Intelligence serverless computing system for telepresence services
Аннотация
This paper addresses the joint service migration and resource allocation for minimum latency optimization problem in the multi-access edge computing network of a telepresence system. Due to access latency, MEC compensates for the shortfall in the capacity of the Cloud by providing services in the vicinity of consumers. The high mobility of devices and the finite resources further aggravate the problem of maintaining the quality of service at edge servers. It is evident that migration itself introduces latency; various users and servers have their requirements and resource management policies, so all this evidently demands a balanced resolution of this issue. Despite the intensive development in the area of telepresence (high-quality video, spatial audio, and virtual and augmented reality), efficient functioning of such systems needs stable infrastructure and low latency of interactions. This article proposes a combined algorithm, namely SMRA+MEC, considering all the details of the telepresence system, making the decision on whether to migrate any service and in which way the resources are to be optimally distributed. This paper proposes a new latency calculation mechanism so that the reduction can be further achieved and hence the resource distribution can be optimized. After all, it turned out that the most effective approach to latency reduction can be realized with the support of combined SMRA+MEC techniques.