Quality Of Service Classes Research Articles

An effective deep-Q learning scheme for QoS improvement in physical layer of software-defined networks

The demands of diverse applications necessitate specific communication requirements, such as jitter, packet loss ratio (PLR), and latency, at the end-to-end (E2E) level within the physical layer of software-defined networking (SDN). In heterogeneous network environments, E2E routes traverse multiple domains with varying quality-of-service (QoS) classes of traffic, posing challenges in meeting unique E2E QoS criteria. Hence the performance on the E2E path is significant to meet the needs of applications. Hence, the issue of providing services on the E2E route is challenging due to several QoS classes in a domain on the source-to-destination route. This paper presents an SDN architecture leveraging Deep Q-learning (DQL) to ascertain the appropriate QoS class for E2E routes in software-defined heterogeneous networks, thereby reducing overall E2E delay. The proposed model with DQL is compared with benchmark reinforcement learning. The proposed framework is validated using various real Internet topologies such as Abilene, USNet, and OS3E, assessing not only E2E delay but also jitter, PLR, and throughput, showcasing its efficacy in optimizing network performance regarding these performance metrics.

Adaptive quality of service for packet loss reduction using OpenFlow meters.

Quality of Service (QoS) is a mechanism used in computer networks to prioritize, classify, and treat packets differently based on certain criteria. This helps the switching devices to schedule and reorder packets if there is congestion in the network. Edge routers experience high traffic congestion as a result of traffic aggregation from the internal network devices. A router can have multiple QoS classes configured, and each class could experience traffic at various rates. However, when a QoS class is underperforming or needs more bandwidth, some bandwidth can be borrowed or leased out to another QoS class to ensure the link is utilized to maximum capacity and the highest throughput is achieved. This article proposes a bandwidth allocation and distribution algorithm that purely uses the flow statistics from the OpenFlow switches to allocate bandwidth to different QoS classes optimally based on their current requirement. The algorithm does not guarantee in advance that the packet loss will be minimized but does guarantee the initial minimum bandwidth allocation. It adjusts the flows' rates with the aim to increase their current throughput. The algorithm uses the Software Defined Networking (SDN) controller's flow monitoring component to query the flow statistics from the switch to first approximate the traffic flow rate and then calculate the optimal bandwidth values to assign to each QoS class. The proposed algorithms will be applied to certain switches in the path with the assumption that all the switches are OpenFlow compatible. The algorithm's performance was compared with the Adaptive Quality of Service (AQoS) algorithm over various traffic scenarios. The results show that the proposed algorithm achieves an average of 9% performance gain compared to the AQoS algorithm.

Large-Scale Service Function Chaining Management and Orchestration in Smart City

In the core networking of smart cities, mobile network operators need solutions to reflect service function chaining (SFC) orchestration policies while ensuring efficient resource utilization and preserving quality of service (QoS) in large-scale networking congestion states. To offer this solution, we observe the standardized QoS class identifiers of smart city scenarios. Then, we reflect the service criticalities via cloning virtual network function (VNF) with reserved resources for ensuring effective scheduling of request queue management. We employ graph neural networks (GNN) with a message-passing mechanism to iteratively update hidden states of VNF nodes with the objectives of enhancing allocation of resource blocks, accurate detection of availability statuses, and duplication of heavily congested instances. The deployment properties of smart city use cases are presented along with their intelligent service functions, and we aim to activate a modular architecture with multi-purpose VNFs and chaining isolation for generalizing global instances. Experimental simulation is conducted to illustrate how the proposed scheme performs under different congestion levels of SFC request rates, while capturing the key performance metrics of average delay, acceptance ratios, and completion ratios.

