A strategy for scheduling tightly coupled parallel applications on clusters

Although various strategies have been developed for scheduling parallel applications with independent tasks, very little work exists for scheduling tightly coupled parallel applications on cluster environments. In this paper, we compare four different strategies based on performance models of tightly coupled parallel applications for scheduling the applications on clusters. In addition to algorithms based on existing popular optimization techniques, we also propose a new algorithm called Box Elimination that searches the space of performance model parameters to determine the best schedule of machines. By means of real and simulation experiments, we evaluated the algorithms on single cluster and multi-cluster setups. We show that our Box Elimination algorithm generates up to 80% more efficient schedule than other algorithms. We also show that the execution times of the schedules produced by our algorithm are more robust against the performance modeling errors.

Cross-Layer Scheduling with Infrequent Channel and Queue Measurements

The downlink scheduling problem in multi-queue multi-server systems under channel uncertainty is considered. Two policies that make allocations based on predicted channel states are proposed. The first is an extension of the well-known dynamic backpressure policy to the uncertain channel case. The second is a variant that improves delay performance under light loads. The stability region of the system is characterised and the first policy is argued to be throughput optimal. A recently proposed policy of Kar et al [1] has lesser complexity, but is shown to be throughput suboptimal. Further, simulations demonstrate better delay and backlog properties for both our policies at light loads.

Scheduling techniques to optimise sugarcane rail transport systems

In Australia, railway systems play a vital role in transporting the sugarcane crop from farms to mills. In this paper, a novel job shop approach is proposed to create a more efficient integrated harvesting and sugarcane transport scheduling system to reduce the cost of sugarcane transport. There are several benefits that can be attained by treating the train scheduling problem as a job shop problem. Job shop is generic and suitable for all trains scheduling problems. Job shop technique prevents operating two trains on one section at the same time because it considers that the section or the machine is unique. This technique is more promising to find better solutions in reasonable computation times.

Delay Optimal Scheduling in a Two-Hop Vehicular Relay Network

We study a scheduling problem in a wireless network where vehicles are used as store-and-forward relays, a situation that might arise, for example, in practical rural communication networks. A fixed source node wants to transfer a file to a fixed destination node, located beyond its communication range. In the absence of any infrastructure connecting the two nodes, we consider the possibility of communication using vehicles passing by. Vehicles arrive at the source node at renewal instants and are known to travel towards the destination node with average speed v sampled from a given probability distribution. Th source node communicates data packets (or fragments) of the file to the destination node using these vehicles as relays. We assume that the vehicles communicate with the source node and the destination node only, and hence, every packet communication involves two hops. In this setup, we study the source node's sequential decision problem of transferring packets of the file to vehicles as they pass by, with the objective of minimizing delay in the network. We study both the finite file size case and the infinite file size case. In the finite file size case, we aim to minimize the expected file transfer delay, i.e. expected value of the maximum of the packet sojourn times. In the infinite file size case, we study the average packet delay minimization problem as well as the optimal tradeoff achievable between the average queueing delay at the source node buffer and the average transit delay in the relay vehicle.

An efficient algorithm for scheduling in bluetooth piconets and scatternets

Bluetooth is a short-range radio technology operating in the unlicensed industrial-scientific-medical (ISM) band at 2.45 GHz. A piconet is basically a collection of slaves controlled by a master. A scatternet, on the other hand, is established by linking several piconets together in an ad hoc fashion to yield a global wireless ad hoc network. This paper proposes a scheduling policy that aims to achieve increased system throughput and reduced packet delays while providing reasonably good fairness among all traffic flows in bluetooth piconets and scatternets. We propose a novel algorithm for scheduling slots to slaves for both piconets and scatternets using multi-layered parameterized policies. Our scheduling scheme works with real data and obtains an optimal feedback policy within prescribed parameterized classes of these by using an efficient two-timescale simultaneous perturbation stochastic approximation (SPSA) algorithm. We show the convergence of our algorithm to an optimal multi-layered policy. We also propose novel polling schemes for intra- and inter-piconet scheduling that are seen to perform well. We present an extensive set of simulation results and performance comparisons with existing scheduling algorithms. Our results indicate that our proposed scheduling algorithm performs better overall on a wide range of experiments over the existing algorithms for both piconets (Das et al. in INFOCOM, pp. 591–600, 2001; Lapeyrie and Turletti in INFOCOM conference proceedings, San Francisco, US, 2003; Shreedhar and Varghese in SIGCOMM, pp. 231–242, 1995) and scatternets (Har-Shai et al. in OPNETWORK, 2002; Saha and Matsumot in AICT/ICIW, 2006; Tan and Guttag in The 27th annual IEEE conference on local computer networks(LCN). Tampa, 2002). Our studies also confirm that our proposed scheme achieves a high throughput and low packet delays with reasonable fairness among all the connections.

