Greedy Scheduling Research Articles

Quick or cheap? Breaking points in dynamic markets

We examine two-sided markets where players arrive stochastically over time. The cost of matching a client and provider is heterogeneous, and the distribution of costs – but not their realization – is known. In this way, a social planner is faced with two contending objectives:(a) to reduce the players’ waiting time before getting matched; and (b) to reduce matching costs. In this paper, we aim to understand when and how these objectives are incompatible. We identify two regimes dependent on the ‘speed of improvement’ of the cost of matching with respect to market size. One regime results in a quick or cheap dilemma without ‘free lunch’: there exists no clearing schedule that is simultaneously optimal along both objectives. In that regime, we identify a unique breaking point signifying a stark reduction in matching cost contrasted by an increase in waiting time. The other regime features a window of opportunity in which free lunch can be achieved. Which scheduling policy is optimal depends on the heterogeneity of match costs. Under limited heterogeneity, e.g., when there is a finite number of possible match costs, greedy scheduling is approximately optimal, in line with the related literature. However, with more heterogeneity greedy scheduling is never optimal. Finally, we analyze a particular model where match costs are exponentially distributed and show that it is at the boundary of the no-free-lunch regime We then characterize the optimal clearing schedule for varying social planner desiderata.

Outage probability analysis of a cooperative NOMA UWOC system with multiuser scheduling under weak oceanic turbulence.

A downlink cooperative non-orthogonal multiple access (NOMA) multiuser underwater wireless optical communication (UWOC) system with a greedy scheduling scheme has been proposed for the Internet of Underwater Things. In particular, the near user plays as a relay to assist the far user, and both decode-and-forward and amplify-and-forward relaying protocols are considered. Relying on the Gauss-Laguerre quadrature formula, the analytical expressions for the outage probability of users are derived considering the degrading effects of the underwater channel, namely, absorption, scattering, and turbulence-induced fading. The outage performance is further analyzed systematically under different water types, targeted data rates, the number of users, the receiver aperture size, and the power allocation coefficient. Numerical results demonstrate that the performance of the far user can be improved by the proposed cooperative NOMA technology. Moreover, the proposed cooperative NOMA system performs better compared with both conventional OMA and non-cooperative NOMA systems. Monte Carlo simulation results are presented to confirm the accuracy of derived expressions, which have a tight agreement with analytical results.

Graph-Based Deadlock Analysis and Prevention for Robust Intelligent Intersection Management

Intersection management systems, with the assistance of vehicular networks and autonomous vehicles, have potential to perform traffic control more precisely than contemporary signalized intersections. However, as infrastructural intersection management controllers do not directly activate motions of vehicles, it is possible that the vehicles fail to follow the instructions from controllers, undermining system properties such as deadlock-freeness and traffic performance. In this paper, we consider a class of robustness issues, the time violations, which stem from possible discrepancies between scheduled orders and real executions. We refine a graph-based intersection model to build our theoretical foundations and analyze potential deadlocks and their resolvability. We develop solutions that mitigate negative effects of time violations. Particularly, we propose a Robustness-Aware Greedy Scheduling (RGS) algorithm for robust scheduling and evaluate the deadlock-free robustness of different intersection models and scheduling algorithms. Experimental results show that the RGS algorithm is able to significantly improve robustness and keep a good balance with traffic performance.

Fast Link Scheduling in Wireless Networks Using Regularized Off-Policy Reinforcement Learning

The centralized-link-scheduling problem in a wireless network graph involves solving the maximum-weighted-independent-set (MWIS) problem on the conflict graph. In this paper, we propose a novel regularized off-policy reinforcement learning-based MWIS solver and use for the scheduling problem. The proposed MWIS algorithm achieves 17% improvement over state-of-the-art heuristic solver KaMIS, 60% over greedy solver, 16% and 17% over RL-based solvers LwD and S2V-DQN, respectively. We show that our scheduler achieves stable throughput values 14% and 22% higher than LwD and a distributed greedy scheduler, respectively. We demonstrate the flexibility of our RL algorithm by modifying it to create a time-since-last-service-aware scheduler.