OPTIMIZATION OF ROUTING USING TRAFFIC CLASSIFICATION IN SOFTWARE DEFINED NETWORKING

Efficient routing is an essential task for any network as it directly impacts the network’s performance. In this paper, we have used traffic classification techniques to optimize routing in Software Defined Network (SDN) and provided a cost aware routing framework. Instead of classifying traffic based on a particular parameter, we perform the traffic classification using a hybrid approach that is based on machine learning techniques. Our framework uses supervised machine learning techniques to classify network flow and detect elephant flows which are too heavy in size and require significant bandwidth. As the requirements of each elephant flow vary with the application, we further perform Quality of Service (QoS) based traffic classification, which classifies these elephant flow into QoS classes as per their QoS requirements. For this purpose, we have used semi-supervised machine learning algorithms. Further, we have proposed a routing algorithm that makes use of the Dijkstra algorithm to compute the best possible shortest path based on the QoS requirements of the elephant flow. The proposed method was implemented and tested on Mininet using a POX controller. The simulation results show that our framework successfully classifies elephant flows with an accuracy of 80% and also computes a low-cost path for each elephant flow based on the actual internet data set.

Design of dynamic resource allocation scheme for real-time traffic in the LTE network

The 5G wireless technology is recently standardized for meeting intense demand. The Long Term Evolution (LTE) technology provides an easy, time-saving, and low-cost method for deploying a 4G/5G network infrastructure. To support multimedia service and higher bandwidth data delivery, an LTE MAC layer has Quality of Service (QoS) support with several QoS class indicator (QCI) levels. Based on LTE current QCI priority and QoS requirements in UEs, the original Max-Rate scheduler, Proportionally Fair (PF) and Modified Largest Weighted Delay First (M-LWDF) algorithms could not achieve their goal owing to the UE’s dynamic physical capacity with a different Channel Quality Indicator (CQI) at run time. For better QoS service than LTE networks, per UE’s CQI state for each resource block (RB) must be considered simultaneously in LTE MAC layer resource allocation with cross-layer support. As downlink real estimated capacity is dynamic owing to a UE’s periodic CQI reporting, the CQI state in LTE scheduling must be considered. This study proposes a smart and flexible scheme for Enhanced Utilization Resource Allocation (EURA) including three novel mechanisms that can dynamically fit UEs’ CQI states. The simulation results in this study demonstrate that the proposed EURA scheme outperforms the contrast schemes (Max-Rate, EXP/PF and M-LWDF), can save more rare radio capacity, and improve the utilization of radio resource assignment.

Network traffic clustering with QoS-awareness

Network traffic classification is essential in supporting network measurement and management. Many existing traffic classification approaches provide application-level results regardless of the network quality of service (QoS) requirements. In practice, traffic flows from the same application may have irregular network behaviors that should be identified to various QoS classes for best network resource management. To address the issues, we propose to conduct traffic classification with two newly defined QoS-aware features, i.e., inter-APP similarity and intra-APP diversity. The inter-APP similarity represents the close QoS association between the traffic flows that originate from the different Internet applications. The intra-APP diversity describes the QoS variety of the traffic even among those originated from the same Internet application. The core of performing the QoS-aware feature extraction is a Long-Short Term Memory neural network based Autoencoder (LSTM-AE). The QoS-aware features extracted by the encoder part of the LSTM-AE are then clustered into the corresponding QoS classes. Real-life data from multiple applications are collected to evaluate the proposed QoS-aware network traffic classification approach. The evaluation results demonstrate the efficacy of the extracted QoS-aware features in supporting the traffic classification, which can further contribute to future network measurement and management.

DeepEdge: A New QoE-Based Resource Allocation Framework Using Deep Reinforcement Learning for Future Heterogeneous Edge-IoT Applications

Edge computing is emerging to empower the future of Internet of Things (IoT) applications. However, due to heterogeneity of applications, it is a significant challenge for the edge cloud to effectively allocate multidimensional limited resources (CPU, memory, storage, bandwidth, etc.) with constraints of applications’ Quality of Service (QoS) requirements. In this paper, we address the resource allocation problem in Edge-IoT systems through developing a novel framework named <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DeepEdge</i> that allocates resources to the heterogeneous IoT applications with the goal of maximizing users’ Quality of Experience (QoE). To achieve this goal, we develop a novel QoE model that considers aligning the heterogeneous requirements of IoT applications to the available edge resources. The alignment is achieved through selection of QoS requirement range that can be satisfied by the available resources. In addition, we propose a novel two-stage deep reinforcement learning (DRL) scheme that effectively allocates edge resources to serve the IoT applications and maximize the users’ QoE. Unlike the typical DRL, our scheme exploits deep neural networks (DNN) to improve actions’ exploration by using DNN to map the Edge-IoT state to joint resource allocation action that consists of resource allocation and QoS class. The joint action not only maximize users’ QoE and satisfies heterogeneous applications’ requirements but also align the QoS requirements to the available resources. In addition, we develop a Q-value approximation approach to tackle the large space problem of Edge-IoT. Further evaluation shows that <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">DeepEdge</i> brings considerable improvements in terms of QoE, latency and application tasks’ success ratio in comparison to the existing resource allocation schemes.

