Fair Scheduling Research Articles

CDF-Based Multiuser Scheduling for Downlink Non-Orthogonal Multiple Access (NOMA)

In this paper, we propose cumulative distribution function (CDF)-based multiuser scheduling (CMS) for downlink non-orthogonal multiple access (NOMA) to accommodate a massive number of IoT devices over limited radio resources and statistically analyze the ergodic capacity of devices. With the proposed CMS algorithm, multiple devices having relatively better channel gain to exploit multiuser diversity are selected while meeting QoS requirements by controlling the amount of resources. We first provide the user selection probability in terms of the weight of each device, which determines the amount of time resources. Based on some selected results, it is shown that equal weights ensure fair resource allocation. Further, for fair resource allocation scenario, the distribution of signal-to-interference-plus-noise ratio (SINR) when a specific device is selected is derived. With this derived statistical result of SINR, we analyze the ergodic capacity of the specific device. Some selected results show that the proposed CMS algorithm allocates the time resource more fairly than proportional fair scheduling. Further, the average throughput and amount of resources can be controlled by adjusting the weights of each device, so different QoS requirements can be guaranteed.

Dynamic Resource Scheduling Optimization With Network Coding for Multi-User Services in the Internet of Vehicles

For Internet of Vehicles (IoV) systems with multiple users, network coding can be introduced to provide efficient error control and throughput improvement services. However, if the heterogeneity characteristics and requirements of the end users (vehicles) are neglected, it will be difficult for an IoV system to provide each end user with fair system services, without which the advantages of network coding cannot be fully achieved and the performance of the multi-user diversity system will be degraded. In this paper, we propose a Dynamic Resource Scheduling Optimization (DRSO) algorithm, a dynamic fair scheduling algorithm combined with network coding for system resource allocation in a multi-user IoV system. We construct a general solution framework for service scheduling: first, we estimate the fairness index for each end user (vehicle) with the key information on Quality of Service (QoS). Second, we construct a service scheduling control model based on the service capability of control entities (multi-access edge computing servers), and propose a new utility evaluation function. Third, based on the fairness index, we select end users into multiple network coding sets. Network coding sets are the basic units of service scheduling. The optimization objective of the scheduling service is to maximize the total utility of all the network coding sets (the utility of the control entity). Finally, we establish a coding cache queue in the control entity based on the scheduling decision. To obtain the global optimal solution for active queue control, we combine a Quantum Particle Swarm Optimization (QPSO) algorithm with a Proportional Integral (PI) model. Then, the optimal scheduling decision can be made. Extensive simulation results show that DRSO outperforms related scheduling algorithms in varying traffic loads, demonstrating that DRSO can effectively guide service resource allocation.

Considering fairness in the load shedding scheduling problem

Every day national power system networks provide thousands of MW of electric power from generating units to consumers, requiring different operations and planning to ensure secure systems. Where demand exceeds supply, load shedding - controlled, enforced reduction in supply - is necessary to prevent system collapse. Should load shedding need to be implemented, a planned schedule is necessary to allocate geographic areas on the required period of shedding. The problem of how to construct a schedule that fairly allocates load shedding responsibilities over geographic areas with minimum economic impacts is addressed in this paper. Two programming models are proposed. The first model consists of a linear integer programming model in which the objective is to minimise the economic cost subject to different fairness allocation constraints, while the second model involves formulation of the problem as a goal programming model in which different conflicting goals are simultaneously optimised. Several case studies are conducted in the context of a realistic, but hypothetical, scenario to explore the possible solutions that the proposed models provide. Results show that a fair schedule requires a high cost whereas lower cost can only be achieved with some sacrifices to the fairness of the schedule.

Deadline-Aware Fair Scheduling for Offloaded Tasks in Fog Computing With Inter-Fog Dependency

A fundamental problem in fog computing networks is how to schedule the deadline-aware offloaded tasks that directly arrive from the end-users and via other fog nodes. The computational resource allocation becomes more challenging when the tasks demand different delay-deadlines. In this letter, we aim to propose a scheduling strategy to maximize the number of the completed tasks within their respective deadlines while making the network strongly stable. We exploit Lyapunov drift-plus-penalty function on the queue length to schedule the tasks in the queues. Subsequently, the scheduling policy decides the amount of task to be offloaded to the underloaded fog nodes to fully utilize the computational resources offered by all fog nodes in the network. Our simulation results reveal that the proposed strategy outperforms the baseline schemes, especially when those tasks have distinct delay-deadlines.

