An Improved Randomized On-Line Algorithm for a Weighted Interval Selection Problem

Given a set of weighted intervals, the objective of the weighted interval selection problem (WISP) is to select a maximum-weight subset such that the selected intervals are pairwise disjoint. We consider on-line algorithms that process the intervals in order of non-decreasing left endpoints. Preemption is allowed, but rejections are irrevocable. This problem has natural applications in various scheduling problems. We study the class of monotone instances of WISP, i.e., we require that the order of right endpoints of the given intervals coincides with that of the left endpoints. This class includes the case where all intervals have the same length. For monotone instances of WISP, the best possible competitive ratio for deterministic on-line algorithms is known to be 1/4. It has long been an open question whether there exists a randomized algorithm with better competitive ratio. In this paper, we present a new randomized algorithm and prove that it achieves a better competitive ratio 1/3 for the special case of monotone WISP where the sequence of weights of the arriving intervals is non-decreasing. Thus we provide the first result towards a solution of the long-standing open question. Furthermore, we show that no randomized algorithm achieves a competitive ratio strictly larger than 4/5. This is the first non-trivial upper bound for randomized algorithms for monotone WISP.     

Online Randomized Multiprocessor Scheduling

The use of randomization in online multiprocessor scheduling is studied. The problem of scheduling independent jobs on m machines online originates with Graham [16]. While the deterministic case of this problem has been studied extensively, little work has been done on the randomized case. For m= 2 a randomized 4/3-competitive algorithm was found by Bartal et al. A randomized algorithm for m ≥ 3 is presented. It achieves competitive ratios of 1.55665, 1.65888, 1.73376, 1.78295, and 1.81681, for m = 3, 4, 5, 6,7 , respectively. These competitive ratios are less than the best deterministic lower bound for m=3,4,5 and less than the best known deterministic competitive ratio for m = 3,4,5,6,7 . Two models of multiprocessor scheduling with rejection are studied. The first is the model of Bartal et al. Two randomized algorithms for this model are presented. The first algorithm performs well for small m , achieving competitive ratios of 3/2 , , for m=2,3,4 , respectively. The second algorithm outperforms the first for m ≥ 5 . It beats the deterministic algorithm of Bartal et al. for all m = 5 ,. . ., 1000 . It is conjectured that this is true for all m . The second model differs in that preemption is allowed. For this model, randomized algorithms are presented which outperform the best deterministic algorithm for small m . Received August 11, 1997; revised February 25, 1998.     

Scheduling parallel jobs to minimize the makespan

We consider the NP-hard problem of scheduling parallel jobs with release dates on identical parallel machines to minimize the makespan. A parallel job requires simultaneously a prespecified, job-dependent number of machines when being processed. We prove that the makespan of any nonpreemptive list-schedule is within a factor of 2 of the optimal preemptive makespan. This gives the best-known approximation algorithms for both the preemptive and the nonpreemptive variant of the problem. We also show that no list-scheduling algorithm can achieve a better performance guarantee than 2 for the nonpreemptive problem, no matter which priority list is chosen. List-scheduling also works in the online setting where jobs arrive over time and the length of a job becomes known only when it completes; it therefore yields a deterministic online algorithm with competitive ratio 2 as well. In addition, we consider a different online model in which jobs arrive one by one and need to be scheduled before the next job becomes known. We show that no list-scheduling algorithm has a constant competitive ratio. Still, we present the first online algorithm for scheduling parallel jobs with a constant competitive ratio in this context. We also prove a new information-theoretic lower bound of 2.25 for the competitive ratio of any deterministic online algorithm for this model. Moreover, we show that 6/5 is a lower bound for the competitive ratio of any deterministic online algorithm of the preemptive version of the model jobs arriving over time.     