Unleashing the potential of QoS-aware pricing within licensed shared access framework

We present a techno-economic analysis of a cellular market that operates within the licensed shared access (LSA) framework, consisting of a mobile network operator (MNO) that leases spectrum to a number of programme making and special events (PMSE) users. The MNO offers two quality-of-service (QoS) classes (high and low), differentiating the price based on the QoS class. The key question that we address is whether and to which extent the MNO has incentive to adopt this form of QoS-aware pricing. After getting some promising insights from the analysis of two case studies, we propose a general methodology. The first step is to model the parameters that are controlled by each PMSE user: (i) the way to choose between the two QoS classes and (ii) the available budget per QoS class. The second step is to compute the maximum revenue of the MNO. Our analysis reveals that the MNO can always tune the prices so as to maximise its revenue for the scenario where all users belong to the high QoS class. This is a consistent result throughout our study that holds for any considered set of user-controlled parameters and of technical parameters. We conclude that the adoption of QoS-aware pricing in an LSA market generates a tussle between the MNO and the regulator. The MNO has incentive to support fewer users but with high QoS and charge them more, which is not aligned with the regulator’s goal for social welfare maximisation.

Downlink scheduling algorithm over LTE networks using genetic algorithm

ABSTRACT The Third Generation Partnership Project (3GPP) has introduced the Long-Term Evolution (LTE) to handles the increasing demand for the wireless network setup. In the LTE setup, radio resource management scheduling algorithms are employed to distribute radio resources to users that have diverse Quality of Service (QoS) requirements. QoS Class identifier radio resource allocation algorithm for LTE downlink uses QoS requirements and channel conditions to decide the transmission order of users. However, this resulted in poor network performance by deteriorating the network data throughput and an increase in average delay due to the lack of proffer use of genetic algorithms (GA). In this paper, a new scheduling algorithm is proposed to improve the poor performance degradation of the network by introducing the idea of GA before radio resources are allowed to users. Simulation is used to test the performance of the proposed algorithm and the results obtained show that it considerably beats the other algorithm according to the measured metrics. An improvement of 65.76% was observed in terms of spectral efficiency, an increase of throughput and the average delay was minimised.

Bidirectional resource scheduling algorithm for advanced long term evolution system

AbstractThe 5G long term evolution‐advanced (LTE‐A) system aims to offer unprecedent data rates, user experience, and minimal latency amongst the end users with the introduction of carrier aggregation (CA) technology. Radio resource management (RRM) with scheduler is responsible for satisfying the user quality of service (QoS) requirements through resource allocation. Due to the need for an uninterrupted bidirectional communication between the eNodeB and user equipment (UE) in LTE‐A, the joint scheduling algorithm is acknowledged as a central research prospect. In this article, a modified joint uplink/downlink (UL/DL) user carrier scheduling algorithm satisfying demand service requests UE, is proposed. The joint optimization of the subcarrier assignment in an orthogonal frequency‐division multiplexing‐based network is utilized to balance the resource coupling between the UL/DL directions. CA is used to determine the weight factors required for the bandwidth allocation within the mobile users based on the QoS class identifier. These weight factors are employed as probability functions for resource allocation. The proposed algorithm will reduce the computational complexity arised in the conventional joint UL/DL scheduling algorithm based on the suboptimal solutions via probability functions. The simulation results show that the packet loss rate and packet delay among the mobile users are substantially reduced for both real‐time and nonreal‐time services.