Workload-Adaptive Configuration Tuning for Hierarchical Cloud Schedulers

Cluster schedulers provide flexible resource sharing mechanism for best-effort cloud jobs, which occupy a majority in modern datacenters. Properly tuning a scheduler's configurations is the key to these jobs’ performance because it decides how to allocate resources among them. Today's cloud scheduling systems usually rely on cluster operators to set the configuration and thus overlook the potential performance improvement through optimally configuring the scheduler according to the heterogeneous and dynamic cloud workloads. In this paper, we introduce AdaptiveConfig, a run-time configurator for cluster schedulers that automatically adapts to the changing workload and resource status in two steps. First, a comparison approach estimates jobs’ performances under different configurations and diverse scheduling scenarios. The key idea here is to transform a scheduler's resource allocation mechanism and their variable influence factors (configurations, scheduling constraints, available resources, and workload status) into business rules and facts in a rule engine, thereby reasoning about these correlated factors in job performance comparison. Second, a workload-adaptive optimizer transforms the cluster-level searching of huge configuration space into an equivalent dynamic programming problem that can be efficiently solved at scale. We implement AdaptiveConfig on the popular YARN Capacity and Fair schedulers and demonstrate its effectiveness using real-world Facebook and Google workloads, i.e., successfully finding best configurations for most of scheduling scenarios and considerably reducing latencies by a factor of two with low optimization time.

Enforcing Bitrate-Stability for Adaptive Streaming Traffic in Cellular Networks

Video streaming over cellular network has become extremely popular in 4G and will be an integral part of future cellular networks. While most modern-day video clients continually adapt quality of video streams, they neither coordinate with network elements nor among each other. Consequently, a streaming client may quickly overload the cellular network, leading to poor Quality of Experience (QoE) for users in the network. Motivated by this problem, we present D-VIEWS — a scheduling paradigm that assures video bitrate stability of adaptive video streams while ensuring better system utilization. D-VIEWS only needs to be aware of the set of video bitrates and requires no changes to streaming clients and other network functions. Through simulations, we also study the performance of proportional fairness scheduler and D-VIEWS in the presence of user arrival and departure events.

Buffer-Aided Adaptive Wireless Powered Communication Network With Finite Energy Storage and Data Buffer

In this paper, the access point (AP) in a wireless network is assumed to provide energy supply via wireless energy transfer to multiple terminals in the downlink, and all the terminals use the harvested energy to transmit their collected data to the AP in the uplink in a time division multiple access (TDMA) manner. Each terminal is provisioned with a finite energy storage and a finite data buffer to store the harvested energy and to buffer the arrived data traffic, respectively. Due to the limited data buffer and energy storage size, there might be data loss due to either data buffer overflow or energy storage depletion. Firstly, we aim at maximizing the long-term weighted sum-rate through energy beamforming vector design, power allocations, rate control, time allocations, and transmission mode selection subject to average transmit power, peak transmit power, data loss ratio requirements, practical data buffer as well as energy storage constraints. Secondly, the weighted max-min scheduling scheme is proposed to guarantee the fair access requirement by multiple terminals. Numerical analyses are presented to show that, the proposed adaptive design can substantially improve the average achievable rate region, while the proposed weighted max-min fair scheduling can effectively ensure the fair access requirements.

HybSMRP: a hybrid scheduling algorithm in Hadoop MapReduce framework

Due to the advent of new technologies, devices, and communication tools such as social networking sites, the amount of data produced by mankind is growing rapidly every year. Big data is a collection of large datasets that cannot be processed using traditional computing techniques. MapReduce has been introduced to solve large-data computational problems. It is specifically designed to run on commodity hardware, and it depends on dividing and conquering principles. Nowadays, the focus of researchers has shifted towards Hadoop MapReduce. One of the most outstanding characteristics of MapReduce is data locality-aware scheduling. Data locality-aware scheduler is a further efficient solution to optimize one or a set of performance metrics such as data locality, energy consumption and job completion time. Similar to all situations, time and scheduling are the most important aspects of the MapReduce framework. Therefore, many scheduling algorithms have been proposed in the past decades. The main ideas of these algorithms are increasing data locality rate and decreasing the response and completion time. In this paper, a new hybrid scheduling algorithm has been proposed, which uses dynamic priority and localization ID techniques and focuses on increasing data locality rate and decreasing completion time. The proposed algorithm was evaluated and compared with Hadoop default schedulers (FIFO, Fair), by running concurrent workloads consisting of Wordcount and Terasort benchmarks. The experimental results show that the proposed algorithm is faster than FIFO and Fair scheduling, achieves higher data locality rate and avoids wasting resources.

