Dynamic scheduling has been always a challenging problem for real-time distributed systems. EDF (Earliest Deadline First) algorithm has been proved to be optimal scheduling algorithm for single processor real-time sys...
详细信息
Deficit Round Robin (DRR) is a scheduling algorithm which provides fair queuing at O(1) complexity. However, due to its round robin structure, its latency properties are not adequate for latency-critical applications,...
详细信息
ISBN:
(纸本)1595935045
Deficit Round Robin (DRR) is a scheduling algorithm which provides fair queuing at O(1) complexity. However, due to its round robin structure, its latency properties are not adequate for latency-critical applications, such as voice. For this reason, router manufacturers implement variants of the DRR algorithm which guarantee lower latencies to one (or a subset of) queue(s). In this paper we evaluate the performance of two such variants, both of which are known as Modified Deficit Round Robin, currently implemented in commercial routers. The comparison is carried out analytically, by deriving the latency and bandwidth sharing properties of both algorithms, and by simulation. Copyright 2006 ACM.
Among modern cloud infrastructures, live migration of virtual machines offers many advantages like scalability and elasticity but also leads to risks in the meantime. Security issues of live migration have been studie...
详细信息
In this paper we analyze achievable throughput guarantees for different opportunistic scheduling algorithms operating in wireless time-division multiplexing networks. We consider a scenario where the average carrier-t...
详细信息
For great change of service time for request, big difference of hardware and software server and different network performance, this paper proposes a dynamic-feedback algorithm based on AHP in the course of studying t...
详细信息
We consider a model where multiple queues are served by a server whose capacity varies randomly and asynchronously with respect to different queues. The problem is to optimally control large deviations of the queues i...
详细信息
ISBN:
(纸本)9781604237924
We consider a model where multiple queues are served by a server whose capacity varies randomly and asynchronously with respect to different queues. The problem is to optimally control large deviations of the queues in the following sense: find a scheduling rule maximizing min i h lim n!1 1 n log P (aiQi n) i , (1) where Qi is the length of i-th queue in a stationary regime, and ai 0 are parameters. Thus, we seek to maximize the minimum of the exponential decay rates of the tails of distributions of weighted queue lengths aiQi. We give a characterization of the upper bound on (1) under any scheduling rule, and of the lower bound on (1) under the exponential (EXP) rule. For the case of two queues, we prove that the two bounds match, thus proving optimality of EXP rule in this case. The EXP rule is not asymptotically invariant with respect to scaling of the queues, which complicates its analysis in large deviations regime. To overcome this, we introduce and prove a refined sample path large deviations principle, or refined Mogulsky theorem, which is of independent interest.
A task graph set is proposed for the evaluation of reconfigurable environment scheduling algorithms. This makes it possible to evaluate scheduling algorithms under the similar conditions. This paper provides an overvi...
详细信息
Computers are taking up great part in technology evolution. They perform many tasks quickly and reliably when compared to humans, opening new avenues. One important aspect of computers is multitasking that enables a s...
详细信息
scheduling algorithms play a key role in overall system performance of broadband wireless systems (BWS) such as WLAN/WMAN. Maximal SNR (MaxSNR) and Round Robin (RR) are two conventional scheduling strategies which emp...
详细信息
Traditional real-time scheduling algorithms are developed employing a standard framework to satisfy the requirements for a rapid execution time, high performance, and low power consumption. However, real-time control ...
详细信息
暂无评论