An Optimized User Prioritized Service Provisioning in LTE Network

An exponential increase in the number of multimedia users over LTE network necessitates user prioritization and differentiated services support. LTE standard has defined Quality of Service (QoS) class-based user priority rather than priority among users within a QoS class. However, when the cell load exceeds the system capacity, Quality of Experience (QoE) of all the users may deteriorate due to lack of radio resources allocated to them. Under these circumstances, some users of interest may be prioritized over other users in the cell during resource allocation to enhance their QoE. In this paper, Proportional Fair (PF) scheduling algorithm Based User Priority (PFBUP) mechanism is proposed to prioritize organizational users over other users. Performance evaluation of the proposed mechanism is carried out using QualNet 7.1 network simulator by varying priority coefficient for the organizational users.

Enhanced Rough K-Means Based Flow Aggregation for QoS Mapping in Heterogeneous Network Environments

Flow aggregation is capable of reducing the burden on core routing and provide the scalability of network. Among the state-of-the-art methods of QoS (Quality of Service) mapping, the existing schemes lack effective means to guarantee the QoS of Internet flows in variable network environment. To tackle this problem, in this paper, we design a dynamic flow aggregation method that uses an enhanced rough ${k}$ -means algorithm (ERKM) to properly aggregate network flows and perform flexible QoS mapping across different networks. More importantly, it is able to adjust the allocation of flows according to the degree of membership in ever-changing Internet environment, and control the distance threshold parameter th to make the algorithm more flexible. The experimental results suggest that the proposed method outperforms other methods in terms of QoS support in the high-load and ever-changing networks. Moreover, it is validated that the proposed method can ensure the consistency and flexibility of QoS class mapping.

An Effective Hierarchical Control Plane for Software-Defined Networks Leveraging TOPSIS for End-to-End QoS Class-Mapping

If an end-to-end (E2E) path includes multiple domains, we need inter-domain collaboration to ensure the E2E quality-of-service (QoS) for the applications. In heterogeneous networks, an E2E path may go through domains with several QoS classes in each domain. However, the prevalent legacy network architecture and the standard software-defined networking (SDN) model lack effective mechanisms for inter-domain collaboration and QoS class mapping. In this study, we propose a hierarchical SDN control plane approach to guarantee the E2E QoS among multiple domains with various QoS classes on the E2E path. We propose a controller module for selecting the most suitable QoS class for each domain in the E2E path based on multi-criteria decision-making by using the technique for order of preference by similarity to ideal solution (TOPSIS). We map the suitable service classes in the global controller (GC) for provisioning the E2E QoS according to the application service requests. First, we propose an SDN-based inter-domain communication scheme and the message processing algorithm for E2E service delivery when multiple QoS classes exist in each domain. Next, we formulate the problem of service class selection with TOPSIS, provide an E2E mapping scheme, and demonstrate it with an example. Finally, we compare the proposed approach with the existing schemes for E2E QoS class mapping in terms of E2E delay, jitter, packet loss rate (PLR), and cost (per bandwidth unit). According to our simulation results, the proposed approach ensures the E2E QoS and guarantees the E2E delay, jitter, PLR, and cost according to the application service requests.

Shared LTE Network Performance on Smart Grid and Typical Traffic Schemes

This paper investigates the possibility of delivering distinct smart grid (SG) demand response (DR) applications in a highly loaded LTE network. In a shared LTE network, the proportion of SG DR traffic is relatively low when compared to typical traffics such as voice over IP, Skype video call, FTP, Youtube video stream, and HTTP. The quality of service (QoS) requirements for the SG DR traffics have to be fulfilled by maintaining the network delays and the packet delivery ratios within certain limits, while not causing significant hindrance to the typical traffics. The Riverbed Modeler network simulations are performed using detailed physical layer propagation models, detailed LTE functionality, and a suburban topology. In the simulation scenarios, three distinct DR applications generate varying amounts of SG DR traffic to the LTE network while the LTE capacity is exceeded by the typical traffics. The results illustrate that satisfactory performance for the SG DR traffics can be maintained due to the constant traffic characteristics and relatively low traffic amount that facilitates the scheduling of channel resources. Typically, the more a DR application generates traffic the higher hindrance it causes for the typical traffics other than the voice over IP that applies the QoS class of highest priority.

