Mobile learning (m-learning) adoption has incredibly increased with the implementation of related computing paradigms. The mobile cloud architectures (MCAs) enable m-learning with several benefits and face limitations with m-learning actors’ changing requirements. IoT, edge, mobile edge, fog, AI, and 5G, bring numerous features and increase m-learning efficiency across educational disciplines. This study investigates the state-of-the-art m-learning architectures, determines a unified m-learning MCA, and explores the related computing paradigms’ characteristics to expand m-learning provision. Also, it evaluates m-learning performance across the MCAs and the emerging computing architectures. It finds the four physical layer’s MCAs and several application layer’s m-learning architectures. Only distance-cloud MCA does explore, and the other three MCAs do ignore by experts. Besides, the performance evaluation in related computing paradigms gives terrific benefits and QoE. MEC offers ultra-low latency for resource-intensive m-learning applications, fog using AI algorithms is exceptional for more complex learning objects, IoT is superior for intelligent learning tools, and 5G Ultra-Wideband services are more significant for intelligent video analytics. Eventually, it identifies the challenges, limitations, presents implications, and raises the future research directions to improve m-learning performance efficiency. The study’s findings help m-learning actors, institutions, and potential stakeholders by following their needs.