Computation offloading in the edge-to-cloud compute continuum: a survey of federated architectural solutions

Abstract Computation offloading involves transferring computation to resourceful nodes to overcome resource limitations, particularly within the emerging federated edge-to-cloud computing infrastructures. The growth of IoT applications has hugely increased the need for effective and decentralised offloading strategies considering the highly dynamic and dispersed edge-to-cloud compute continuum. Recently, researchers have used various implementation techniques—ranging from rule-based systems, heuristic, and machine learning-based intelligence—to investigate issues related to offloading, including edge computing, fog computing, and edge-to-cloud compute continuum. Although several review papers are available that provide a comprehensive analysis of existing research works on computation offloading, most surveys have largely overlooked the specific challenges and characteristics of federated and distributed edge-to-cloud execution models. In contrast, this paper aims to compile and synthesise computation offloading research with a specific focus on the federated edge-to-cloud ecosystem and distributed execution solutions. We first propose a detailed taxonomy focused on the decentralisation aspect and federated coordination. This taxonomy is then used as a unified framework to critically review the existing research landscape. Finally, we identify and discuss the key challenges that require further attention, providing insights to guide future developments in distributed, federated offloading systems.