A Design of Packet Scheduling Algorithm to Enhance QoS in High-Speed Downlink Packet Access (HSDPA) Core Network

Voice over Internet Protocol (VOIP) in an efficient manner is a basic requirement of modern era. The real time and non-real traffics demand customized communication provisioning to get guarantee of service. For this we proposed a user fulfillment design for facilitating packets switching in 3G cellular network to insure provisioning of QoS (quality of service) in DiffServ (Differentiated Services) Network. To enhance QoS for real time traffic by reducing delay, packet loss and jitter, we proposed Low latency queuing (LLQ) algorithm. In this paper, we focused on packet scheduling, Diffserv and QoS classes mapping into Universal Mobile telecommunication System (UMTS) classes and buffering. To associate different types of real time multimedia traffic, the QoS provisioning mechanism used different code points of Diffserv. The new idea in LLQ is to map the video and voice traffics against two separate queues and used priority queuing in Low latency queuing for voice traffic. The results got from reproductions shows that proposed calculation meets the QoS prerequisites.

A Delay Optimal Multiclass Packet Scheduler for General M2M Uplink

The uplink data arriving at the Machine-to-Machine (M2M) Application Server (AS) via M2M Aggregators (MAs) is fairly heterogeneous along several dimensions such as maximum permissible packet delay, arrival rate, and payload size, thus necessitating the design of Quality-of-Service (QoS) aware packet schedulers. In this paper, we classify the M2M uplink data into multiple QoS classes and use sigmoidal function to map the delay requirements of each class onto utility functions. We propose a delay-optimal multiclass packet scheduler at AS obtained by iteratively searching for the optimal fraction of time-sharing between all preemptive priority scheduling policies, such that it maximizes a proportionally fair system utility metric. The iterative optimization process leads to reduced complexity, thus facilitating its online implementation. We then extend this work to determine a jointly optimal packet scheduler at the MAs and AS. We significantly reduce its computational complexity by iteratively solving a distributed optimization problem, independently at MAs and AS. Using Monte Carlo simulations, we show that the proposed joint packet scheduler outperforms other state-of-the-art packet schedulers and results in (desirable) near-minimal delay-jitter for delay-sensitive traffic at the expense of (tolerable) higher delay-jitter for delay-tolerant traffic.

Quality of service class mapping and scheduling scheme for converged LTE-WiFi in the next generation networks

In this paper, a novel quality of service (QoS) class mapping and scheduling scheme for the LTE-WiFi converged network is proposed to ensure that end-to-end QoS support is provisioned seamlessly. A mapping among QoS classes of LTE and WiFi is performed first, then scheduling scheme is outlined. For scheduling scheme, traffic is separated into real-time (RT) and non-real-time (NRT) components, then a two-stage complex queuing strategy is performed. In the first stage, deficit weighted round robin queuing (DWRRQ) and class-based weighted fair queuing (CBWFQ) are used for RT and NRT applications respectively to segregate and forward traffic according to resource requirements. Whereas in the second stage, rate-control priority queuing (RCPQ) is used for all types of traffic to give each class appropriate priority level. Performance evaluation of the converged network was done based on different metrics such as jitter, end-to-end delay, throughput, and packet loss ratio. Simulation results show significant improvements for RT with slight degradation on NRT applications after using the new scheduling scheme.

End-user traffic policing for QoS assurance in polyservice RINA networks

Looking at the ever-increasing amount of heterogeneous distributed applications supported on current data transport networks, it seems evident that best-effort packet delivery falls short to supply their actual needs. Multiple approaches to Quality of Service (QoS) differentiation have been proposed over the years, but their usage has always been hindered by the rigidness of the TCP/IP-based Internet model, which does not even allow for applications to express their QoS needs to the underlying network. In this context, the Recursive InterNetwork Architecture (RINA) has appeared as a clean-slate network architecture aiming to replace the current Internet based on TCP/IP. RINA provides a well-defined QoS support across layers, with standard means for layers to inform of the different QoS guarantees that they can support. Besides, applications and other processes can express their flow requirements, including different QoS-related measures, like delay and jitter, drop probability or average traffic usage. Greedy end-users, however, tend to request the highest quality for their flows, forcing providers to apply intelligent data rate limitation procedures at the edge of their networks. In this work, we propose a new rate limiting policy that, instead of enforcing limits on a per QoS class basis, imposes limits on several independent QoS dimensions. This offers a flexible traffic control to RINA network providers, while enabling end-users freely managing their leased resources. The performance of the proposed policy is assessed in an experimental RINA network test-bed and its performance compared against other policies, either RINA-specific or adopted from TCP/IP. Results show that the proposed policy achieves an effective traffic control for high QoS traffic classes, while also letting lower QoS classes to take profit of the capacity initially reserved for the former ones when available.