A massively parallel implementation of multilevel Monte Carlo for finite element models

The Multilevel Monte Carlo (MLMC) method has proven to be an effective variance-reduction statistical method for Uncertainty Quantification (UQ) in Partial Differential Equation (PDE) models, combining model computations at different levels to create an accurate estimate. Still, the computational complexity of the resulting method is extremely high, particularly for 3D models, which requires advanced algorithms for the efficient exploitation of High Performance Computing (HPC). In this article we present a new implementation of the MLMC in massively parallel computer architectures, exploiting parallelism within and between each level of the hierarchy. The numerical approximation of the PDE is performed using the finite element method but the algorithm is quite general and could be applied to other discretization methods. The two key ingredients of the implementation are a good processor partition scheme together with a good scheduling algorithm to assign work to different processors. We introduce a multiple partition of the set of processors that permits the simultaneous execution of different levels and we develop a dynamic scheduling algorithm to exploit it. The problem of finding the optimal scheduling of distributed tasks in a parallel computer is an NP-complete problem. We propose and analyze a new greedy scheduling algorithm to assign samples and we show that it is a 2-approximation, which is the best that may be expected under general assumptions. On top of this result we design a distributed memory implementation using the Message Passing Interface (MPI) standard. Finally we present a set of numerical experiments illustrating its scalability properties.

Reinforcement learning - based adaptation and scheduling methods for multi-source DASH

Dynamic adaptive streaming over HTTP (DASH) has been widely used in video streaming recently. In DASH, the client downloads video chunks in order from a server. The rate adaptation function at the video client enhances the user?s quality-of-experience (QoE) by choosing a suitable quality level for each video chunk to download based on the network condition. Today networks such as content delivery networks, edge caching networks, contentcentric networks, etc. usually replicate video contents on multiple cache nodes. We study video streaming from multiple sources in this work. In multi-source streaming, video chunks may arrive out of order due to different conditions of the network paths. Hence, to guarantee a high QoE, the video client needs not only rate adaptation, but also chunk scheduling. Reinforcement learning (RL) has emerged as the state-of-the-art control method in various fields in recent years. This paper proposes two algorithms for streaming from multiple sources: RL-based adaptation with greedy scheduling (RLAGS) and RL-based adaptation and scheduling (RLAS). We also build a simulation environment for training and evaluation. The efficiency of the proposed algorithms is proved via extensive simulations with real-trace data.

Accelerating the scheduling of the network resources of the next-generation optical data centers

Data centers (DCs) play a key role in the evolving IT applications and they rely heavily on the optical interconnects to improve their performance and scalability. Optically switched DCs most often exploit the slotted Time Division Multiplexing Access (TDMA) operation and the Wavelength Division Multiplexing (WDM) technology and rely on the effective scheduling of the TDMA frames to decide in real time the end-to-end connections that include the network links, switches and ports. This task becomes computationally intensive as the communication requests increase.The current paper builds on a greedy scheduling algorithm to introduce a parallel technique that accelerates the scheduling process and improves optical DC’s performance. The proposed technique handles efficiently the scheduler’s data structures, minimizes the communication among the scheduler’s processors and it is scalable. Moreover, this work presents the technique’s performance results for a variety of scheduling scenarios and DC sizes executed on an algorithm-specific Single Instruction Multiple Data (SIMD) accelerator architecture and on a Graphics Processing Unit (GPU). The performance of the GPU and the SIMD accelerator implemented on FPGA validate the parallel scheduler technique.

Complete Targets Coverage in Energy Harvesting Internet of Things With Ambient Backscatter

This article considers deriving set covers or the set of active nodes that are responsible for monitoring all targets in an Internet of Things (IoT) network. A key distinction to prior works is that this article considers sensor nodes aided by backscatter communications, which allow them to communicate with negligible energy cost using ambient radio frequency (RF) signals. The article contains three main novelties. First, it presents a mixed integer linear program (MILP) that can be used to compute the global optimal solution. Second, it also outlines a centralized greedy scheduling (CGS) algorithm that selects nodes based on their energy and the number of covered targets. Third, it presents a distributed greedy scheduling (DGS) algorithm that selects nodes according to their energy level. The simulation results show that equipping nodes with ambient backscattering capability increases the ratio of complete targets coverage by more than 100% as compared to existing techniques.

