DPLB

AcronymDefinition
DPLBDécouverte Pédestre Lavandou Bormes (French walking club)
DPLBDésolé pour le Bruit (French: Sorry for the Noise)
DPLBData-Parallel Load-Balancing
References in periodicals archive ?
In order to evaluate the performance of DPLB, four assessment indicators are defined in the experiments as follows:
Under the same experimental condition and environment, we implement the Min-Min algorithm, Max-Min algorithm, WRR algorithm, GAs algorithm, HBB-LB algorithm and DPLB algorithm in the CloudSim platform, and compare their performance in four indicators which are defined in section 4.2.
Part I Experiments with smooth tasks set, and Comparing Min-Min algorithm, Max-Min algorithm, WRR[18] algorithm, GAs[12] algorithm, HBB-LB[11] algorithm and DPLB algorithm.
5, we can see that DPLB algorithm keeps a lower value, it is more fit to a large number of tasks, it becomes the lowest since 150 tasks, it is obvious better than the Min-Min algorithm and WRR algorithm which are prone to load imbalance, and it also better than Max-Min, GAs algorithm and HBB-LB algorithm, it proves that the DPLB algorithm can effectively reduce the overall task completion time.
6, we can see that the DPLB algorithm keeps a lower value, it becomes the lowest since 150 tasks, it proves that the DPLB algorithm has a good load balancing characteristics.
7, we can see that with the increasing number of tasks, the success completion ratio shows a downward trend, and DPLB algorithm keeps the highest success completion ratio all the time, namely the DPLB algorithm can ensure more tasks to be completed before deadline, it proves that the design of task's dynamic priority which considers the urgency of task execution has a good influence on DPLB algorithm, it insures that the task which has the shorter deadline can be first scheduled and completed.
8 we can see that with the increasing number of tasks, the DPLB algorithm keeps the highest value since 150 tasks, namely the DPLB algorithm can bring more benefits per second.
Part II Experiments with recomputed task sets, and comparing dynamic balancing algorithm which includes GAs algorithm, HBB-LB algorithm, DPLB algorithm.
In order to verify the dynamic load balancing performance of DPLB algorithm, we design a new task sets based on Part I 450 task set.
9, we can see that with the increasing number of tasks, the DPLB algorithm keeps the lowest value in Makespan and load balancing degree, keeps the highest value in success completion ratio and service profit per second, it proves that the DPLB algorithm also has a better performance in dynamic environment, it can effectively reduce the overall task completion time (Makespan) and load balancing degree, and effectively advance the success completion ratio and service profit per second.
Does every [C.sub.0](L) space, L being a locally compact Hausdorff space, solve the dplb problem?