into a many-to-many two-sided matching problem between the VM instances and the PMs. Motivated by the student project allocation problem, we develop an extended two-sided matching algorithm with lower computational complexity for solving the matching problem. Simulation results are presented to show the effectiveness of our proposed algorithm, and the normalization factor is of great significance to obtain the lower total cost.

Impact of Processing-Resource Sharing on the Placement of Chained Virtual Network Functions

Network Function Virtualization (NFV) provides higher flexibility for network operators and reduces the complexity in network service deployment. Using NFV, Virtual Network Functions (VNF) can be located in various network nodes and chained together in a Service Function Chain (SFC) to provide a specific service. Consolidating multiple VNFs in a smaller number of locations would allow decreasing capital expenditures. However, excessive consolidation of VNFs might cause additional latency penalties due to processing-resource sharing, and this is undesirable, as SFCs are bounded by service-specific latency requirements. In this paper, we identify two different types of penalties (referred as "costs") related to the processingresource sharing among multiple VNFs: the context switching costs and the upscaling costs. Context switching costs arise when multiple CPU processes (e.g., supporting different VNFs) share the same CPU and thus repeated loading/saving of their context is required. Upscaling costs are incurred by VNFs requiring multi-core implementations, since they suffer a penalty due to the load-balancing needs among CPU cores. These costs affect how the chained VNFs are placed in the network to meet the performance requirement of the SFCs. We evaluate their impact while considering SFCs with different bandwidth and latency requirements in a scenario of VNF consolidation.

Reliability-Aware VNF Placement Using a Probability-Based Approach

Network function virtualization (NFV) is a new network architecture concept that simplifies the deployment of network services and improves service management. However, it is challenging for a network service provider (NSP) to decide where to place virtual network functions (VNFs). Most previous studies have considered only single-chain services, wherein the VNFs for a request are executed in sequence. In contrast to previous approaches, we consider more general and practical situations in which the VNFs of a request can be executed in parallel and are represented as a forwarding graph. Our objective is to maximize the profits earned by providing network services while satisfying the delay requirements of requests. We formulate the VNF placement problem as an integer linear programming (ILP) problem. Due to the complexity of this problem, we propose a probability-based approach called PBP, in which the placements of the VNFs are determined based on their probabilities of contributing to the profit. Furthermore, we propose a heuristic reliability-aware algorithm to guarantee service reliability, in which each VNF of a request is assigned a backup that can be shared with other requests. Simulation experiments show that PBP achieves a much shorter computation time than previous algorithms while earning higher profit, and furthermore, our reliability-aware algorithm provides the same reliability as a previous algorithm while yielding much higher profit.