Telecom service providers are adopting a Network Functions Virtualization (NFV) based service delivery model, in response to the unprecedented traffic growth and an increasing customers demand for new high-quality network services. In NFV, telecom network functions are virtualized and run on top of commodity servers. Ensuring network performance equivalent to the legacy non-virtualized system is a determining factor for the success of telecom networks virtualization. Whereas in virtualized systems, achieving carrier-grade network performance such as low latency, high throughput, and high availability to guarantee the quality of experience (QoE) for customer is challenging.

In this thesis, we focus on addressing the latency challenge. We investigate the delay overhead of virtualization by comprehensive network performance measurements and analysis in a controlled virtualized environment. With this, a break-down of the latency incurred by the virtualization and the impact of co-locating virtual machines (VMs) of different workloads on the end-to-end latency is provided. We exploit this result to develop an optimization model for placement and provisioning of the virtualized telecom network functions to ensure both the latency and cost-efficiency requirements.

To further alleviate the latency challenge, we propose a multipath transport protocol MDTCP, that leverage Explicit Congestion Notification (ECN) to quickly detect and react to an incipient congestion to minimize queuing delays, and achieve high network utilization in telecom datacenters.

The demand from customers for high-quality and customized network services has increased. Telecom service providers have adopted a Network Functions Virtualization (NFV) based service delivery model, in response to the unprecedented traffic growth and increasing customer demand. However, in virtualized systems, achieving carrier-grade network performance such as low latency to guarantee the quality of experience (QoE) of customers is challenging. Moreover, queuing delays that may occur both in the datacenter networks and the IP network infrastructure inhibit the deployment of emerging low-latency services.

In this thesis, we focus on addressing the problem of network latency. We study the delay overhead of virtualization by comprehensive network performance measurements and analysis, in a controlled virtualized environment. The study of virtualization delay provides a break-down of the latency imposed by the virtualization and the impact of the consolidation of virtualized applications of different workloads on the end-to-end latency. On the basis of our study, we developed an optimization model for the placement and provisioning of virtualized telecom applications subject to both the latency and cost-efficiency requirements.

To mitigate network latency that results from queuing delays as well as to improve multipath network capacity utilization of a datacenter network, we propose a multipath congestion control, Multipath Datacenter TCP (MDTCP), that leverages Explicit Congestion Notification (ECN) to detect and react to queuing delays caused by incipient congestion. Furthermore, we extend MDTCP with the Low Latency Low Loss and Scalable Throughput (L4S) Internet service architecture support so that it can also be used in the Internet. This ensures the low-latency demand of delay-sensitive applications and improves QoE of Internet users by exploiting the multi-access or multi-connectivity technologies of user devices.

The demand for high-quality network services has increased. Telecom service providers have adopted the NFV-based service delivery model, in response to the unprecedented traffic growth and increasing customer demand. In virtualized systems, achieving carrier-grade network performance such as low latency to guarantee the QoE of customers is challenging. Moreover, queuing delays that may occur both in the datacenter and IP networks inhibit the deployment of low-latency services.

This thesis addresses the problem of network latency. We study the delay overhead of virtualization by comprehensive network performance measurements and obtain the break-down of the latency imposed by the virtualization and the impact of the consolidation of virtualized applications of different workloads on the end-to-end latency. Then, we developed an optimization model for the placement and provisioning of virtualized telecom applications subject to both the latency and cost-efficiency requirements.

To mitigate queuing delays and improve multipath network capacity utilization of a datacenter network, we propose MDTCP that leverages ECN to detect and react to queuing delays caused by incipient congestion. We extend MDTCP with the L4S architecture support so that it can also be used in the Internet. This ensures the demand of delay-sensitive applications and improves QoE of Internet users by exploiting the multi-access or multi-connectivity technologies of user devices.