Development of an Original Adaptive Network Coding Framework for Reliable Multicast Delivery Over Heterogeneous Wireless Networks in Nigeria

Multicast content delivery over heterogeneous wireless networks in Nigeria, including community health information broadcasts, educational content distribution, and public safety alerts, requires reliable delivery guarantees across diverse receiver link quality conditions without the per-receiver feedback overhead that makes unicast repair approaches unscalable. Network coding enables novel approaches to reliable multicast that exploit linear combinations of data packets, yet adaptive network coding frameworks optimized for Nigerian heterogeneous wireless environments are absent from the literature. This doctoral study develops an original adaptive network coding framework for reliable multicast delivery over heterogeneous wireless networks in Nigeria, with theoretical analysis of decoding delay bounds and practical algorithm design. An original adaptive random linear network coding scheme is developed in which the coding coefficient field size and generation size are dynamically adjusted based on estimated channel conditions at heterogeneous receivers, minimizing decoding delay while bounding the computational overhead at mobile terminals. Theoretical decoding delay bounds for the proposed adaptive scheme are derived using Markov chain analysis, providing closed-form expressions as functions of receiver channel statistics and network coding parameters. The bounds are parameterized using measured loss rate statistics from Nigerian mobile networks in the South South and North West zones. Simulation evaluation demonstrates that the adaptive framework achieves 99.2% reliable delivery to the 95th percentile receiver while reducing average decoding delay by 34% compared to fixed-parameter random linear network coding, with receiver computation overhead within the processing budget of mid-range Android smartphones. Keywords: network coding, multicast delivery, heterogeneous network, reliability, Nigeria.