An online parallel scheduling method with application to energy-efficiency in cloud computing

This paper considers online energy-efficient scheduling of virtual machines (VMs) for Cloud data centers. Each request is associated with a start-time, an end-time, a processing time and a capacity demand from a Physical Machine (PM). The goal is to schedule all of the requests non-preemptively in their start-time-end-time windows, subjecting to PM capacity constraints, such that the total busy time of all used PMs is minimized (called MinTBT-ON for abbreviation). This problem is a fundamental scheduling problem for parallel jobs allocation on multiple machines; it has important applications in power-aware scheduling in cloud computing, optical network design, customer service systems, and other related areas. Offline scheduling to minimize busy time is NP-hard already in the special case where all jobs have the same processing time and can be scheduled in a fixed time interval. One best-known result for MinTBT-ON problem is a g-competitive algorithm for general instances and unit-size jobs using First-Fit algorithm where g is the total capacity of a machine. In this paper, a $(1+\frac{g-2}{k}-\frac{g-1}{k^{2}})$ -competitive algorithm, Dynamic Bipartition-First-Fit (BFF) is proposed and proved for general case, where k is the ratio of the length of the longest interval over the length of the second longest interval for k>1 and g≥2. More results in general and special cases are obtained to improve the best-known bounds.     

Optimal and online preemptive scheduling on uniformly related machines

We consider the problem of preemptive scheduling on uniformly related machines. We present a semi-online algorithm which, if the optimal makespan is given in advance, produces an optimal schedule. Using the standard doubling technique, this yields a 4-competitive deterministic and an e≈2.71-competitive randomized online algorithm. In addition, it matches the performance of the previously known algorithms for the offline case, with a considerably simpler proof. Finally, we study the performance of greedy heuristics for the same problem.     

Online scheduling of bounded length jobs to maximize throughput

We consider an online scheduling problem, motivated by the issues present at the joints of networks using ATM and TCP/IP. Namely, IP packets have to be broken down into small ATM cells and sent out before their deadlines, but cells corresponding to different packets can be interwoven. More formally, we consider the online scheduling problem with preemptions, where each job j is revealed at release time r _j , and has processing time p _j , deadline?d _j , and weight w _j . A?preempted job can be resumed at any time. The goal is to maximize the total weight of all jobs completed on time. Our main results are as follows. Firstly, we prove that when the processing times of all jobs are at most k, the optimum deterministic competitive ratio is ??(k/log?k). Secondly, we give a deterministic algorithm with competitive ratio depending on the ratio between the smallest and the largest processing time of all jobs. In particular, it attains competitive ratio 5 in the case when all jobs have identical processing times, for which we give a lower bound of 2.598. The latter upper bound also yields an O(log?k)-competitive randomized algorithm for the variant with processing times bounded by k.     

Barely Random Algorithms for Multiprocessor Scheduling

We consider randomized algorithms for on-line scheduling on identical machines. For two machines, a randomized algorithm achieving a competitive ratio of was found by Bartal et al. (1995). Seiden has presented a randomized algorithm which achieves competitive ratios of 1.55665, 1.65888, 1.73376, 1.78295, and 1.81681, for m=3,4,5,6,7, respectively (Seiden, 2000). A barely random algorithm is one which is a distribution over a constant number of deterministic strategies. The algorithms of Bartal et al. and Seiden are not barely random–in fact, these algorithms potentially make a random choice for each job scheduled. We present the first barely random on-line scheduling algorithms. In addition, our algorithms use less space and time than the previous algorithms, asymptotically.     

Minimizing total weighted completion time with an unexpected machine unavailable interval

In the past two decades, scheduling with machine availability constraints has received considerable attention. Until recently most research has focused on the setting where all machine unavailability information is known at the beginning of the scheduling horizon. In reality, this is not practical in some cases. The machine may become unavailable to process jobs due to a machine breakdown or an occurrence of an emergent job that has to be processed immediately. In both cases, the start time of the unavailable interval is unknown beforehand, and the length of the interval may not be known until the end of the interval. In this article, we consider the situation in which the scheduler has to make scheduling decisions without any knowledge of the machine unavailable intervals. Of particular interest is the problem of minimizing total weighted completion time. When there are two or more unavailable intervals on a single machine, Fu et al. (2009) have shown that the problem is exponentially inapproximable even when jobs’ weights are equal to their processing times and one has full knowledge of unavailability. So in this paper we consider the scheduling problem on a single machine with a single unavailable period. And, we assume that every job has a weight proportional to its processing time. Based on whether the unavailable interval is due to a breakdown or an emergent job, we have the breakdown model and the emergent job model. First we show that no $\tfrac{\sqrt{5}+1}{2}$ -competitive online algorithm exists for the breakdown model, and no $\tfrac{11-\sqrt{2}}{7}$ -competitive online algorithm exists for the emergent job model. Next, we show that the simple LPT rule can give a 2- and a $\tfrac{9}{5}$ -competitive ratio for the breakdown model and the emergent job model, respectively. Further, we show that the ratios are tight by examples. For the offline case, we show that the First Fit LPT (FF-LPT) rule can give a tight approximation ratio of 2 and 4/3 for the breakdown model and the emergent job model, respectively. Finally, our experimental results show that, in practice, both LPT and FF-LPT perform very well and the performance improves when the number of jobs $n$ increases. In both models, when $n \ge 50$ , the worst case error ratio is much better than the theoretical bounds.     