Techno-Economics of Licensed Shared Access with Mobile Network Operators Leasing Spectrum to PMSE Users

Licensed shared access (LSA) has been adopted as a new paradigm by several different regulatory and industrial stakeholders. The European LSA two-tier approach opens the possibility to lease the spectrum more dynamically among network operators implementing different wireless technologies. The initial use cases considered mobile network operators (MNOs) leasing spectrum from other technologies. However, recent initiatives emphasized an important symmetric use case, where programme making and special events (PMSE) users could lease spectrum from MNOs short-term for concerts, conferences, etc. In this market, the MNOs should first have economic incentives to temporarily lease the spectrum that they do not fully use to PMSE users. The goal of this work is to address the relevance of economic aspects for spectrum sharing under LSA. Importantly, this is also of broader regulatory interest, since spectrum could be utilized more efficiently using LSA-like schemes only if this is techno-economically attractive in practice. To this end we present an extensive techno-economic study, analysing the interactions between the MNO and a number of PMSE users. For the MNO, the key challenge is to determine the pricing policy that maximizes its revenue for scenarios where users belong to two distinct quality-of-service (QoS) classes, suitable for different audio applications. Our contributions are as follows. From a technical perspective, we perform extensive simulations in ns-3 for several deployment scenarios, whereby we estimate the maximum number of supported PMSE users. We consider the effect of target QoS throughput, carrier frequency, base station (BS) transmit power, bandwidth, and propagation environment. From an economic perspective, we propose three QoS-aware pricing policies for the MNO that correspond to applications of second and third degree of price discrimination, compute the corresponding revenue, and compare them with the baseline policy of no price discrimination, as currently adopted by the UK regulator Ofcom. Our analysis highlights the importance of the purchasing power of the users, in particular the ones that belong to the low QoS class. Whenever all of them can afford a fixed price set by the regulator, similar to the one suggested by Ofcom, we expect that, in almost all technical configuration settings, the market will operate at a region where it is more profitable for the MNO to support only these low-QoS PMSE users. Therefore, there is no need for price discrimination. On the contrary, whenever the MNO needs to reduce the face value (i.e. the initial price) for some of them, QoS-aware pricing schemes can generate higher revenue. For these cases, price discrimination should be the de facto policy and scenarios with users with high QoS, as well as mixed QoS scenarios, can be supported. Our results show that there is a strong interplay between economic and technical conditions, which ought to be considered when designing regulatory measures for encouraging spectrum sharing in such markets.

Internet multimedia traffic classification from QoS perspective using semi-supervised dictionary learning models

To address the issue of finegrained classification of Internet multimedia traffic from a Quality of Service (QoS) perspective with a suitable granularity, this paper defines a new set of QoS classes and presents a modified K-Singular Value Decomposition (K-SVD) method for multimedia identification. After analyzing several instances of typical Internet multimedia traffic captured in a campus network, this paper defines a new set of QoS classes according to the difference in downstream/upstream rates and proposes a modified K-SVD method that can automatically search for underlying structural patterns in the QoS characteristic space. We define bag-QoS-words as the set of specific QoS local patterns, which can be expressed by core QoS characteristics. After the dictionary is constructed with an excess quantity of bag-QoS-words, Locality Constrained Feature Coding (LCFC) features of QoS classes are extracted. By associating a set of characteristics with a percentage of error, an objective function is formulated. In accordance with the modified K-SVD, Internet multimedia traffic can be classified into a corresponding QoS class with a linear Support Vector Machines (SVM) classifier. Our experimental results demonstrate the feasibility of the proposed classification method.