Decision-Theoretic Safety Assessment of Persona-Driven Multi-Agent Systems in O-RAN
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866908953923813376 |
|---|---|
| author | Nezami, Zeinab Zaidi, Syed Ali Raza Hafeez, Maryam Powell, Louis Talari, Vara Prasad Tatipamula, Mallik |
| author_facet | Nezami, Zeinab Zaidi, Syed Ali Raza Hafeez, Maryam Powell, Louis Talari, Vara Prasad Tatipamula, Mallik |
| contents | Autonomous network management in Open Radio Access Networks requires intelligent decision making across conflicting objectives, yet existing LLM based multi agent systems employ homogeneous strategies and lack systematic predeployment validation. We introduce a persona driven multi agent framework where configurable behavioral personas structured specifications encoding optimization priorities, risk tolerance, and decision making style influence five specialized agents (planning, coordination, resource allocation, code generation, analysis). To enable rigorous validation, we develop a three dimensional evaluation framework grounded in decision theory, measuring normative compliance (optimality adherence), prescriptive alignment (behavioral guideline consistency), and behavioral dynamics (emergent system properties). We evaluate 486 persona configurations across two ORAN optimization challenges (energy efficient resource allocation and network load balancing). Results demonstrate that persona agent alignment significantly impacts both individual performance (14.3 percent) and emergent multi agent coordination, with retrieval architecture (GraphRAG vs. RAG) fundamentally constraining customization effectiveness. Single agent persona modifications propagate system wide through cascading effects, with certain combinations exhibiting detectable fundamental incompatibilities. Our framework provides systematic validation mechanisms for deploying LLM based automation in mission critical telecommunications infrastructure. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_09682 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Decision-Theoretic Safety Assessment of Persona-Driven Multi-Agent Systems in O-RAN Nezami, Zeinab Zaidi, Syed Ali Raza Hafeez, Maryam Powell, Louis Talari, Vara Prasad Tatipamula, Mallik Networking and Internet Architecture Artificial Intelligence Autonomous network management in Open Radio Access Networks requires intelligent decision making across conflicting objectives, yet existing LLM based multi agent systems employ homogeneous strategies and lack systematic predeployment validation. We introduce a persona driven multi agent framework where configurable behavioral personas structured specifications encoding optimization priorities, risk tolerance, and decision making style influence five specialized agents (planning, coordination, resource allocation, code generation, analysis). To enable rigorous validation, we develop a three dimensional evaluation framework grounded in decision theory, measuring normative compliance (optimality adherence), prescriptive alignment (behavioral guideline consistency), and behavioral dynamics (emergent system properties). We evaluate 486 persona configurations across two ORAN optimization challenges (energy efficient resource allocation and network load balancing). Results demonstrate that persona agent alignment significantly impacts both individual performance (14.3 percent) and emergent multi agent coordination, with retrieval architecture (GraphRAG vs. RAG) fundamentally constraining customization effectiveness. Single agent persona modifications propagate system wide through cascading effects, with certain combinations exhibiting detectable fundamental incompatibilities. Our framework provides systematic validation mechanisms for deploying LLM based automation in mission critical telecommunications infrastructure. |
| title | Decision-Theoretic Safety Assessment of Persona-Driven Multi-Agent Systems in O-RAN |
| topic | Networking and Internet Architecture Artificial Intelligence |
| url | https://arxiv.org/abs/2604.09682 |