Showing papers on "Differentiated service published in 2021"

Time-preference-based on-spot bundled cloud-service provisioning

Abstract: The cloud computing spot instance is one offering that vendors are leveraging to provide differentiated service to an expanding pay-per-use computing market. Spot instances have cost advantages, albeit at a trade-off of interruptions that can occur when the user's bid price falls below the spot price. The interruptions are often exacerbated since customers often require resources in bundles. For these reasons, customers might have to wait for a long time before their jobs are completed. In this paper, we propose a behavioral-economic model in the form of time-preference-based bids, wherein users are willing to use and bid for services at other times if the vendor cannot provide the resources at the preferred time. Given such bids, we consider the problem of provisioning for such service requests. We develop a time-preference-based optimization model. Since the optimization model is NP-Hard, we develop rule-based genetic algorithms. We have obtained very encouraging results with respect to standard commercial solver as a benchmark. In turn, our results provide evidence for the viability of our approach for online service-provisioning problems.

Research on User Access Selection Mechanism Based on Maximum Throughput for 5G Network Slicing

Abstract: With the development of Internet of Things (IoT) and network technologies, traditional networks cannot cope with the growth of network traffic and the changes in service requirements. The 5-th Generation Mobile Communication (5G) technology improves network transmission performance. In the communication network, 5G combines Software Defined Network (SDN) and Network Function Virtualization (NFV), through the deployment of end-to-end network slicing, to meet the challenge of differentiated service requirements in the complex environment. In the mobile network, users need to choose appropriate slices for access. Its performance is related to the quality of service and determines the efficiency of system resources utilization. We research the problem of slice re-access and slice resource scheduling caused by user mobility in 5G network slicing architecture and propose a slice access mechanism based on maximum throughput. A slice access selection algorithm based on genetic algorithm (GA) is proposed. Related simulations and comparative tests are carried out to prove the effectiveness and superiority of the algorithm.

A Multi-Tier Inspection Queueing System With Finite Capacity for Differentiated Border Control Measures

Abstract: Both uniform quarantine and isolation measures, due to the COVID-19 pandemic, have brought forth unprecedented and severe socio-economic impacts. For the global post-COVID economic recovery, it is of great significance to explore scientific ways to reopen the borders with consideration of both risk and efficiency. With the development of international travel health certificate or digital travel pass, differentiated inspection and quarantine measures can be implemented to accelerate the recovery of international travel. In this paper, we study a multi-tier inspection queueing system with finite capacity based on a differentiated level of risk classification. A queueing analysis is conducted for the stochastic process of inspecting cross-border travelers under differentiated service for inspection and quarantine. Besides, we develop a computing method to determine the steady-state probability and several performance indices of the proposed queueing system, and an illustrative example is also set to introduce a step-by-step process for the method. Furthermore, we figure out the relationship between the model parameters and system performance of interest by means of a series of numerical experiments. In the data analysis, we also illustrate the monotonic and concave effects on the system performance, which can provide a visualized understanding of the trade-off between safety and efficiency in the studied multi-server queueing system with hierarchical inspection channels and finite capacity. Our findings can reveal some managerial insight into the border control problems, which could reconcile the efficiency with safety in the current epidemic prevention and control tasks.

Research on QoS Classification of Network Encrypted Traffic Behavior Based on Machine Learning

Abstract: In recent years, privacy awareness is concerned due to many Internet services have chosen to use encrypted agreements. In order to improve the quality of service (QoS), the network encrypted traffic behaviors are classified based on machine learning discussed in this paper. However, the traditional traffic classification methods, such as IP/ASN (Autonomous System Number) analysis, Port-based and deep packet inspection, etc., can classify traffic behavior, but cannot effectively handle encrypted traffic. Thus, this paper proposed a hybrid traffic classification (HTC) method based on machine learning and combined with IP/ASN analysis with deep packet inspection. Moreover, the majority voting method was also used to quickly classify different QoS traffic accurately. Experimental results show that the proposed HTC method can effectively classify different encrypted traffic. The classification accuracy can be further improved by 10% with majority voting as K = 13. Especially when the networking data are using the same protocol, the proposed HTC can effectively classify the traffic data with different behaviors with the differentiated services code point (DSCP) mark.

