Candidate: Kaige Qu
Title: Dynamic Resource Provisioning and Scheduling for Embedded Services in SDN/NFV-Enabled Core Networks
Date: November 30, 2020
Time: 2:00 PM
Place: REMOTE ATTENDANCE
Supervisor(s): Zhuang, Weihua
The service-oriented fifth-generation (5G) core networks are featured by customized network services with differentiated quality-of-service (QoS) requirements, which can be provisioned through network slicing enabled by the software defined networking (SDN) and network function virtualization (NFV) paradigms. Multiple network services are embedded in a common physical infrastructure, generating service-customized network slices. Each network slice supports a composite service via virtual network function (VNF) chaining, with dedicated packet processing functionality at each VNF. For a network slice with a target traffic load, the end-to-end (E2E) service delivery is enabled by VNF placement at NFV nodes (e.g., data centers and commodity servers) and traffic routing among corresponding NFV nodes, with static resource allocations. To provide continuous QoS performance guarantee over time, it is essential to develop dynamic resource management schemes for the embedded services experiencing traffic dynamics in different time granularities during virtual network operation. In this thesis, we focus on processing resources and investigate three research problems on dynamic processing resource provisioning and scheduling for embedded delay-sensitive services, in presence of both large-timescale traffic statistical changes and bursty traffic dynamics in smaller time granularities.
In problem I, we investigate a dynamic flow migration problem for multiple embedded services, to accommodate the large-timescale changes in the average traffic rates with average E2E delay guarantee, while addressing a trade-off between load balancing and flow migration overhead. We develop optimization problem formulations and efficient heuristic algorithms, based on a simplified M/M/1 queueing model with Poisson traffic arrivals. Motivated by the limitations of Poisson traffic model, in problem II, we restrict to a local network scenario and study a dynamic VNF scaling problem based on a real-world traffic trace with nonstationary traffic statistics in large timescale. Under the assumption that the nonstationary traffic trace can be partitioned into non-overlapping stationary traffic segments with unknown change points in time, a change point detection driven traffic parameter learning and resource demand prediction scheme is proposed, based on which dynamic VNF migration decisions are made at variable-length decision epochs via deep reinforcement learning. The long-term trade-off between load balancing and migration overhead is studied. A fractional Brownian motion (fBm) traffic model is employed for each detected stationary traffic segment, based on properties of Gaussianity and self-similarity of the real-world traffic. In Problem III, we focus on a sufficiently long time duration with given VNF placement and stationary traffic statistics, and study a delay-aware VNF scheduling problem to coordinate VNF scheduling for multiple services, which achieves network utility maximization with timely throughput guarantee for each service, in presence of bursty and unpredictable small-timescale traffic dynamics, while using a realistic state-of-the-art time quantum (slot) for CPU processing resource scheduling among VNF software processes. Based on the Lyapunov optimization technique, an online distributed VNF scheduling algorithm is derived, which greedily schedules a VNF at each NFV node based on a weight incorporating the backpressure-based differential backlogs of required CPU cycles, the timely throughput performance, and the packet delay.
With the proposed dynamic resource management framework, resources can be efficiently and fairly allocated to the embedded services, to avoid congestion and QoS degradation in the presence of traffic dynamics. This research provides some insights in dynamic resource management for delay-sensitive services in a virtualized network environment with CPU processing resources.