Greedy Scheduling with Complex Obejectives

Heterogeneous-ISA multi-core architectures have emerged as a promising design paradigm given the ever-increasing demands on single threaded performance. Such architectures comprise multiple cores that differ not just in micro-architectural parameters (e.g., fetch width, ROB size) but also in their Instruction Set Architectures (ISAs). These architectures extract previously latent performance gains by executing different phases of the program on the core (and ISA) best suited to it, as opposed to executing the entire program on a single ISA. In such a computing paradigm, maximum performance is only extracted when we ensure that at every point in the program’s execution, the program runs on the core best suited to it. In this work, we propose a migration framework that practically and accurately decides when to migrate the program across different cores (and ISAs) to extract maximum performance gains. Under the covers, this framework combines a regression based performance modelling technique with a greedy scheduling algorithm. Our performance modelling technique leverages hardware performance counters prevalent in all major processors today to accurately estimate the performance of the program on different ISAs to within an error of 6%. Putting it together with our greedy scheduler enables our framework to achieve single thread performance speedups of 29.6% with respect to a baseline single ISA heterogeneous architecture.

Edge computing provides a feasible technical solution to solve the problem of resource and energy supply constraints in terminal device deep learning applications. By migrating part of the deep learning application computing from the terminal to the cloud edge server, the computing pressure of the terminal device can be greatly relieved, and the energy consumption of the terminal can also be reduced to a certain extent. However, in the process of computing migration, the context switching loss of deep learning applications results in a long delay. This paper studies the problem of minimizing the completion time of deep learning applications under edge computing. We model the deep learning application as a task graph, and consider the constraints such as task switching delay, energy consumption and task dependency. Without changing the structure of the task graph, we transform it into a task scheduling-based application completion time minimization problem. In this paper, we first design a parallel scheduling algorithm based on greedy search. Considering the loss of context switching and the dependence of task data, we can allocate tasks to multiple processors for parallel processing by calculating the expected completion time of tasks, thus reducing the completion time of applications. Based on the experimental results, the greedy scheduling algorithm and comparison algorithm are analyzed and verified to optimize the completion time of different depth learning applications.

In this paper, we consider scheduling in multicast switching networks, which aims to minimize the multicast latency for a set of multicast requests. Such a problem has been proved to be NP-complete. We propose a simple, fast greedy multicast scheduling algorithm and derive a lower bound and an upper bound on the performance of the algorithm. As can be seen, while a lower bound is fairly straightforward, the upper bound is much more difficult to obtain. By translating the multicast scheduling problem into a graph theory problem and employing a random graph approach, we are able to obtain a probabilistic upper bound on the performance of the multicast scheduling algorithm. Our analytical and simulation results show that the performance of the proposed multicast scheduling algorithm is quite close to the lower bound and is statistically guaranteed by the probabilistic upper bound.

Workloads with precedence constraints due to data dependencies are common in various applications. These workloads can be represented as directed acyclic graphs (DAG), and are often data-intensive, meaning that data loading cost is the dominant factor and thus cache misses should be minimized. We address the problem of parallel scheduling of a DAG of data-intensive tasks to minimize makespan. To do so, we propose greedy online scheduling algorithms that take load balancing, data dependencies, and data locality into account. Simulations and an experimental evaluation using an Apache Spark cluster demonstrate the advantages of our solutions.

Considering a heterogeneous network where both macro base station (BS) and small cell (SC) nodes are equipped with massive number of antennas, this paper studies the performance for multiple-input multiple-output downlinks when the macro and small cells share the same spectrum, and hence interfere with each other. Suppose that the large-scale antenna arrays at both macro BS and SC nodes employ maximum-ratio transmission (MRT) or zero-forcing transmission (ZFT) precoding, and transmit data streams to served users simultaneously. A new pilot reuse pattern among SCs is proposed for channel estimation. Taking into account imperfect channel state information (CSI), capacity lower bounds for MRT and ZFT are derived, respectively, in closed-form expressions involving only statistical CSI, followed by asymptotic analyses for massive arrays under specific power scaling laws. Subsequently, two user scheduling algorithms, greedy scheduling and asymptotical scheduling algorithm (ASA), are proposed based on derived capacity lower bounds and asymptotic analyses, respectively. ASA is demonstrated to be a near optimal in the asymptotic regime and has low complexity. Finally, the derived closed-form expressions are verified to be accurate predictors of the system performance by Monte Carlo simulations. Numerical results demonstrate the effectiveness of the asymptotic analysis and proposed user scheduling schemes.

