A Novel Theoretical Model for Channel Estimation in Massive MIMO Systems for Nigerian 5G Deployment Under Pilot Contamination

Massive MIMO technology deployed in Nigeria's emerging 5G networks faces severe pilot contamination in urban environments where the number of simultaneously active users exceeds the pilot sequence length, causing correlated channel estimation errors that degrade the promised capacity gains, and no theoretical model for pilot contamination impact calibrated to Nigerian urban deployment conditions exists. This study developed a novel theoretical model for channel estimation accuracy in massive MIMO systems under pilot contamination specific to Nigerian urban 5G deployment conditions. A comprehensive measurement campaign characterised antenna correlation and user density characteristics at 3.5 GHz in Lagos (Ikeja and Victoria Island), Abuja (Central Business District), and Kano (Sabon Gari) using a 32-element virtual ULAR testbed constructed from commodity USRP hardware. Three novel theoretical contributions were made. First, a spatially correlated massive MIMO channel model for Nigerian urban environments was derived, parameterised from the measurement data, showing higher spatial correlation than 3GPP Urban Macro models in Lagos due to building geometry. Second, an original closed-form pilot contamination interference expression was derived as a function of user density, channel correlation structure, and pilot reuse factor, incorporating the Nigerian-specific channel statistics. Third, a novel Interference-Aware Pilot Assignment Algorithm (IAPA) was developed, exploiting spatial channel correlation structure to reduce pilot contamination through data-driven pilot orthogonalisation across correlated users, achieving 34.7 percent reduction in pilot contamination power versus random assignment. The theoretical framework was validated through 5G NR system-level simulation. Expert review confirmed the original channel modelling and algorithm contributions.