Performance Analysis of Integrated Satellite-Terrestrial Multiantenna Relay Networks With Multiuser Scheduling

In this article, we investigate the performance of a multiuser integrated satellite-terrestrial relay network (ISTRN) with the threshold-based decode-and-forward protocol, where the satellite-relay link undergoes shadowed-Rician (SR) fading, while the relay-user links experience Nakagami-m fading. We first formulate a constrained optimization problem with an objective to maximize the system capacity in order to determine the beamforming weight vectors at the relay and thereby develop two multiuser scheduling schemes, namely, best user scheduling and user fairness scheduling. Then, we present a closed-form expression for the probability density function (PDF) of the square sum of independent and identically distributed SR random variables, which is more accurate and concise than the existing solutions. Based on the new PDF results, we derive closed-form expressions for the outage probability (OP) and ergodic capacity of the considered network using the two scheduling schemes. Furthermore, the asymptotic OP expressions at high signal-to-noise ratio are also deduced to reveal the asymptotic behavior of the considered ISTRN. Finally, simulation results are provided to validate our theoretical analysis and show the impact of various parameters on the system performance.

Improving MapReduce scheduler for heterogeneous workloads in a heterogeneous environment

SummaryBig data is largely influencing business entities and research sectors to be more data‐driven. Hadoop MapReduce is one of the cost‐effective ways to process large scale datasets and offered as a service over the Internet. Even though cloud service providers promise an infinite amount of resources available on‐demand, it is inevitable that some of the hired virtual resources for MapReduce are left unutilized and makespan is limited due to various heterogeneities that exist while offering MapReduce as a service. As MapReduce v2 allows users to define the size of containers for the map and reduce tasks, jobs in a batch become heterogeneous and behave differently. Also, the different capacity of virtual machines in the MapReduce virtual cluster accommodate a varying number of map/reduce tasks. These factors highly affect resource utilization in the virtual cluster and the makespan for a batch of MapReduce jobs. Default MapReduce job schedulers do not consider these heterogeneities that exist in a cloud environment. Moreover, virtual machines in MapReduce virtual cluster process an equal number of blocks regardless of their capacity, which affects the makespan. Therefore, we devised a heuristic‐based MapReduce job scheduler that exploits virtual machine and MapReduce workload level heterogeneities to improve resource utilization and makespan. We proposed two methods to achieve this: (i) roulette wheel scheme based data block placement in heterogeneous virtual machines, and (ii) a constrained 2‐dimensional bin packing to place heterogeneous map/reduce tasks. We compared heuristic‐based MapReduce job scheduler against the classical fair scheduler in MapReduce v2. Experimental results showed that our proposed scheduler improved makespan and resource utilization by 45.6% and 47.9% over classical fair scheduler.

Datacenter congestion control

Recent years have seen a slew of papers on datacenter congestion control mechanisms. In this editorial, we ask whether the bulk of this research is needed for the common case where congestion control involves hosts responding to simple congestion signals from the network and the performance goal is reducing some average measure of flow completion time. We raise this question because we find that, out of all the possible variations one could make in congestion control algorithms, the most essential feature is the switch scheduling algorithm. More specifically, we find that congestion control mechanisms that use Shortest-Remaining-Processing-Time (SRPT) achieve superior performance as long as the rate-setting algorithm at the host is reasonable. We further find that while SRPT's performance is quite robust to host behaviors, the performance of schemes that use scheduling algorithms like FIFO or Fair Queuing depend far more crucially on the rate-setting algorithm, and their performance is typically worse than what can be achieved with SRPT. Given these findings, we then ask whether it is practical to realize SRPT in switches without requiring custom hardware. We observe that approximate and deployable SRPT (ADS) designs exist, which leverage the small number of priority queues supported in almost all commodity switches, and require only software changes in the host and the switches. Our evaluations with one very simple ADS design shows that it can achieve performance close to true SRPT and is significantly better than FIFO. Thus, the answer to our basic question - whether the bulk of recent research on datacenter congestion control algorithms is needed for the common case - is no.

Reflections on "analysis and simulation of a fair queueing algorithm"

This article discusses my personal view of the events that led to the publication of the paper "Analysis and Simulation of a Fair Queueing Algorithm" that won a Test-of-Time Award in 2007.