Space-Based Sensor Tasking Using Deep Reinforcement Learning

To maintain a robust catalog of resident space objects (RSOs), space situational awareness (SSA) mission operators depend on ground- and space-based sensors to repeatedly detect, characterize, and track objects in orbit. Although some space sensors are capable of monitoring large swaths of the sky with wide fields of view (FOVs), others—such as maneuverable optical telescopes, narrow-band imaging radars, or satellite laser-ranging systems—are restricted to relatively narrow FOVs and must slew at a finite rate from object to object during observation. Since there are many objects that a narrow FOV sensor could choose to observe within its field of regard (FOR), it must schedule its pointing direction and duration using some algorithm. This combinatorial optimization problem is known as the sensor-tasking problem. In this paper, we developed a deep reinforcement learning agent to task a space-based narrow-FOV sensor in low Earth orbit (LEO) using the proximal policy optimization algorithm. The sensor’s performance—both as a singular sensor acting alone, but also as a complement to a network of taskable, narrow-FOV ground-based sensors—is compared to the greedy scheduler across several figures of merit, including the cumulative number of RSOs observed and the mean trace of the covariance matrix of all of the observable objects in the scenario. The results of several simulations are presented and discussed. Additionally, the results from an LEO SSA sensor in different orbits are evaluated and discussed, as well as various combinations of space-based sensors.

Optimizing Age-of-Information in Adversarial and Stochastic Environments

We design efficient online scheduling policies to maximize the freshness of information delivered to the users in a cellular network under both adversarial and stochastic channel and mobility assumptions. The information freshness achieved by a policy is investigated through the lens of a recently proposed metric - <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Age-of-Information</i> (AoI). We show that a natural greedy scheduling policy is competitive against any optimal offline policy in minimizing the AoI in the adversarial setting. We also derive universal lower bounds to the competitive ratio achievable by any online policy in the adversarial framework. In the stochastic setting, we show that a simple index policy is near-optimal for minimizing the average AoI in two different mobility scenarios. Further, we prove that the greedy scheduling policy minimizes the peak AoI for static users in the stochastic setting. Simulation results show that the proposed policies perform well under realistic conditions.

Power control scheme using particle swarm optimization method in resource allocation process on D2D underlaying communication

Fast growing number of cellular telecommunication technology resulting on the increasing number of the user equipment. This condition increased the eNodeB load. To overcome this problem, the device-to-device (D2D) underlaying communication is introduced. In underlaying scheme, the D2D user equipment (DUE) will do the communication process using the same radio resources with the conventional cellular user equipment (CUE). To avoid a severe interference between these two types of user in the system, a good resource allocation is needed. In this work, a power control scheme using particle swarm optimization (PSO) is proposed, to manage the transmit power on each user on the system. The power control scheme take place after the greedy scheduling algorithm, after all user is given a resource block (RB) to do the communication process. The power transmit for each user is managed to reach a better system capacity, and to reduce the power consumed in one communication process. From the simulation, the PSO power control can improve the sumrate and spectral efficiency up to 12.97% and 3.38% respectively. The PSO power control also can reduce the power consumed by the system up to 8.84%. The fairness happens among the CUEs also can be maintained, despite of the decreasing fairness among DUEs.

Resource-Efficient Pareto-Optimal Green Scheduler Architecture

Rapidly developing cloud technology with enormous number of clients creates need of reducing power consumption of data centers. VM live migration is the most promising tool to achieve resource consolidation but it creates overheads in terms of additional CPU, disk I/O and network bandwidth utilization. This paper proposes a power-aware VM live migration based dynamic VM consolidation mechanism that focuses on reduction in datacenter’s resource utilization. Proposed mechanism is Pareto Optimal because during live migration it not only optimize the migration overheads but also select the VM and destination server by considering all the performance overheads to be generated during and after live migration. The proposed algorithm reduces nearly 60% of the VMs migration overheads. In terms of energy saving the proposed mechanism is 43% more efficient than the greedy scheduling approach and about 47% more energy efficient than the round-robin approach and thus achieves green computing goal.

A storage computing architecture with multiple NDP devices for accelerating compaction performance in LSM-tree based KV stores