Patient-Centered, Sustainable Hypertension Care: The Case for Adopting a Differentiated Service Delivery Model for Hypertension Services in Low- and Middle-Income Countries.

Abstract: Expanding hypertension services in low- and middle-income countries requires efficient and effective service delivery approaches that meet the needs and expectations of people living with hypertension within the resource constraints of existing national health systems. Ideally, a hypertension program will extend treatment coverage while maintaining service quality, maximizing efficient resource utilization and improving clinical outcomes. In this article, we discuss lessons learned from HIV differentiated service delivery initiatives, and make the case that the same approach should be adopted for hypertension programs.

Multi-Dimensional Resource Allocation for Diverse Safety Message Transmissions in Vehicular Networks

Abstract: To enhance driving safety and road intelligence for connected vehicles, the transmission of safety messages is critical in vehicular networks. In this paper, we focus on urban vehicular networks with deployed roadside units, and both vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) connections can be leveraged for message transmissions. We consider three types of safety messages: periodic messages for vehicular status notification, event-driven messages for urgent situation notification, and messages to achieve collective perception. To support different safety-related services, we develop a multi-dimensional resource allocation scheme to jointly optimize the sensing resource allocation (i.e., selecting vehicles as perception data providers), the V2I/V2V transmission mode selection, and the corresponding communication resource allocation. As the decisions on sensing resource allocation and wireless resource allocation are coupled, an iterative algorithm is proposed to solve the joint optimization problem by taking the differentiated service priorities into consideration. Extensive simulation results are presented to validate the effectiveness of the proposed resource allocation scheme.

A Novel Traffic Shaping Algorithm for SDN-Sliced Networks using a New WFQ Technique

Abstract: Managing traditional networks comes with number of challenges due to their limitations, in particular, because there is no central control. Software-Defined Networking (SDN) is a relatively new idea in networking, which enables networks to be centrally controlled or programmed using software applications. Novel traffic shaping (TS) algorithms are proposed for the implementation of a Quality of Service (QoS) bandwidth management technique to optimise performance and solve network congestion problems. Specifically, two algorithms, namely “Packet tagging, Queueing and Forwarding to Queues” and “Allocating Bandwidth”, are proposed for implementing a Weighted Fair Queuing (WFQ) technique, as a new methodology in an SDN-sliced testbed to reduce congestion and facilitate a smooth traffic flow. This methodology aimed at improving QoS that does two things simultaneously, first, making traffic conform to an individual rate using WFQ to make the appropriate queue for each packet. Second, the methodology is combined with buffer management, which decides whether to put the packet into the queue according to the proposed algorithm defined for this purpose. In this way, the latency and congestion remain in check, thus meeting the requirements of real-time services. The Differentiated Service (DiffServ) protocol is used to define classes in order to make network traffic patterns more sensitive to the video, audio and data traffic classes, by specifying precedence for each traffic type. SDN networks are controlled by floodlight controller(s) and FlowVisor, the slicing controller, which characterise the behaviour of such networks. Then, the network topology is modelled and simulated via the Mininet Testbed emulator platform. To achieve the highest level of accuracy, The SPSS statistical package Analysis of Variance (ANOVA) is used to analyse particular traffic measures, namely throughput, delay and jitter as separate performance indices, all of which contribute to QoS. The results show that the TS algorithms do, indeed, permit more advanced allocation of bandwidth, and that they reduce critical delays compared to the standard FIFO queueing in SDN.

A Price-Aware Congestion Control Protocol for Cloud Services

Abstract: In current infrastructure-as-a service (IaaS) cloud services, customers are charged for the usage of computing/storage resources only, but not the network resource. The difficulty lies in the fact that it is nontrivial to allocate network resource to individual customers effectively, especially for short-lived flows, in terms of both performance and cost, due to highly dynamic environments by flows generated by all customers. To tackle this challenge, in this paper, we propose an end-to-end Price-Aware Congestion Control Protocol (PACCP) for cloud services. PACCP is a network utility maximization (NUM) based optimal congestion control protocol. It supports three different classes of services (CoSes), i.e., best effort service (BE), differentiated service (DS), and minimum rate guaranteed (MRG) service. In PACCP, the desired CoS or rate allocation for a given flow is enabled by properly setting a pair of control parameters, i.e., a minimum guaranteed rate and a utility weight, which in turn, determines the price paid by the user of the flow. Two pricing models, i.e., a coarse-grained VM-Based Pricing model (VBP) and a fine-grained Flow-Based Pricing model (FBP), are proposed. The optimality of PACCP is verified by both large scale simulation and small testbed implementation. The price-performance consistency of PACCP are evaluated using real datacenter workloads. The results demonstrate that PACCP provides minimum rate guarantee, high bandwidth utilization and fair rate allocation, commensurate with the pricing models.

