边缘计算任务处理时延数据集

本数据集记录了所设计的边缘计算资源分配算法在算力受限的硬件平台中处理图像识别任务的时延，用于计算不同强度的随机计算负载下图像任务执行的超时概率。通过对比现有贪婪算法和固定资源管理算法，验证所提出的边缘计算资源管理算法能够更充分地利用平台算力、降低计算时延,达到了降低任务超时数75%的效果。

This dataset records the execution latency of the proposed edge computing resource allocation algorithm when processing image recognition tasks on a hardware platform with constrained computing power, which is used to calculate the timeout probability of image task execution under random computing loads of varying intensities. By comparing with existing greedy algorithms and fixed resource management algorithms, it is verified that the proposed edge computing resource management algorithm can make more efficient use of the platform's computing power and reduce computing latency, achieving a 75% reduction in the number of task timeouts.

搜集汇总

背景与挑战

背景概述

该数据集包含边缘计算资源分配算法在受限硬件平台处理图像识别任务的时延数据，通过对比现有算法验证了新算法能降低75%任务超时数。数据集由深圳大学创建，属于国家重点研发计划项目成果。