Large-scale events like Hajj require impeccable network performance and seamless connectivity to ensure safety, communication, and smooth operations for millions of attendees.  

Mobily and Ericsson, leveraging the advanced Cognitive Network Solution’s AI-driven software, delivered an unprecedented Level 4 (L4) network management automation (as defined by TM Forum) and improved optimization efficiency during Hajj 2025. These algorithms, also currently offered as rApps, have proven to be key in automating complex optimization processes and enhancing end-user experience.  

Managing the connectivity demands during Hajj presents unique and complex challenges, such as:

• Handling massive, fluctuating network traffic (uploads) in a geographically concentrated environment
• Detecting and mitigating uplink interference issues (both 4G and 5G)
• Rapidly identifying network capacity bottlenecks, such as IBS (in-building solutions) anomalies and coverage gaps
• Ensuring near real-time responses to failures or network performance degradation during the event, when minutes matter
• Coordinating distributed and centralized AI frameworks to harmonize decision-making and automation across multiple sites

Recognizing that traditional manual operations, assisted monitoring models, and the legacy Self Organizing Network (SON) solutions were inadequate for the immense demands and critical nature of Hajj, Mobily and Ericsson saw the need to advance toward Level 4 autonomy, a high degree of automation with closed-loop AI-driven controls.

As network complexity grows and performance expectations rise, only AI-based, closed-loop use cases that automatically adapt to evolving network conditions can ensure seamless operations during such a large-scale event.

The project brought together a set of tightly integrated AI-driven capabilities to detect, diagnose, and automatically remediate RAN issues at Hajj scale.

Event insights and real-time proactive optimization: This was powered by Markaz, a custom platform developed for real time event monitoring and optimization. It delivers powerful user experience insights with interactive dashboards, detailed user-movement heatmaps reports, and instant visibility into failure causes. The solution was integrated seamlessly with Ericsson Network Management (ENM), delivering measurement resolution of just 1-3 minutes, and over 50 new KPIs for live failure insight. Dynamic load balancing and coverage-gap mitigation were achieved through coordinated centralized and distributed AI agents across cells, enabling instant CGM (Coverage Gap Mitigator) adjustments for near real-time network optimization implementing 598 automated corrective actions across 116 sites. In addition, optimization activities were performed in specialized indoor designs through IBS (In-Building Solutions) Issues Detector.

AI powered 5G and 4G Uplink Interference Optimizers: These optimizers, part of the Ericsson Cognitive Network Solutions (CNS) portfolio, powered advanced issue detection, root-cause analysis, and proposed actions. Customized root-cause analysis (RCA) was enabled by CNS Software API extensions and on-demand RCA prioritization with spatial dashboards. Closed-loop, zero-touch automation delivered continuous parameter recommendations and auto-tuning, such as uplink power control.

Together, these capabilities delivered fully automated detection, diagnosis, and corrective action across LTE and 5G layers, achieving zero-touch operations and continuous network optimization for the Hajj event.

Event management

Mobily used its event management experience to implement a customized AI-powered software suite for Hajj 2025. This included:

• Near real-time dashboard: delivering live performance insights with minute resolution, enabling rapid detection of issues

• Deployment of over 50 new Key Performance Indicators (KPIs) and counters to instantly identify failures and performance drops

• Customized APIs for Root-Cause Analysis (RCA), enabling proactive, data-driven decisions

• Handling over 23,000 AI-driven decisions per hour and executing more than 15,000 automated corrective actions before the event, guaranteeing network resilience

• Seamless harmonization of centralized and distributed AI agents to dynamically manage resources and optimize network loads

As a result of these innovations, Hajj 2025 achieved what is believed to be a first-of-its-kind large-scale autonomous network operation with instant, closed-loop optimization capabilities.

Network optimization AI software

The backbone of this success was the sophisticated AI-powered network optimization software stack. In this scenario, different algorithms were used:

• Cognitive Network Solutions AI-based software. Issue detection, classification, root-cause analysis, and proposed changes are fully automated within closed-loop implementation cycle, enabling zero-touch operations.

• 5G and 4G Uplink Interference Optimizers: Reduced interference by improving SINR by approximately 15%, enabling a roughly 9% increase in higher-order modulation efficiency

• Explainable AI (XAI): Provided transparency and trust in AI decisions during the mega event, an essential factor for operator confidence

• IBS issues detector: Continuously identified in-building coverage anomalies.

• Advanced dynamic load balancing through network functionalities: Traffic distribution dynamically adapted to avoid congestion and improve overall network throughput.

Intelligent closed-loop coverage gap mitigator: Addressed and closed coverage holes proactively to maintain seamless connectivity in a closed-loop setup: Executed over 598 actions on 116 sites with a 1-minute decision resolution time, ensuring near real-time automated network tuning.

This comprehensive AI-driven diagnostics and proactive remediation ecosystem transformed network operations from reactive firefighting to predictive, autonomous management.

Hajj 2025 showcased the transformative potential of Cognitive Network Solutions’ AI-powered applications for network management in event settings, highlighted by Mobily and Ericsson’s collaboration. The integration of customizations, such as Markaz’s AI-based Autonomous Framework, was instrumental in delivering instant closed-loop adaptations tailored to the unique demands of Hajj.

Key enhancements include:

• Best Hajj performance, based on bit rate and latency measurements during the event, resulting in the highest communication quality for millions of attendees
• Approximately 280% Year-over-Year (YoY) improvement in network actions driven by cognitive network solutions
• A 15% enhancement in SINR, 9% improvement in higher order modulation, 7% enhancement in spectral efficiency and a 10% 5G throughput gain. These achievements enabled more efficient use of the network's spectrum without increasing capacity on actionable cells, alongside better communication quality despite the increasing number of users
• Delivery of zero-touch operations through closed-loop L4 autonomy

This project sets a new standard for future large-scale events demanding flawless connectivity and demonstrates how AI-driven automation is reshaping network operation paradigms.

Looking ahead, Mobily is poised to bring cutting-edge technologies to  deliver unmatched service quality and operational efficiency through intelligent network optimization.