Multiple-input multiple-output (MIMO) wireless communications systems have attracted a lot of attentions, because of their potential for high gain in channel capacity. Recently, attention has shifted to multiuser MIMO (MU-MIMO) systems. In the schemes for downlink MU-MIMO system, dirty paper coding (DPC) and block diagonalization (BD) algorithms can provide higher sum rate capacity of MU-MIMO systems; however, exhaustive user scheduling for them may cause prohibitive computation complexity. The greedy scheduling algorithms were proposed for DPC and for BD; and proportionally fair (PF) scheduling for MU-MIMO was also discussed in the previous works. However, the discussions in where mainly based on the system model with equal average signal to noise power ratio (SNR) users. In this paper, we focus on the user scheduling for the downlink of MU-MIMO systems with DPC transmission scheme, where each user is assumed to have only one receive antenna, which will simplify the analysis and is expected to give more insight of PF scheduling. We also assume each individual user has different average transmit SNR in the system. We modify the conventional user scheduling algorithm for DPC to modified Gram-Schmidt orthogonalization (M-GSO) based algorithm and extend it into PF scheduling. We also propose a modified PF metric corresponding to the unequal SNR users system model, that increases the sum rate capacity of MU-MIMO system.

This study investigates the scheduling problem for a single-cell downlink distributed antenna systems (DASs) with a massive number of remote access units (RAUs). To reduce signalling overhead under limited backhaul capacity, the authors make use of local long-term channel state information (CSI) in coordinated scheduling design. They first derive the ergodic rate expressions for both the single RAU transmission and the cooperative RAU transmission modes as functions of long-term CSI. Then greedy scheduling algorithms (GSAs) aiming for the maximum ergodic sum rate for the massive DAS using long-term CSI are proposed. To mitigate the intra-cell interference, a two-stage GSA with hybrid transmission mode is devised. Asymptotic analysis reveals that as the number of RAUs goes to infinity, intra-cell interference can be effectively mitigated. Simulation results verify the analysis and demonstrate that the two-stage GSA exhibits a higher ergodic sum-rate.

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.

The Internet of Things (IoT) is defined as interconnected digital and mechanical devices with intelligent and interactive data transmission features over a defined network. The ability of the IoT to collect, analyze and mine data into information and knowledge motivates the integration of IoT with grid and cloud computing. New job scheduling techniques are crucial for the effective integration and management of IoT with grid computing as they provide optimal computational solutions. The computational grid is a modern technology that enables distributed computing to take advantage of a organization’s resources in order to handle complex computational problems. However, the scheduling process is considered an NP-hard problem due to the heterogeneity of resources and management systems in the IoT grid. This paper proposed a Greedy Firefly Algorithm (GFA) for jobs scheduling in the grid environment. In the proposed greedy firefly algorithm, a greedy method is utilized as a local search mechanism to enhance the rate of convergence and efficiency of schedules produced by the standard firefly algorithm. Several experiments were conducted using the GridSim toolkit to evaluate the proposed greedy firefly algorithm’s performance. The study measured several sizes of real grid computing workload traces, starting with lightweight traces with only 500 jobs, then typical with 3000 to 7000 jobs, and finally heavy load containing 8000 to 10,000 jobs. The experiment results revealed that the greedy firefly algorithm could insignificantly reduce the makespan makespan and execution times of the IoT grid scheduling process as compared to other evaluated scheduling methods. Furthermore, the proposed greedy firefly algorithm converges on large search spacefaster , making it suitable for large-scale IoT grid environments.

