Adaptive Service-Oriented Mobile Applications: A Declarative Approach

Abstract

Modern society increasingly relies on mobile devices and on distributed applications that use them. To increase development efficiency and shorten time-to-market, mobile applications are typically developed by composing together ad-hoc developed components, services available on-line, and other third-party mobile applications. To cope with unpredictable changes and failures, but also with the various settings offered by the plethora of devices, mobile applications need to be adaptive. We address this issue by proposing a declarative approach. The advantages of the proposed solution are demonstrated through an example inspired by an existing worldwide distributed mobile application.

Learning automata-based adaptive web services composition

“…A repair technology [6] is used with planning graph of AI planning for the adaptation. In [7], A SelfMotion Approach is proposed that separates abstract and concrete actions and supports one-to-many mappings. And a declarative language supports design time activities and a middleware is responsible for run-time activities.…”

Self-Adaptation in Mobile Apps: a Systematic Literature Study

“…If we examine the duration attribute in Figure 3d, we observe that most of the primary studies support goals with a persistent validity, as opposed to only 5 studies supporting temporary goals. The approach presented in [8] is an example of study dealing with temporary goals; the purpose of the approach is to allow developers to develop mobile applications in a declarative manner; then it will be the responsibility of the system to adapt the application to the device on which it is deployed. Since the adaptation goal is limited to the deployment phase, we can consider it as temporary.…”

“…Differently, changes are foreseeable when they are not known at design time, but they can be resolved at runtime and there is a plan for managing them during the execution of the system [19]. Examples of foreseeable changes include: the backend of the app has a failure [5], the GPS sensors of the mobile device produces incorrect data [8], etc. Finally, it is interesting to note that no primary study considers unforeseen changes, i.e., drastic changes that have been not planned for and that are unknown until their first occurrence [19].…”

Mechanisms for Leveraging Models at Runtime in Self-adaptive Software

“…These search-based approaches prescribe goals that the running system should achieve, and guided by utility functions they try to find the best or at least a suitable system configuration fulfilling these goals [21,22,39,76]. Other search-based mechanisms are based on model checking techniques to find plans on how to adapt the running system [90].…”