Dynamic service deployment for budget‐constrained mobile edge computing

Abstract

Summary Currently, Mobile edge computing (MEC) is facing a great challenge that is how to make full use of edge resources to provide a seamless support for compute‐intensive latency‐sensitive applications. Prior studies often make a simple assumption that tasks can be executed upon every edge server, but the assumption does not hold in practical scenarios. Because a specific application task often corresponds to a certain service that provides the corresponding running environment, whereas an edge server only has limited resources and cannot offer too many services. How to decide service deployment of so many types of services among multiple edge servers is also a big challenge. To address the challenge, we study dynamic service deployment for latency‐sensitive applications. We first model the long‐term budget‐constrained latency minimization problem as a multi‐slot latency minimization problem based on the Lyapunov framework. By doing this, the hardness of a problem is significantly reduced, since we never require future information to solve the long‐term optimization. Furthermore, we extend our study by joining the task scheduling optimization, where every edge server is fully utilized in an even more efficient collaborative manner. Our extensive experiments show that the proposed algorithms can bring short latency with low cost.

Dynamic Service Placement in Multi-Access Edge Computing: A Systematic Literature Review

“…Fourth, the cache update cost is usually omitted in content caching methods [15], which is not the case in dynamic service placement. Finally, task sharing among servers [22], [23] and service chaining make dynamic service placement a distinctive problem. The costs of service access and cache update coupled with the interplay between the available services in the cache and the computing capacity at the base stations make the service caching and task-assignment policies more challenging to design.…”

“…Additionally, the limitation hinders frequent reconfiguration of services in edge servers. Zhou et al [22], [23] introduce a budget-constrained dynamic service deployment method in which the cost model comprises service installation cost, computation cost, and traffic cost. The service installation cost refers to the expenses of service copyright and maintenance, computation cost corresponds to the utilisation of computational resources in the edge servers, and finally, traffic refers to the expenditures paid for incoming and outgoing data transmission.…”

“…• Considering a user device within coverage areas of multiple base stations, the first design issue is to decide whether user devices are in charge of selecting a base station or the base stations have to select their serving user devices. Some researchers in [22], [23], [171] put the burden on user devices by selecting a base station according to the quality of its wireless channels, while some others [113], [168] consider that the base stations can select user devices based on the number of occupied and available channels.…”

“…• The second design issue is to identify the possible implications of using coverage overlap. The coverage overlap can be used for wireless channel management [22], [23], [171]. Besides, a user device may have direct access to the cached services of multiple base stations within its reach.…”

Cloud computing–based big data processing and intelligent analytics

“…How to decide service deployment of so many types of services among multiple edge servers is also a big challenge. To address the challenge, Zhou et al study dynamic service deployment for latency‐sensitive applications and first model the long‐term budget‐constrained latency minimization problem as a multi‐slot latency minimization problem based on the Lyapunov framework. Furthermore, the task scheduling optimization is also considered here, which makes every edge server be fully utilized in an even more efficient collaborative manner.…”