Task allocation algorithm and optimization model on edge collaboration

Xiaoheng Deng,Jun Li,Enlu Liu,Honggang Zhang

DOI: https://doi.org/10.1016/j.sysarc.2020.101778

IF: 5.836

2020-11-01

Journal of Systems Architecture

Abstract:<p>This paper investigates a mobile edge computing environment for video analysis tasks where edge nodes provide their computation capacities to process the computation intensive tasks submitted by end users. First, we introduce a Cloudlet Assisted Cooperative Task Assignment (CACTA) system that organizes edge nodes that are geographically close to an end user into a cluster to collaboratively work on the user's tasks. It is challenging for the system to find an optimal strategy that assigns workload to edge nodes to meet the user's optimization goal. To address the challenge, this paper proposes multiple algorithms for different situations. Firstly, considering the situation that historical data cannot be obtained, a multi-round allocation algorithm based on EMA prediction is proposed, and the experimental results prove the efficiency and necessity of multiple rounds of transmission. To address the second case of obtaining historical data, this paper introduces a prediction-based dynamic task assignment algorithm that assigns workload to edge nodes in each time slot based on the prediction of their capacities/costs and an empirical optimal allocation strategy which is learned from an offline optimal solution from historical data. Experimental results demonstrate that our proposed algorithm achieves significantly higher performance than several other algorithms, and especially its performance is very close to that of an offline optimal solution. Finally, we propose an online task assignment algorithm based on Q-learning, which uses the model-free Q-learning algorithm to actively learn the allocation strategy of the system, and the experimental results verify the superiority and effectiveness of this algorithm.</p>

computer science, software engineering, hardware & architecture

What problem does this paper attempt to address?

### Problems Addressed by the Paper

This paper primarily explores how to optimize workload distribution for video analysis tasks in a mobile edge computing environment. Specifically, the paper proposes a system called "Cloudlet Assisted Cooperative Task Assignment (CACTA)" and designs several different algorithms to tackle workload distribution challenges under various conditions.

#### Main Issues:
1. **Workload Distribution Strategy**: How to find the optimal workload distribution strategy on highly heterogeneous, dynamic, and mobile edge nodes to meet user optimization goals (such as minimizing task completion time or jointly minimizing total task completion time and system cost).
2. **Utilization of Historical Data**: When the system cannot obtain long-term historical data, how to achieve efficient task distribution through multi-round allocation algorithms (based on EMA prediction); when historical data is available, how to use this data for prediction and optimize task distribution.
3. **Real-time Resource Scheduling**: Proposes an online task allocation algorithm based on Q-learning, achieving real-time load balancing and joint optimization of resources through classical reinforcement learning methods.

#### Research Background:
With the development of IoT technology, edge computing has gained widespread attention due to its ability to provide low latency and high computing power. However, the high heterogeneity, mobility, and dynamic nature of edge nodes make designing effective task distribution strategies extremely challenging. Therefore, the paper aims to address this issue and improve task processing efficiency in edge computing environments.