2025 Vol.8 Oct.N05

• Title: Resource Optimization Allocation Method for Heterogeneous Experimental Tasks Using MAS-RL
• Name: Xiaojun Cheng
• Company: Jiangsu Union Technical Institute，Yancheng Mechanical and Electrical Branch，Yancheng，224005，China
• Abstract:

   Resource allocation methods for heterogeneous experimental tasks predominantly rely on static rules or single-objective optimization algorithms, matching tasks to hardware through predefined priority queues or greedy strategies. However, the lack of comprehensive consideration for multidimensional resource constraints and dynamic load variations results in suboptimal allocation balance. To address this, we propose a MAS-RL-based resource optimization allocation method for heterogeneous experimental tasks. A multidimensional state vector incorporating queue depth and memory utilization is constructed to quantify hardware real-time load and efficiency based on heterogeneous devices' computational characteristics. By concatenating device states with task demands and incorporating temporal step information, a global state vector is formed. Defining a distributed action space enables each agent to independently make device allocation decisions. A hybrid reward function is designed to integrate three objectives: task delay, system throughput, and energy consumption. Training the optimization strategy achieves a weighted balance between efficiency and cost. Attention weights are introduced to calculate the importance of other agents, enabling dynamic coordination and communication among agents. Experiments validate the proposed method's allocation balance. Comparative test results clearly demonstrate that when applying the proposed method to allocate resources for heterogeneous experimental tasks, the average resource utilization stabilizes at 89.2%, achieving an ideal allocation effect.

  
  

• Keyword: Heterogeneous experimental tasks; MAS-RL; Resource optimization allocation; Allocation fairness;
• DOI: 10.12250/jpciams2025091018
• Citation form: Xiaojun Cheng.Resource Optimization Allocation Method for Heterogeneous Experimental Tasks Using MAS-RL[J]. Computer Informatization and Mechanical System,2025,Vol.8,pp.

[1] Lin X, Fan Y, Zhang L, et al. Resource Allocation for RIS-Aided mmWave System With Cooperative and Non-Cooperative Base Stations[J]. IEEE Transactions on Vehicular Technology, 2025, 74(3):4419-4431.

[2] Shi J, Li Z, Liao X. Deep reinforcement learning-aided resource allocation in ultradense networks. Resource Optimization in Wireless Communications, 2025:241-257.

[3] An N, Yang F, Liu C, et al. Resource Allocation for Fairness Enhancement in Multi-Cell Vehicular VLC System With Optical IRS: A Cooperative Transmission Approach[J]. IEEE Internet of Things Journal, 2025:1-1.

[4] Liakath J A, Natesan G, Krishnadoss P, et al. Efficient resource allocation in heterogeneous clouds: genetic water evaporation optimization for task scheduling[J]. Signal, Image and Video Processing, 2024, 18(5):3993-4002.

[5] Feng L, Jiang X, Sun Y, et al. Resource Allocation for Metaverse Experience Optimization: A Multi-Objective Multi-Agent Evolutionary Reinforcement Learning Approach[J]. IEEE Transactions on Mobile Computing, 2025, 24(4):3473-3488.

[6] Aldrees A, Min H, Daradkeh Y I, et al. Optimization of 6G resource allocation using CyberTwin function-based service enhancement scheme[J]. EURASIP Journal on Wireless Communications and Networking, 2025, 2025(1):1-27.

[7] Cai J, Yin C, Ding Y. Optimization of resource allocation in FDD massive MIMO systems[J]. Digital Communications and Networks, 2024, 10(1):117-125.

[8] Papar C M, Schirliu T H. Logistics Optimization for Resource Allocation and Scheduling Using Time Slots[J]. International Journal of Advanced Statistics and IT&C for Economics and Life Sciences, 2024, 14(1):131-146.

[9] Zhang Z, Xu C, Xu S, et al. Towards optimized scheduling and allocation of heterogeneous resources via graph-enhanced EPSO algorithm [J]. Journal of Cloud Computing, 2024, 13(1):1-23.

[10] Rabaaoui S, Héla Hachicha, Zagrouba E. An Efficient and Autonomous Dynamic Resource Allocation in Cloud Computing with Optimized Task Scheduling [J]. Procedia Computer Science, 2024, 246(000):3654-3663.

[11] Mishra R, Gupta M. MHDORA-LBA: Dynamic and Optimized Resource-Aware Load Balancing Approach for Resource Allocation[J]. SN Computer Science, 2024, 5(6):1-20.

[12] Zhao L, Feng Y, Hawbani A, et al. Optimized Resource Allocation in Vehicle Edge Computing through Platoon Collaboration[J]. IEEE Internet of Things Journal, 2025:1-1.

[13] Rangaiah Y P, Yaragudri H, Rani K P, et al. Blockchain and Machine Learning for Optimized Resource Allocation in Cloud Computing Services[J]. 2024 7th International Conference on Contemporary Computing and Informatics (IC3I), 2024:1495-1501.

[14] Zhang G, Mingwei X U. Push and Pull Tor Users' Guards Through Optimized Resource Portfolios[J]. Journal of Tsinghua University (Science and Technology), 2024, 64(8):1293-1305.

[15] Anjaneyulu P, Priyanka D, Prakash K J, et al. Assessment on Significant SVM and MLP-Based Optimized Resource Allocation for Load Balancing [J]. Communications in Computer and Information Science, 2025:402-412.