In this paper we are considering the problem of tasks scheduling executed simultaneously on multiple identical machines with a cost-criterion, which is the product of the tasks execution time and the number of used machines. Moreover, it is assumed that between the tasks performed sequentially there must be setup of the machines performed. Solution to the problem comes down to a generalization of a two-dimensional packing problem. The paper presents simulated annealing algorithm with different variants of packing strategy. The conducted computational experiments proved that designated solution differs little from some lower bounds for the constraints of the objective function.

With the rapid expansion of optical networks, the number of network cutover tasks has increased significantly. Manual scheduling of cutover tasks typically takes several days, leading to inefficient resource utilization. To address this issue, we propose systematic intelligent task scheduling methods of cutover tasks in optical networks for what we believe to be the first time. The rules and attributes of cutover tasks are abstracted into graphical representations or mathematical expressions. Then, the scheduling process employs layered topological scheduling methods based on greedy heuristic algorithm, maximal cliques and particle swarm optimization. The scheduling results are verified in the real-time resource library by the conflict verification method. When conflicts arise, the scheduling process is iterated to find a conflict-free schedule, and prior rules are updated. The field trial results of national optical networks demonstrated that our proposed intelligent schedule method realizes second-level scheduling without conflicts. Compared with the layered topological scheduling method based on greedy algorithm, the scheduling methods based on maximal clique and particle swarm optimization can reduce the conflict rate by 20% and 30%, enhance the ability of conflict adjustment by 12% and 43%, and improve scheduling efficiency by 12% and 9%, respectively. The proposed intelligent scheduling methods can also autonomously generate multi-period conflict-free schedules.

This paper considers the channel-assignment and scheduling in wireless mesh networks that employ multiple radios and multiple channels. In contrast to the various algorithms available in the literature, we explicitly model the delay overhead that is incurred during channel switching, and use that delay in the design of algorithms. We prove that the well known Greedy Maximal Scheduling (GMS) algorithm does not have any provable efficiency ratio when the switching overhead is considered. We present a centralized algorithm (CGSSO), and a dynamic algorithm (DMSSO), both of which consider switching overhead. Simulation results show that the proposed algorithms significantly outperform other algorithms in packet throughput and average packet delay metrics. Results also show that the improvements in performance become more pronounced as the switching delay increases.

We present a new scheduling method for batch jobs on massively parallel processor architectures. This method is based on the First-come-first-serve strategy and emphasizes the notion of fairness. Severe fragmentation is prevented by using gang scheduling which is only initiated by highly parallel jobs. Good worst-case behavior of the scheduling approach has already been proven by theoretical analysis. In this paper we show by simulation with real workload data that the algorithm is also suitable to be applied in real parallel computers. This holds for several different scheduling criteria like makespan or sum of the flow times. Simulation is also used for determination of the best parameter set for the new method.

Cloud datacenters consume enormous energy and generate heat, which affects the environment. Hence, there must be proper management of resources in the datacenter for optimum usage of energy. Virtualization enabled computing improves the performance of the datacenters in terms of these parameters. Therefore, Virtual Machines (VMs) management is a required activity in the datacenter, which selects the VMs from the overloaded host for migration, VM migration from the underutilized host, and VM placement in the suitable host. In this paper, a method (SMA-LinR) has been developed using the Simple Moving Average (SMA) integrated with Linear Regression (LinR), which predicts the CPU utilization and determines the overloading of the host. Further, this predicted value is used to place the VMs in the appropriate PM. The main aim of this research is to reduce energy consumption (EC) and service level agreement violations (SLAV). Extensive simulations have been performed on real workload data, and simulation results indicate that SMA-LinR provides better EC and service quality improvements.

In this paper, we address job scheduling in Distributed Computing Infrastructures, that is a loosely coupled network of autonomous acting High Performance Computing systems. In contrast to the common approach of mutual workload exchange, we consider the more intuitive operator's viewpoint of load-dependent resource reconfiguration. In case of a site's over-utilization, the scheduling system is able to lease resources from other sites to keep up service quality for its local user community. Contrary, the granting of idle resources can increase utilization in times of low local workload and thus ensure higher efficiency. The evaluation considers real workload data and is done with respect to common service quality indicators. For two simple resource exchange policies and three basic setups we show the possible gain of this approach and analyze the dynamics in workload-adaptive reconfiguration behavior.