International Peer-Reviewed JournalOpen AccessISSN 2456-8880
irejournals@gmail.com+91-7433024337

Home / Current Issue / Paper 1719075

1719075PublishedVol 9 · Issue 12

Cooperative Multi-Agent Negotiation for Closed-Loop Database Tuning and Index Optimization

Shahid Moosa

Subject area: Science,Engineering and Technology  ·  Area of research: Cloud Database

DOI: https://doi.org/10.64388/IREV9I12-1719075

Abstract

Modern database systems operate under increasingly dynamic workloads that render static configuration and manual tuning obsolete. Closed-loop database tuning, where the system automatically observes the performance metrics, diagnoses the bottlenecks and applies corrective actions, has emerged as a promising paradigm for self-managing databases. But the tuning problem is inherently multi-objective and multi-component: knob configuration, index selection, query rewriting and materialized view maintenance each need specialized optimization strategies which can conflict with each other. In this paper, we propose a cooperative multi-agent negotiation framework in which autonomous agents are responsible for a specific tuning dimension and engage in structured negotiation protocols to reach collectively optimal decisions. We formalize the negotiation as a constrained multi-objective optimization problem, design a coalition formation mechanism based on cooperative game theory, and evaluate the framework on the TPC-H and TPC-DS benchmarks against state-of-the-art single-agent and heuristic baselines. Experimental results show that the proposed framework improves the aggregate throughput by 18-34% and the tail latency by 22-41% over existing approaches while being stable against the workload phase transitions. The paper presents a principled theoretical basis for multi-agent database tuning, and a practical architecture that can be deployed in production environments.

Keywords

Multi-Agent Systems, Database Tuning, Index Optimization, Closed Loop Control, Cooperative Game Theory, Self Managing Databases, Query Optimization

How to cite this paper

Shahid Moosa "Cooperative Multi-Agent Negotiation for Closed-Loop Database Tuning and Index Optimization" Iconic Research And Engineering Journals Volume 9 Issue 12 2026 Page 2120-2127 https://doi.org/10.64388/IREV9I12-1719075
Shahid Moosa "Cooperative Multi-Agent Negotiation for Closed-Loop Database Tuning and Index Optimization" Iconic Research And Engineering Journals, vol. 9, no. 12, Jun. 2026, doi: https://doi.org/10.64388/IREV9I12-1719075
Shahid Moosa (2026). Cooperative Multi-Agent Negotiation for Closed-Loop Database Tuning and Index Optimization. Iconic Research And Engineering Journals, 9(12). doi: https://doi.org/10.64388/IREV9I12-1719075
Shahid Moosa "Cooperative Multi-Agent Negotiation for Closed-Loop Database Tuning and Index Optimization" Iconic Research And Engineering Journals, vol. 9, no. 12, Jun. 2026. Crossref, https://doi.org/10.64388/IREV9I12-1719075
@article{1719075,
      author = {Shahid Moosa},
      title = {Cooperative Multi-Agent Negotiation for Closed-Loop Database Tuning and Index Optimization},
      journal = {Iconic Research And Engineering Journals},
      year = {2026},
      volume = {9},
      number = {12},
      pages = {2120-2127},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1719075.pdf},
      abstract = {Modern database systems operate under increasingly dynamic workloads that render static configuration and manual tuning obsolete. Closed-loop database tuning, where the system automatically observes the performance metrics, diagnoses the bottlenecks and applies corrective actions, has emerged as a promising paradigm for self-managing databases. But the tuning problem is inherently multi-objective and multi-component: knob configuration, index selection, query rewriting and materialized view maintenance each need specialized optimization strategies which can conflict with each other. In this paper, we propose a cooperative multi-agent negotiation framework in which autonomous agents are responsible for a specific tuning dimension and engage in structured negotiation protocols to reach collectively optimal decisions. We formalize the negotiation as a constrained multi-objective optimization problem, design a coalition formation mechanism based on cooperative game theory, and evaluate the framework on the TPC-H and TPC-DS benchmarks against state-of-the-art single-agent and heuristic baselines. Experimental results show that the proposed framework improves the aggregate throughput by 18-34% and the tail latency by 22-41% over existing approaches while being stable against the workload phase transitions. The paper presents a principled theoretical basis for multi-agent database tuning, and a practical architecture that can be deployed in production environments.},
      keywords = {Multi-Agent Systems, Database Tuning, Index Optimization, Closed Loop Control, Cooperative Game Theory, Self Managing Databases, Query Optimization},
      month = {June},
      doi = {https://doi.org/10.64388/IREV9I12-1719075}
  }