A design simulation of traffic light intersection using SimEvents MATLAB

The smoothness of traffic flow is one indicators of the excellent management by road managers, especially in the settings of traffic lights on road intersection. The time duration the red and green light must be set precisely according to the traffic density, so that the optimal and fairness queue length and waiting time are obtained for road users in each lane. Some problems at the intersection of the highway are the occurrence of very long queues in one lane, but quite short on the other lane. This happens because the setting is not appropriate on the traffic light time duration. In this article, we propose a model and simulation for intersection with traffic light using MATLAB SimEvents tools. The proposed model is to place an inhibitor module on each path at the intersection, where the entity does not flow for a certain duration of time. Through this simulation, it can be predicted the exact duration time of traffic lights based on the traffic density on each lane.

Implementation of Max-SNR Opportunistic Scheduling in Cross Layer Design

The spectrum efficiency in wireless networks is becoming increasingly important to satisfy the expanding need for remote services, particularly reasonable remote Internet services. Various wireless users will encounter distinct circumstances on the channel, which are time-varying and location dependent at a given time. To exploit the wireless time-varying nature, a cross-layer design method called Opportunistic scheduling is used. Opportunistic scheduling increases the overall system performance and user fairness requirements under certain Quality of Service (QoS). The main idea behind this opportunistic scheduling algorithm is to make use of the time-varying channel and a user with the highest channel condition should be scheduled at a specified moment. The progressions in Wireless innovation made opportunistic scheduling a famous research point as of late. The demand for QoS provisioning is increasing and using a scheme which allows only users with best channel conditions to transmit at high transmission power cannot be satisfied. The objective of this paper is to implement opportunistic scheduling while adhering to fairness and QoS constraints, using Max-Min fair algorithm. In brief, a wireless network has been simulated using Qualnet simulator. An opportunistic scheduler that uses Max-Min fairness scheduling has been implemented, at the Media Access Control (MAC) layer. Here, the base station gathers the Signal to Noise Ratio (SNR) of all the nodes and then schedules the users using Max-Min algorithm, based on these SNR values. The same scenario is then implemented using Strict Priority, which is a Non-opportunistic scheduling algorithm. The resulting throughput, fairness, delay and jitter of both the algorithms are then compared.

A Comparison of Reinforcement Learning Algorithms
 in Fairness-Oriented OFDMA Schedulers

Due to large-scale control problems in 5G access networks, the complexity of radioresource management is expected to increase significantly. Reinforcement learning is seen as apromising solution that can enable intelligent decision-making and reduce the complexity of differentoptimization problems for radio resource management. The packet scheduler is an importantentity of radio resource management that allocates users’ data packets in the frequency domainaccording to the implemented scheduling rule. In this context, by making use of reinforcementlearning, we could actually determine, in each state, the most suitable scheduling rule to be employedthat could improve the quality of service provisioning. In this paper, we propose a reinforcementlearning-based framework to solve scheduling problems with the main focus on meeting the userfairness requirements. This framework makes use of feed forward neural networks to map momentarystates to proper parameterization decisions for the proportional fair scheduler. The simulation resultsshow that our reinforcement learning framework outperforms the conventional adaptive schedulersoriented on fairness objective. Discussions are also raised to determine the best reinforcement learningalgorithm to be implemented in the proposed framework based on various scheduler settings.

5G 네트워크에서의 무선자원 할당: 신경망을 이용한 접근법

The minimum frequency-time unit that can be allocated to User Equipments (UEs) in the fifth generation (5G) cellular networks is a Resource Block (RB). A RB is a channel composed of a set of OFDM subcarriers for a given time slot duration. 5G New Radio (NR) allows for a large number of block shapes ranging from 15 kHz to 480 kHz. In this paper, we address the problem of RBs allocation to UEs. The RBs are allocated at the beginning of each time slot based on the channel state of each UE. The problem is formulated based on the Generalized Proportional Fair (GPF) scheduling. Then, we model the problem as a 2-Dimension Hopfield Neural Networks (2D-HNN). Finally, in an attempt to solve the problem, the energy function of 2D-HNN is investigated. Simulation results show the efficiency of the proposed approach.