Deep Reinforcement Learning-Based Resource Allocation for Smart Grid in RAN Network Slice

Abstract: Driven by the construction of the Ubiquitous Electricity Internet of things, various services have increasingly higher requirements for wireless communication indicators. The 5G network with low latency, large connection and large bandwidth is urgently needed to fundamentally meet various business requirements and network security requirements of the smart grid. Because of the type of service arrival of smart grid is unknown and lacks prior knowledge, reinforcement learning (RL) is used to conduct this research. Considering the differentiated service characteristics of the smart grid and the challenges of flexibility and adaptability of the communication platform, this paper aims to solve the resource allocation problem of the radio access network (RAN) slice of smart grid. In this paper, we firstly introduce three typical power services and propose a service priority concept. After reviewing the fundamental concepts and proving the convergence of RL algorithm, we propose a dynamic resource allocation strategy of the RAN slice for smart grid based on RL. Finally, simulation results prove that the proposed RL algorithm can achieve resource utilization and quality of experience (QoE) improvement against the fair allocation scheme.

A novel dynamic bandwidth allocation algorithm in optical access systems

Abstract: With the emergence of differentiated service applications, optical access networks have developed as the most promising reliable technology for broadband access systems. One of the most critical characteristics of the access system is dynamic bandwidth allocation (DBA) which facilitates sharing of a common upstream channel for diverse categories of bursty data. It is expected that an ideal DBA mechanism shall reserve transmission resources for higher prioritized real-time applications along with a fair allocation of best effort (BE) services. This requires a provision of prioritized service class transmission on multiple uplink wavelengths in a cost-effective manner. This paper proposes a novel improved hybrid slot size/rate (IHSSR) DBA scheme that allows data traffic with three diverse priority classes to be communicated on multiple upstream wavelengths. The transmission cycle is segmented equally into two segments, in which the first section is reserved exclusively for the highest priority-expedite forwarding data traffic. On the other hand, the remaining part of the time cycle is further distributed between two services namely assured forwarding and best effort (BE) traffic. Furthermore, the proposed IHSSR algorithm computes the minimum scheduled channel length to achieve higher channel utilization. The proposed DBA mechanism is comprehensively evaluated in terms of quality of service metrics like average frame delay, throughput, average jitter, and channel utilization on varying traffic load. The simulation results indicate an improved performance from 4 to more than 20% in comparison with similar existing schemes. The range of congestion-free traffic load is also prolonged in the proposed algorithm. Thus it can be visualized that the proposed mechanism outperforms other schemes and met its objective.

Deep reinforcement learning-based resource reservation method for Power Emergency Internet-of-things Slice

Abstract: Aiming at the ultra-low latency service demand of power emergency Internet of Things (PEIoT), a multi-slice network architecture for ultra-low delay transmission of emergency Internet of Things was designed, and a PEIoT slice resource reservation and multi-heterogeneous slice resource sharing framework was proposed. The proposed framework adopts the deep reinforcement learning method to realize the automatic prediction and allocation of real-time resource requirements among heterogeneous slices. Simulation results show that the method based on resource reservation enables PEIoT slice to explicitly retain resources and provides a better level of security isolation. Deep reinforcement learning can ensure the accurate and real-time update of resource reservation and effectively consider the resource utilization rate and the differentiated service quality requirements of slices. The comparison with two existing algorithms shows that Dueling DQN has better performance advantages.

