Resilient Resource Allocation for Service Placement in Mobile Edge Clouds

TECoSA PhD student Peiyue Zhao will defend his thesis – abstract below.  All are welcome to join. Please contact peiyue @ kth.se for the link.

Abstract

Mobile edge computing makes available distributed computation and storage resources in close proximity to end users and allows to provide low-latency and high-capacity services within mobile networks. Therefore, mobile edge computing is emerging as a promising architecture for hosting critical services with stringent latency and performance requirements, which otherwise are challenging to be addressed in conventional cloud computing architectures. Notable use cases of mobile edge computing include real-time data analytic services, industrial process control, and computation offloading for massive Internet of things devices. However, those services rely on efficient resource management, including resource dimensioning and service placement, and require to be resilient to cyber-attacks, to faulty components and to operation mistakes. The work in this thesis proposes models of resilient resource management that support rapid response to incidents in mobile edge computing and develops efficient algorithms for the resulting resource management problems.​

In the first part of the thesis, we consider resilient resource management for edge computing systems in which failover is realized by restoring additional service instances in different mobile edge computing nodes in case of failures. We first develop a placement algorithm based on Benders decomposition and linear relaxation to determine the mobile edge computing nodes to be opened and to compute the placement of the service instances with respect to a set of considered failure scenarios, with the objective of minimizing operation costs. Upon the occurrence of a failure scenario, service migration is to be triggered to migrate the service instances from one placement to another placement, for which we further develop service migration algorithms to schedule migration under time constraints, so as to minimize service interruptions.

In the second part of the thesis, we consider resilient resource management in mobile edge computing for services with different levels of resilience requirements. Resilience is achieved by synchronizing states of the services to two types of standby instances that maintain the trade-off between energy consumption and activation time such that the standby instances can take over the service seamlessly as an instantaneous failure response. We formulate the joint problem of resource dimensioning and service placement for minimizing energy consumption and prove that it is NP-hard. We propose an efficient approximation algorithm based on Lagrangian relaxation to decide the type, amount, and locations of the computation resources and to compute the placement of service instances and their associated standby instances. We then consider the same resilience model but for hosting periodic services in mobile edge computing systems with resources portioned into availability zones, under schedulability constraints. We formulate the corresponding resilient resource management problem as a non-linear programming problem and prove that it is NP-hard. We propose efficient solutions based on approximation programming and primal-dual approaches for resilient service placement.

By considering different models of resilient service placement in mobile edge computing, the results in this thesis provide effective, efficient, and scalable resource management algorithms for emerging mobile edge computing systems.