Opportunistic schedulers and asymptotic price for fairness

We consider the well-known wireless fair opportunistic schedulers (mainly the alpha-fair schedulers) and analyze their price of fairness (PoF). Efficient scheduler, designed from the system perspective, maximizes the sum of accumulated utilities of all the agents, accumulated over several time slots. On the other hand, the fair schedulers deviate from such a schedule to provide a given level of fairness to various customers utilizing the system. This obviously results in a lower (total) accumulated utility. We study this loss, using the well-known performance measure, the price of fairness. Previous studies show that the PoF mostly increases, as the number of agents increases. We have very different results for opportunistic schedulers. We group agents into finite classes, each class having identical utilities and QoS requirements (inspired by wireless cellular networks), to obtain the asymptotic PoF (APoF). This is always below one. Further, in many cases, the PoF actually decreases to zero/negligible value as the number of agents increases. We also consider the case of multiple resources with bounded utilities. We again have zero/small APoF, depending upon the distributions of the utilities and the relative proportions of various classes. We derive closed-form/easily computable expressions for the APoF, using extreme-value theory, center-order statistics and the maximum theorem.

Semi-Clairvoyant Scheduling in Data Analytics Systems

Popular data analytics systems including Apache Hadoop, Dryad, and Apache Spark abstract jobs as directed acyclic graphs (DAGs). Speeding up completions for DAG jobs matter in practice in order to support real-time decisions. State-of-the-art works propose clairvoyant schedulers to optimize these goals, however, they assume complete job information as a prior knowledge which includes the precise DAG structure, and fine-grained resource requirement and duration time of each task. This assumption limits their applicability. In this paper, to be more practical, we relax the complete prior knowledge assumption and rely solely on partial prior information, based on which, we design a semi-clairvoyant task scheduler Cobra operating within each job. When managing resources for a job, Cobra adaptively adjusts its resource desires in a multiplicative-increase multiplicative-decrease manner on the basis of the its current resource utilization and the presence of current waiting tasks. When assigning tasks to run on the allocated resources, Cobra strives to satisfy task locality preference by tolerating each task waiting for some time that is bounded by a parameterized threshold. Even with the partial prior job information, when a set of jobs in which each employing Cobra as its task scheduler, run on a cluster that employs the fair job scheduler, we theoretically prove the produced makespan and average job response time are O(1)-competitive in different settings. We implement our design in Spark on YARN system, and use experiments from both real deployments and simulations on Google's trace to verify the performance promotion and sensitivity of Cobra.

Fault-tolerant Real-time Fair Scheduling on Multiprocessor Systems with Cold-standby

The ability to maintain functional and temporal correctness in the presence of faults is a key requirement in many safety-critical embedded systems. This work proposes an efficient fault recovery mechanism for real-time multiprocessor systems scheduled using a low overhead, semi-partitioned optimal proportional fair scheduling technique. We assume a system that can handle a single permanent processor fault at any time, using cold back-ups (with pre-specified activation / recovery time subsequent to the detection of a fault). As a result of the fault, the system may suffer transient overloads during such recovery periods, potentially leading to unacceptable fairness deviations and consequent rejections / early terminations of critical jobs. The proposed fault-tolerant scheduler, called Fault Tolerant Fair Scheduler (FT-FS), attempts to minimize such job terminations / rejections during recovery, by judiciously redistributing slacks accumulated by a subset of jobs, delivering more sustainable performance in the process. Experimental results reveal that the proposed FT-FS algorithm performs appreciably even under high system loads. Practical applicability of our proposed scheme has been illustrated using a case study on aircraft flight control system.

Adaptive proportional fair scheduling with global-fairness

In practical communication systems, there are always multiple subscribers competing for limited resources, such as time and frequency, hence effective user scheduling is essential to multi-user communications in achieving good system throughput and fairness performance. The conventional proportional fair (PF) scheduling achieves fairness at the cost of system spectral efficiency (SE) loss. Such fairness is of long-term feature, i.e., all the users’ scheduling probabilities become approximately the same only when the observation time is long enough. Therefore, PF cannot guarantee the fairness for subscribers who enter the system temporarily or stay in the system for a short period of time. In addition, delay requirement of real-time-service users can hardly be met with conventional PF. In order to remedy these deficiencies, we propose adaptive proportional fair (APF) scheduling algorithms. In each time slot, the infrastructure node, e.g., base station, dynamically adjusts the forgetting factor based on the variance of all the subscribers’ scheduling priorities, so that users’ scheduling weights can be adaptively updated. Our in-depth simulation results show that compared to conventional PF, APF can not only achieve both long-term and short-term fairness which we refer to global-fairness, but also obtain high system SE. Moreover, users’ delay performance can be obviously improved.