A solution framework for scheduling a BPM with non-identical job dimensions

A Batch Processing Machine (BPM) is one which processes a number of jobs simultaneously as a batch with common beginning and ending times. Also, a BPM, once started cannot be interrupted in between (Pre-emption not allowed). This research is motivated by a BPM in steel casting industry. There are three main stages in any steel casting industry viz., pre-casting stage, casting stage and post-casting stage. A quick overview of the entire process, is shown in Figure 1. There are two BPMs : (1) Melting furnace in the pre-casting stage and (2) Heat Treatment Furnace (HTF) in the post casting stage of steel casting manufacturing process. This study focuses on scheduling the latter, namely HTF. Heat-treatment operation is one of the most important stages of steel casting industries. It determines the final properties that enable components to perform under demanding service conditions such as large mechanical load, high temperature and anti-corrosive processing. In general, different types of castings have to undergo more than one type of heat-treatment operations, where the total heat-treatment processing times change. To have a better control, castings are primarily classified into a number of job-families based on the alloy type such as low-alloy castings and high alloy castings. For technical reasons such as type of alloy, temperature level and the expected combination of heat-treatment operations, the castings from different families can not be processed together in the same batch.

An efficient and optimized bluetooth scheduling algorithm for piconets

Bluetooth is an emerging standard in short range, low cost and low power wireless networks. MAC is a generic polling based protocol, where a central Bluetooth unit (master) determines channel access to all other nodes (slaves) in the network (piconet). An important problem in Bluetooth is the design of efficient scheduling protocols. This paper proposes a polling policy that aims to achieve increased system throughput and reduced packet delays while providing reasonably good fairness among all traffic flows in a Bluetooth Piconet. We present an extensive set of simulation results and performance comparisons with two important existing algorithms. Our results indicate that our proposed scheduling algorithm outperforms the Round Robin scheduling algorithm by more than 40% in all cases tried. Our study also confirms that our proposed policy achieves higher throughput and lower packet delays with reasonable fairness among all the connections.

Complexity of scheduling for minimum power on a GMAC

Two decision versions of a combinatorial power minimization problem for scheduling in a time-slotted Gaussian multiple-access channel (GMAC) are studied in this paper. If the number of slots per second is a variable, the problem is shown to be NP-complete. If the number of time-slots per second is fixed, an algorithm that terminates in O (Length (I)N+1) steps is provided.

Data-aware task scheduling for all-to-all comparison problems in heterogeneous distributed systems

Solving large-scale all-to-all comparison problems using distributed computing is increasingly significant for various applications. Previous efforts to implement distributed all-to-all comparison frameworks have treated the two phases of data distribution and comparison task scheduling separately. This leads to high storage demands as well as poor data locality for the comparison tasks, thus creating a need to redistribute the data at runtime. Furthermore, most previous methods have been developed for homogeneous computing environments, so their overall performance is degraded even further when they are used in heterogeneous distributed systems. To tackle these challenges, this paper presents a data-aware task scheduling approach for solving all-to-all comparison problems in heterogeneous distributed systems. The approach formulates the requirements for data distribution and comparison task scheduling simultaneously as a constrained optimization problem. Then, metaheuristic data pre-scheduling and dynamic task scheduling strategies are developed along with an algorithmic implementation to solve the problem. The approach provides perfect data locality for all comparison tasks, avoiding rearrangement of data at runtime. It achieves load balancing among heterogeneous computing nodes, thus enhancing the overall computation time. It also reduces data storage requirements across the network. The effectiveness of the approach is demonstrated through experimental studies.

