Application models for ubiquitous systems with sporadic communication availability

Ubiquitous computing systems are predominantly characterized by limited computing power, communication capabilities and energy sources. Yet novel usage scenarios shift their roles from basic data acquisition and forwarding to performing more sophisticated tasks such as data processing, local service provisioning and actuation. To meet their goals, horizontal or vertical collaboration and task distribution are often required; utilizing resources identified in the nearby communication environment or offered by remote, cloud- or edge-based services. However, different infrastructural and energy-related constraints can severely influence device's communication capabilities thus dictating the level of provided functionality. In this paper, we focus on the class of ubiquitous systems without permanent communication links, utilizing only sporadic communication with global computing infrastructure. Such systems exploit the short-term presence of third-party communication and computational resources in their environment, adapting their service level to available resource properties. Several identified communication and computation scenarios are described for such a resource constrained and time-variable environment, along with the incentive model for using and provisioning of short-term resources. A general model of adaptive distributed applications exploiting such sporadic communication and computation environment is presented, relying on the cloud- and edge-computing paradigms and microservices as its building blocks.