基于多智能体深度强化学习的分布式任务与计算资源联合分配数据

数据内容：此数据集针对分布式边缘计算场景下的任务分配，旨在联合优化分布式任务与计算资源分配，以最大化系统中用户体验质量之和。该数据集中包含多个边缘服务器资源和性能特征数据，以评估基于多智能体深度强化学习分布式计算任务分配算法的效果。具体内容包括：（a）多边缘服务器资源拓扑及资源特征：7个边缘服务器计算频率、两两边缘服务器间通信速率和通信路径传播时延;（b)仿真相关参数配置：用于仿真的随机任务特征生成范围，随机因子;（b）任务分配优化：通过优化算法进行任务与资源分配优化结果，包含用户总QoE，用户平均服务时延，以及用户在进行任务分配后任务被分配到不同位置用户的分布。数据集内含1个数据集说明文件，5个支撑数据文件。
数据来源：该数据集通过Python软件进行仿真生成。模拟了不同用户数量情况下，基于多智能体强化学习进行任务分配优化的训练过程。
时间及地点：采集地点为之江实验室，采集时间为2022年10月至2023年10月。
设备情况：实验中使用了个人计算机、远程服务器等设备。

Data Content: This dataset focuses on task allocation in distributed edge computing scenarios, aiming to jointly optimize distributed tasks and computing resource allocation to maximize the total sum of user quality of experience (QoE) in the system. It includes multiple edge server resource and performance characteristic datasets for evaluating the effectiveness of multi-agent deep reinforcement learning-based distributed computing task allocation algorithms. Specific contents are as follows:
(a) Multi-edge server resource topology and resource characteristics: computing frequencies of 7 edge servers, communication rates between every pair of edge servers, and propagation delays of communication paths;
(b) Simulation-related parameter configurations: ranges for generating random task features during simulation, and random factors;
(b) Task allocation optimization: optimization results of task and resource allocation obtained via optimization algorithms, including total user QoE, average user service latency, and the distribution of tasks allocated to users at different locations after task allocation.
The dataset contains 1 dataset description file and 5 supporting data files.

Data Source: This dataset is generated through Python-based software simulation, which simulates the training process of task allocation optimization based on multi-agent reinforcement learning under different user counts.

Time and Location: The data was collected at Zhijiang Laboratory from October 2022 to October 2023.

Equipment: Personal computers, remote servers and other equipment were used in the experiments.

搜集汇总

背景与挑战

背景概述

该数据集针对分布式边缘计算场景，通过多智能体深度强化学习优化任务与计算资源的联合分配，以提升用户体验质量。它包含边缘服务器资源特征、仿真参数及优化结果等数据，由Python仿真生成，采集于2022年至2023年间的之江实验室。