A Microservice-based Approach to Facilitate Multi-Services in Smart Space
Jie Sun, Ming Jin, Yang Wang · 2021
Smart space refers to the integration of compute resources, information equipment and multi-modal sensing devices in the work or living space so that diverse services can be shared between different users in a smart way. Such a form of pervasive computing can not only make compute resources and information services universally available in such a way suitable for people's special uses but also bridge the digital information space and the physical world. However, given the properties of the smart space-each service with different resource requirements and goals, the smart space could cause resource conflicts in uses between different services and participants, and thus lead to the resource contention. In order to solve this problem while fulfilling the requirements of variable services in the smart space, in this paper, we leverage the concept of microservice to propose a universal resource scheduling and a process-evolution method for multiple concurrent services in the smart space. At the same time, to ensure that the microservice-based method is easy to implement, we also design a corresponding scheduling architecture based on the docker engine, which can functionally decompose application into a set of collaborative services. Finally, we carry out a experiment to show that the proposed architecture can achieve reasonable resource allocation, resolve resource contention, and exhibit the virtue of closed loop within different scenes and cross-scenario linkages.