Research Paper

Core Innovation

This paper introduces a hidden Markov model-based approach to link mobile trajectories with CSI evolution, enabling blind angular power map construction without location labels. It demonstrates theoretical bounds on localization error and validates performance with real multi-cell massive MIMO data, achieving practical localization accuracy from limited measurements.

Why It Matters

Accurate localization and radio resource management in massive MIMO networks are hindered by costly and impractical location-labeled CSI data collection. This solution reduces reliance on labeled data, enabling scalable and efficient network optimization and improved user experience. It supports better resource allocation and network planning at scale.

Potential Customers & Pain Points

• Mobile network operators–High cost and complexity of CSI data labeling
• Telecom equipment vendors–Need for advanced localization tools
• Smart city planners–Require precise radio environment maps for IoT deployment
• Autonomous vehicle platforms–Depend on accurate wireless localization
• Wireless infrastructure providers–Seek improved network resource management.

Business Model

SaaS platform licensing to mobile network operators and telecom vendors with tiered pricing based on network size and data volume.

Competitive Landscape

• Google Maps APIs
• HERE Technologies
• Cisco DNA Spaces
• Qualcomm Location Services

Implementation Challenges

• Integration with existing network infrastructure
• Data privacy and security concerns
• Variability in real-world deployment environments

Validation Strategy

• Pilot deployments with telecom operators in urban environments
• Benchmarking against existing localization and radio mapping solutions
• Collecting feedback to refine model accuracy and usability