Improved results for scheduling batched parallel jobs by using a generalized analysis framework

We present two improved results for scheduling batched parallel jobs on multiprocessors with mean response time as the performance metric. These results are obtained by using a generalized analysis framework where the response time of the jobs is expressed in two contributing factors that directly impact a scheduler’s competitive ratio. Specifically, we show that the scheduler IGDEQ is 3-competitive against the optimal while AGDEQ is 5.24-competitive. These results improve the known competitive ratios of 4 and 10, obtained by Deng et al. and by He et al., respectively. For the common case where no fractional allotments are allowed, we show that slightly larger competitive ratios can be obtained by augmenting the schedulers with the round-robin strategy.     

On-Line Edge-Coloring with a Fixed Number of Colors

We investigate a variant of on-line edge-coloring in which there is a fixed number of colors available and the aim is to color as many edges as possible. We prove upper and lower bounds on the performance of different classes of algorithms for the problem. Moreover, we determine the performance of two specific algorithms, First-Fit and Next-Fit . Specifically, algorithms that never reject edges that they are able to color are called fair algorithms. We consider the four combinations of fair/ not fair and deterministic/ randomized. We show that the competitive ratio of deterministic fair algorithms can vary only between approximately 0.4641 and 1/2, and that Next-Fit is worst possible among fair algorithms. Moreover, we show that no algorithm is better than 4/7-competitive. If the graphs are all k -colorable, any fair algorithm is at least 1/2-competitive. Again, this performance is matched by Next-Fit while the competitive ratio for First-Fit is shown to be k/(2k-1) , which is significantly better, as long as k is not too large.     

A better online algorithm for the parallel machine scheduling to minimize the total weighted completion time

The identical parallel machine scheduling problem with the objective of minimizing total weighted completion time is considered in the online setting where jobs arrive over time. An online algorithm is proposed and is proven to be (2.5–1/2m)-competitive based on the idea of instances reduction. Further computational experiments show the superiority over other algorithms in the average performance.     

Semi-online two-level supply chain scheduling problems

We consider two-level supply chain scheduling problems where customers release jobs to a manufacturer that has to process the jobs and deliver them to the customers. Processed jobs are grouped into batches, which are delivered to the customers as single shipments. The objective is to minimize the total cost which is the sum of the total flow time and the total delivery cost. Such problems have been considered in the off-line environment where future jobs are known, and in the online environment where at any time there is no information about future jobs. It is known that the best possible competitive ratio for an online algorithm is 2. We consider the problem in the semi-online environment, assuming that a lower bound P for all processing times is available a priori, and present a semi-online algorithm with competitive ratio \(\frac{2D}{D+P}\) where D is the cost of a delivery. Also, for the special case where all processing times are equal, we prove that the algorithm is \(1.045\sqrt{\frac{2-u}{u}}\)-competitive, where u is the density of the instance.     

A tight lower bound for job scheduling with cancellation

The Job Scheduling with Cancellation problem is a variation of classical scheduling problems in which jobs can be cancelled while waiting for execution. In this paper we prove a tight lower bound of 5 for the competitive ratio of any deterministic online algorithm for this problem, for the case where all jobs have the same processing time.     

Balanced Scheduling toward Loss-Free Packet Queuing and Delay Fairness

In current networks, packet losses can occur if routers do not provide sufficiently large buffers. This paper studies how many buffers should be provided in a router to eliminate packet losses. We assume a network router has m incoming queues, each corresponding to a single traffic stream, and must schedule at any time on-line from which queue to take the next packet to send out. To exclude packet losses with a small amount of buffers, the maximum queue length must be kept low over the entire scheduling period. We call this new on-line problem the balanced scheduling problem (BSP). By competitive analysis, we measure the power of on-line scheduling algorithms to prevent packet losses. We show that a simple greedy algorithm is Θ(log m)-competitive which is asymptotically optimal, while Round-Robin scheduling is not better than m-competitive, as actually is any deterministic on-line algorithm for BSP. We also give a polynomial time algorithm for solving off-line BSP optimally. We also study another on-line balancing problem that tries to balance the delay among the m traffic streams.     

