Table of Contents
Fetching ...

QACM: QoS-Aware xApp Conflict Mitigation in Open RAN

Abdul Wadud, Fatemeh Golpayegani, Nima Afraz

TL;DR

This work addresses intra-component conflicts among xApps in Open RAN's Near-RT-RIC by introducing QoS-Aware Conflict Mitigation (QACM). It formulates QACM as a constrained optimization within a Conflict Management System (CMS) and uses KPI-to-utility conversion via z-score with KPI prediction to balance competing QoS requirements. Compared to state-of-the-art methods NSWF and EG, QACM demonstrates improved QoS satisfaction for more xApps under both priority and non-priority scenarios, supported by case studies and simulations. The approach offers a scalable, vendor-agnostic framework for conflict resolution in Open RAN, with plans to enhance KPI prediction and real-world validation in future work.

Abstract

The advent of Open Radio Access Network (RAN) has revolutionized the field of RAN by introducing elements of native support of intelligence and openness into the next generation of mobile network infrastructure. Open RAN paves the way for standardized interfaces and enables the integration of network applications from diverse vendors, thereby enhancing network management flexibility. However, control decision conflicts occur when components from different vendors are deployed together. This article provides an overview of various types of conflicts that may occur in Open RAN, with a particular focus on intra-component conflict mitigation among Extended Applications (xApps) in the Near Real Time RAN Intelligent Controller (Near-RT-RIC). A QoS-Aware Conflict Mitigation (QACM) method is proposed that finds the optimal configuration of conflicting parameters while maximizing the number of xApps that have their Quality of Service (QoS) requirements met. We compare the performance of the proposed QACM method with two benchmark methods for priority and non-priority cases. The results indicate that our proposed method is the most effective in maintaining QoS requirements for conflicting xApps.

QACM: QoS-Aware xApp Conflict Mitigation in Open RAN

TL;DR

This work addresses intra-component conflicts among xApps in Open RAN's Near-RT-RIC by introducing QoS-Aware Conflict Mitigation (QACM). It formulates QACM as a constrained optimization within a Conflict Management System (CMS) and uses KPI-to-utility conversion via z-score with KPI prediction to balance competing QoS requirements. Compared to state-of-the-art methods NSWF and EG, QACM demonstrates improved QoS satisfaction for more xApps under both priority and non-priority scenarios, supported by case studies and simulations. The approach offers a scalable, vendor-agnostic framework for conflict resolution in Open RAN, with plans to enhance KPI prediction and real-world validation in future work.

Abstract

The advent of Open Radio Access Network (RAN) has revolutionized the field of RAN by introducing elements of native support of intelligence and openness into the next generation of mobile network infrastructure. Open RAN paves the way for standardized interfaces and enables the integration of network applications from diverse vendors, thereby enhancing network management flexibility. However, control decision conflicts occur when components from different vendors are deployed together. This article provides an overview of various types of conflicts that may occur in Open RAN, with a particular focus on intra-component conflict mitigation among Extended Applications (xApps) in the Near Real Time RAN Intelligent Controller (Near-RT-RIC). A QoS-Aware Conflict Mitigation (QACM) method is proposed that finds the optimal configuration of conflicting parameters while maximizing the number of xApps that have their Quality of Service (QoS) requirements met. We compare the performance of the proposed QACM method with two benchmark methods for priority and non-priority cases. The results indicate that our proposed method is the most effective in maintaining QoS requirements for conflicting xApps.
Paper Structure (37 sections, 3 equations, 12 figures, 5 tables, 1 algorithm)

This paper contains 37 sections, 3 equations, 12 figures, 5 tables, 1 algorithm.

Figures (12)

  • Figure 1: Taxonomy of potential conflicts in Open RAN wadud2023conflict.
  • Figure 2: Potential conflicting areas in Open RAN wadud2023conflict.
  • Figure 3: Example model for direct, indirect and implicit conflict.
  • Figure 4: The CMS Framework in the Near-RT-RIC.
  • Figure 5: Control Loop for the CMS.
  • ...and 7 more figures