In the era of big data, log-structured merge-tree (LSM-tree) based key–value stores (KV stores) have become an effective alternative to the traditional relational database owing to their high performance and scalability. KV stores are widely used to manage big unstructured data for data-intensive applications. A compaction operation is triggered to sort data and compact space, but this can degrade write performance, causing write amplification to increase. The compaction process results in extensive data movement and computation. Recently, near-data processing (NDP) models, which place computation close to data sources, aiming to reduce the cost of data movement between the device and the host, have been widely applied to optimize the performance of KV stores. The existing researches on NDP-based KV stores mainly focus on a single NDP-enabled storage device (NDP device for short). In this paper, we present a storage computing architecture with multiple NDP devices named MStore for accelerating compaction performance in LSM-tree based KV stores. MStore can achieve high performance owing to its storage and computing convergence. It employs a multi-column LSM-tree data layout based on key ranges. Compaction tasks on multiple NDP devices are performed in parallel, improving compaction performance. MStore increases the width of the LSM-tree level to reduce the depth, which reduces the execution time and write amplification of compaction. Following the data organization, we design load balancing mechanisms, including storage load balancing by key range adjustment and computation load balancing across compaction tasks. In addition, we develop a batch read interface and a greedy scheduling scheme for user I/Os. MStore aims to fully utilize the computing and I/O resources to boost the compaction performance in KV stores. Compared with single-device NDP systems (such as TStore), MStore with four NDP devices achieves 3.88× speedup under random-write DB_Bench. When running YCSB-C with a random-write workload, MStore is 4.32× faster than TStore. The performance of read and scan on MStore is improved by 2.23× and 2.18×, respectively, compared with TStore. These results have strong implications for the implementations of MStore, and their use as replacements for single NDP based KV stores.

Sliding-Window-Based RNC Scheme in UAV Multicasting: Performance Analysis and Network Optimization

Unmanned aerial vehicles (UAVs)-enabled multicasting network has attracted significant attention in recent years. However, there are still some disadvantages of existing multicasting schemes used in these systems, such as low transmission efficiency and high feedback overhead. Accordingly, we propose a sliding coding window (SCW)-based random network coding (SCWRNC) scheme for a UAV multicasting network where one UAV base station is dispatched to the multicast data stream to multiple user equipments (UEs). The proposed scheme includes an SCW scheduling original packets for encoding, a lower triangular coding structure enabling UEs to decode out information even without receiving a full set of coded packets, and a feedback-compete mechanism requiring only one UE to send feedback information. The packet scheduling process is described as a five-tuple Markov decision process. Then, we give a theoretical analysis of the proposed scheme, based on which the sliding steps of SCW and the UAV hovering location are jointly optimized to maximize the system throughput. The optimal sliding steps are obtained by applying the Greedy scheduling technique, while the UAV optimal position is obtained by minimizing the maximum outage probability of all UEs. Furthermore, we also propose a flexible feedback mechanism, which enables more than one UE to send feedback for systems with sufficient resources and a “F-SCWRNC” scheme for systems where no UE is allowed to send feedback. Numerical results show that both the proposed SCWRNC scheme and F-SCWRNC scheme could achieve significant throughput gain over the existing ones.

An Energy-Saving Task Scheduling Model via Greedy Strategy under Cloud Environment

Cloud computing, an emerging computing paradigm, has been widely concerned due to its high scalability and availability. An essential stage of cloud computing is cloud resource management. Currently, the existing research about cloud computing technology has two prevalent disadvantages: high energy consumption and low resource utilization. Considering greedy scheduling is an effective strategy for cloud resource management technology in cloud computing, particularly in improving resource utilization and reducing energy consumption, we consider the heterogeneous characteristics of resources to save energy consumption of datacenter when tasks are the fundamental element of cloud datacenter. Meanwhile, granular computing is a complex problem-solving strategy through a granulation method. Thus, we introduce granular computing theory into cloud task scheduling and propose a greedy scheduling strategy based on different information granules, dividing the tasks into three types (i.e., CPU, memory, and hybrid type). Finally, we assign various scheduling strategies for cloud tasks with different characteristics. All the numerical experiments on the CloudSim platform show that our method has significant effects on energy consumption optimization and is a practical task scheduling algorithm.

Exact and heuristic methods for the berth allocation problem with multiple continuous quays in tidal bulk terminals

The Berth Allocation Problem (BAP) is a primary seaside operations planning problem in bulk terminals. It consists of allocating quayside space to incoming vessels. In this article, the BAP for multiple continuous quays and dynamic arrivals is considered. The formulation considers the tidal constraints typical to exporting bulk terminals and restrictions regarding each vessel's possible quays. To solve the problem, a mixed-integer linear model is first presented. As the complexity of the problem grows exponentially with the size of the instances, an Iterated Local Search (ILS) approach is proposed to solve industrial-sized instances. The ILS approach works on sequences of vessels that are decoded using a bottom-left constructive heuristic. The initial sequences are obtained using a greedy scheduling heuristic that provides suitable starting solutions, especially for congested terminals. Extensive numerical experiments are carried on randomly generated instances for tidal bulk terminals and reported benchmark sets in literature. The results prove that the proposed ILS can provide good-quality solutions.