On-Line Edge-Coloring with a Fixed Number of Colors

Abstract. We investigate a variant of on-line edge-coloring in which there is a fixed number of colors available and the aim is to color as many edges as possible. We prove upper and lower bounds on the performance of different classes of algorithms for the problem. Moreover, we determine the performance of two specific algorithms, First-Fit and Next-Fit . Specifically, algorithms that never reject edges that they are able to color are called fair algorithms. We consider the four combinations of fair/ not fair and deterministic/ randomized. We show that the competitive ratio of deterministic fair algorithms can vary only between approximately 0.4641 and 1/2, and that Next-Fit is worst possible among fair algorithms. Moreover, we show that no algorithm is better than 4/7-competitive. If the graphs are all k -colorable, any fair algorithm is at least 1/2-competitive. Again, this performance is matched by Next-Fit while the competitive ratio for First-Fit is shown to be k/(2k-1) , which is significantly better, as long as k is not too large.     

Balanced Scheduling toward Loss-Free Packet Queuing and Delay Fairness

In current networks, packet losses can occur if routers do not provide sufficiently large buffers. This paper studies how many buffers should be provided in a router to eliminate packet losses. We assume a network router has m incoming queues, each corresponding to a single traffic stream, and must schedule at any time on-line from which queue to take the next packet to send out. To exclude packet losses with a small amount of buffers, the maximum queue length must be kept low over the entire scheduling period. We call this new on-line problem the balanced scheduling problem (BSP). By competitive analysis, we measure the power of on-line scheduling algorithms to prevent packet losses. We show that a simple greedy algorithm is (log m)-competitive which is asymptotically optimal, while Round-Robin scheduling is not better than m-competitive, as actually is any deterministic on-line algorithm for BSP. We also give a polynomial time algorithm for solving off-line BSP optimally. We also study another on-line balancing problem that tries to balance the delay among the m traffic streams.     

Online Scheduling with Interval Conflicts

In the problem of Scheduling with Interval Conflicts, there is a ground set of items indexed by integers, and the input is a collection of conflicts, each containing all the items whose index lies within some interval on the real line. Conflicts arrive in an online fashion. A scheduling algorithm must select, from each conflict, at most one survivor item, and the goal is to maximize the number (or weight) of items that survive all the conflicts they are involved in. We present a centralized deterministic online algorithm whose competitive ratio is O(lgσ), where σ is the size of the largest conflict. For the distributed setting, we present another deterministic algorithm whose competitive ratio is $2\left \lceil {\lg\sigma} \right \rceil $ in the special contiguous case, in which the item indices constitute a contiguous interval of integers. Our upper bounds are complemented by two lower bounds: one that shows that even in the contiguous case, all deterministic algorithms (centralized or distributed) have competitive ratio Ω(lgσ), and that in the non-contiguous case, no deterministic oblivious algorithm (i.e., a distributed algorithm that does not use communication) can have a bounded competitive ratio.     

Scheduling with Conflicts on Bipartite and Interval Graphs

In this paper, we consider the on-line scheduling of jobs that may be competing for mutually exclusive resources. We model the conflicts between jobs with a conflict graph, so that the set of all concurrently running jobs must form an independent set in the graph. This model is natural and general enough to have applications in a variety of settings; however, we are motivated by the following two specific applications: traffic intersection control and session scheduling in high speed local area networks with spatial reuse. Our results focus on two special classes of graphs motivated by our applications: bipartite graphs and interval graphs. The cost function we use is maximum response time. In all of the upper bounds, we devise algorithms which maintain a set of invariants which bound the accumulation of jobs on cliques (in the case of bipartite graphs, edges) in the graph. The lower bounds show that the invariants maintained by the algorithms are tight to within a constant factor. For a specific graph which arises in the traffic intersection control problem, we show a simple algorithm which achieves the optimal competitive ratio.