ALWAYS-CONNECTED IOT: OPTIMISING BANDWIDTH AND COST FOR DATA STREAMING OVER CELLULAR NETWORKS

Abstract

The current paper researches the maximisation of bandwidth and costs in constantly connected Internet of Things (IoT) systems, which are based on constant data transfer over cellular networks, where constant connectivity creates a great problem in relation to spectral efficiency, operation costs, and device maintenance. The main idea is to critically review the current strategies, which are adaptive bitrate encoding, SDN-based dynamic bandwidth allocation, geo-distributed streaming analytics, and energy-aware hybrid networking, and to synthesise the contributions with each other into a framework of scalable IoT rollout. The study methodologically uses secondary data based on peer-reviewed technical literature and proceeds to perform thematic analysis in order to find out the repetitive technical patterns, trade-offs and interdependencies between the cross layers. The strategy helps to synthesise the empirical results rigorously without repeating experimental designs, which is a methodological efficiency and conceptual richness. Findings indicate that adaptive bitrate encoding offers better spectral allocation and Quality of Experience, though hardware support is needed to reduce computational overhead. SDN-enabled allocation offers programmable traffic control and QoS implementation, but controller scaling is a bottleneck. Geo-distributed analytics incurs less cost in inter-datacenter transfers, whereas WAN pricing limits efficiency. Together, the results highlight the fact that the sustainable deployment of IoT requires combined orchestration of codecs, programmable control planes, distributed analytics, and interface-aware scheduling, and balancing the latency, cost, energy and scalability of cellular IoT ecosystems.