Heuristic Greedy Scheduling of Electric Vehicles in Vehicle-to-Grid Microgrid Owned Aggregators.

In on-grid microgrids, electric vehicles (EVs) have to be efficiently scheduled for cost-effective electricity consumption and network operation. The stochastic nature of the involved parameters along with their large number and correlations make such scheduling a challenging task. This paper aims at identifying pertinent innovative solutions for reducing the relevant total costs of the on-grid EVs within hybrid microgrids. To optimally scale the EVs, a heuristic greedy approach is considered. Unlike most existing scheduling methodologies in the literature, the proposed greedy scheduler is model-free, training-free, and yet efficient. The proposed approach considers different factors such as the electricity price, on-grid EVs state of arrival and departure, and the total revenue to meet the load demands. The greedy-based approach behaves satisfactorily in terms of fulfilling its objective for the hybrid microgrid system, which is established of photovoltaic, wind turbine, and a local utility grid. Meanwhile, the on-grid EVs are being utilized as an energy storage exchange location. A real time hardware-in-the-loop experimentation is comprehensively conducted to maximize the earned profit. Through different uncertainty scenarios, the ability of the proposed greedy approach to obtain a global optimal solution is assessed. A data simulator was developed for the purposes of generating evaluation datasets, which captures uncertainties in the behaviors of the system’s parameters. The greedy-based strategy is considered applicable, scalable, and efficient in terms of total operating expenditures. Furthermore, as EVs penetration became more versatile, total expenses decreased significantly. Using simulated data of an effective operational duration of 500 years, the proposed approach succeeded in cutting down the energy consumption costs by about 50–85%, beating existing state-of-the-arts results. The proposed approach is proved to be tolerant to the large amounts of uncertainties that are involved in the system’s operational data.

Bio-inspired negotiation approach for smart-grid colocation datacenter operation.

Demand response programs allow consumers to participate in the operation of a smart electric grid by reducing or shifting their energy consumption, helping to match energy consumption with power supply. This article presents a bio-inspired approach for addressing the problem of colocation datacenters participating in demand response programs in a smart grid. The proposed approach allows the datacenter to negotiate with its tenants by offering monetary rewards in order to meet a demand response event on short notice. The objective of the underlying optimization problem is twofold. The goal of the datacenter is to minimize its offered rewards while the goal of the tenants is to maximize their profit. A two-level hierarchy is proposed for modeling the problem. The upper-level hierarchy models the datacenter planning problem, and the lower-level hierarchy models the task scheduling problem of the tenants. To address these problems, two bio-inspired algorithms are designed and compared for the datacenter planning problem, and an efficient greedy scheduling heuristic is proposed for task scheduling problem of the tenants. Results show the proposed approach reports average improvements between 72.9% and 82.2% when compared to the business as usual approach.

Anarchy in the UJ: Coordination mechanisms for minimizing the number of late jobs

We consider the distributed scheduling problem on parallel machines with the central objective of maximizing the number of on-time jobs. Jobs are self-interested utility-maximizers that can choose the machines they are processed on and are exclusively interested in reducing their own private objective function. Each machine processes the jobs according to a local policy. We discuss Nash equilibria in the resulting schedules and perform a thorough analysis of the resulting (absolute) prices of anarchy for various parallel machine environments, utilities of the agents, and local policies of the machines. We show that local policies that are based on simple sorting-based procedures like shortest processing time first (SPT) and earliest due date first (EDD) lead to big losses in welfare compared to the global optimum. However, when employing Moore-Hodgson’s algorithm as a local policy, we can prove a price of anarchy of (2m−1)/m for identical machines and a price of anarchy of 2 for related and unrelated parallel machines. Moreover, we show how these results can be used to prove approximation ratios for greedy scheduling algorithms. This paper is the first to prove approximation ratios for two greedy scheduling procedures, turning them from simple heuristics into actual approximation ratios with a provable approximation ratio.

Timeline-Based Space Operations Scheduling with External Constraints

We describe a timeline-based scheduling algorithm developed for mission operations of the EO-1 earth observing satellite. We first describe the range of operational constraints for operations focusing on maneuver and thermal constraints that cannot be modeled in typical planner/schedulers. We then describe a greedy heuristic scheduling algorithm and compare its performance to both the prior scheduling algorithm - documenting an over 50% increase in scenes scheduled with estimated value of millions of dollars US. We also compare to a relaxed optimal scheduler showing that the greedy scheduler produces schedules with scene count within 15% of an upper bound on optimal schedules.