Expansion of a National Differentiated Service Delivery Model Increases Access to Treatment for HIV and Other Chronic Conditions in South Africa

Abstract: Lingrui Liu Global Health Leadership Initiative, Yale School of Public Health Sarah Christie Global Health Leadership Initiative, Yale School of Public Health Maggie Munsamy National Department of Health, Pretoria, South Africa Phil Roberts Project Last Mile, Pretoria, South Africa Merlin Pillay Project Last Mile, Pretoria, South Africa Sheela Shenoi Department of Medicine, Yale School of Medicine Mayur Desai Department of Chronic Disease Epidemiology, Yale School of Public Health Erika Linnander (  erika.linnander@yale.edu ) Global Health Leadership Initiative, Yale School of Public Health

Pricing and Strategies in Queuing Perspective Based on Prospect Theory

Abstract: Due to customers’ heterogeneity, enterprises/service providers usually adopt a service classification for different customers. However, service classification will result in a redistribution of waiting time, reducing wait time for priority customers by increasing wait time for regular customers. In this way, customers will form a psychological utility by comparing the expected waiting time between different queues. In this paper, we study a traditional non-preemptive M/M/1 queuing system in which incorporate customer preferences (loss aversion and gain seeking) will generate a psychological utility, which leads to the switch of customers and further impacts the revenue. In our paper, we analyze the monopoly queuing system in which customers can’t go away freely, and study it from three perspectives: revenue, social welfare, and customer utility. Firstly, we find that from the perspective of revenue maximization, enterprises should choose visual queue for queue classification. Next, enterprises should adopt unobservable queues for service classifications from social welfare maximization. Then, from customer utility maximization, enterprises should cancel service classification and keep regular customers only. Our results not only reaffirm existing research on the benefits of offering differentiated service and pricing by the service providers but also challenge some commonly accepted practices.

Differentiated Service Support in Wireless Networks With Multibeam Antennas

Abstract: Multibeam antenna arrays (MBAAs) have the capability to improve the capacity of a wireless network by facilitating simultaneous transmissions to multiple users. However, in practical deployments of wireless personal or local area networks (WPANs/WLANs) where piconet coordinators or access points (PNCs/APs) are deployed with MBAAs, it is quite likely to observe non-uniform node densities in various regions. To optimally utilize MBAAs in such scenarios, concurrent transmission scheduling in WPANs/WLANs is formulated as a multi-objective optimization problem. Then, a practical heuristic transmission scheduler that aims to maximize the number of concurrent communications by dynamically configuring the directions of beams is proposed. In addition, a service tag based fair scheduler is also proposed to achieve weighted fairness in WPANs/WLANs with MBAAs. Our results show that the performance improvements provided by the proposed heuristic scheduler are higher for antennas with lower beamwidths, and as the non-uniformity in the network increases, the traffic supported in the network increases in the range 24–41%, compared to the existing methods. The proposed service tag based scheduler can further improve the network throughput and achieve better fairness at the cost of a few beam direction reconfigurations, as compared to the existing methods and our heuristic scheduler.

E-commerce warehousing: learning a storage policy.

Abstract: E-commerce with major online retailers is changing the way people consume. The goal of increasing delivery speed while remaining cost-effective poses significant new challenges for supply chains as they race to satisfy the growing and fast-changing demand. In this paper, we consider a warehouse with a Robotic Mobile Fulfillment System (RMFS), in which a fleet of robots stores and retrieves shelves of items and brings them to human pickers. To adapt to changing demand, uncertainty, and differentiated service (e.g., prime vs. regular), one can dynamically modify the storage allocation of a shelf. The objective is to define a dynamic storage policy to minimise the average cycle time used by the robots to fulfil requests. We propose formulating this system as a Partially Observable Markov Decision Process, and using a Deep Q-learning agent from Reinforcement Learning, to learn an efficient real-time storage policy that leverages repeated experiences and insightful forecasts using simulations. Additionally, we develop a rollout strategy to enhance our method by leveraging more information available at a given time step. Using simulations to compare our method to traditional storage rules used in the industry showed preliminary results up to 14\% better in terms of travelling times.

Proactive Dynamic Calendar Allocation Scheme for 5G/B5G Transport Network Slicing Based Flexible Ethernet

Abstract: This paper proposed a proactive calendar allocation scheme that enhances utilization for optical network resource and provides differentiated service for network slicing. The simulation shows utilization improvement from 72.8% to 81.2% under the evaluated scenario.