Efficient scheduling of sensor activity for information coverage in wireless sensor networks

In this paper, we are concerned with algorithms for scheduling the sensing activity of sensor nodes that are deployed to sense/measure point-targets in wireless sensor networks using information coverage. Defining a set of sensors which collectively can sense a target accurately as an information cover, we propose an algorithm to obtain Disjoint Set of Information Covers (DSIC), which achieves longer network life compared to the set of covers obtained using an Exhaustive-Greedy-Equalized Heuristic (EGEH) algorithm proposed recently in the literature. We also present a detailed complexity comparison between the DSIC and EGEH algorithms.

Optimal sleep-wake scheduling for quickest intrusion detection using sensor networks

We consider the problem of quickest detection of an intrusion using a sensor network, keeping only a minimal number of sensors active. By using a minimal number of sensor devices, we ensure that the energy expenditure for sensing, computation and communication is minimized (and the lifetime of the network is maximized). We model the intrusion detection (or change detection) problem as a Markov decision process (MDP). Based on the theory of MDP, we develop the following closed loop sleep/wake scheduling algorithms: (1) optimal control of Mk+1, the number of sensors in the wake state in time slot k + 1, (2) optimal control of qk+1, the probability of a sensor in the wake state in time slot k + 1, and an open loop sleep/wake scheduling algorithm which (3) computes q, the optimal probability of a sensor in the wake state (which does not vary with time), based on the sensor observations obtained until time slot k. Our results show that an optimum closed loop control on Mk+1 significantly decreases the cost compared to keeping any number of sensors active all the time. Also, among the three algorithms described, we observe that the total cost is minimum for the optimum control on Mk+1 and is maximum for the optimum open loop control on q.

Scheduling tools to reduce the cost of cane railway transport

Cane railway systems provide empty bins for harvesters to fill and full bins of cane for the factory to process. These operations need to be conducted in a timely fashion to minimise delays to harvesters and the factory and to minimise the cut-to-crush delay, while also minimising the cost of providing this service. A range of tools has been provided over the years to assist in this process. This paper reviews the objectives of the cane transport system and the tools available to achieve those objectives. The facilities within these tools to assist in the control of costs are highlighted.

A real coded genetic algorithm for data partitioning and scheduling in networks with arbitrary processor release time

The problem of scheduling divisible loads in distributed computing systems, in presence of processor release time is considered. The objective is to find the optimal sequence of load distribution and the optimal load fractions assigned to each processor in the system such that the processing time of the entire processing load is a minimum. This is a difficult combinatorial optimization problem and hence genetic algorithms approach is presented for its solution.

A generalized Linear Programming based approach to optimal divisible load scheduling

In this paper we propose a general Linear Programming (LP) based formulation and solution methodology for obtaining optimal solution to the load distribution problem in divisible load scheduling. We exploit the power of the versatile LP formulation to propose algorithms that yield exact solutions to several very general load distribution problems for which either no solutions or only heuristic solutions were available. We consider both star (single-level tree) networks and linear daisy chain networks, having processors equipped with front-ends, that form the generic models for several important network topologies. We consider arbitrary processing node availability or release times and general models for communication delays and computation time that account for constant overheads such as start up times in communication and computation. The optimality of the LP based algorithms is proved rigorously.

Scheduling parallel batch processors with incompatible job families using ant colony optimization

The present work concerns with the static scheduling of jobs to parallel identical batch processors with incompatible job families for minimizing the total weighted tardiness. This scheduling problem is applicable in burn-in operations and wafer fabrication in semiconductor manufacturing. We decompose the problem into two stages: batch formation and batch scheduling, as in the literature. The Ant Colony Optimization (ACO) based algorithm called ATC-BACO algorithm is developed in which ACO is used to solve the batch scheduling problems. Our computational experimentation shows that the proposed ATC-BACO algorithm performs better than the available best traditional dispatching rule called ATC-BATC rule.