Managing network configuration and improving service experience effectively is essential for cellular service providers (CSPs). This is challenging because of cellular networks' large scale and complexity, the wide variety of configuration parameters, and the performance impact tradeoffs resulting across multiple metrics and geographical locations. This paper focuses on learning and using network contexts to recommend performance-improving configurations. While learning contexts, one must carefully account for the configuration parameter dependency, performance impact confusion that can arise due to co-occurring unrelated changes, and uneven change deployment distribution across locations. We present a new solution Chroma that addresses the above challenges. Using real-world data collected from a large operational LTE and 5G cellular service provider, we thoroughly evaluate and demonstrate the efficacy of Chroma. We successfully trial Chroma on an operational cellular network and highlight its benefits in practical settings.

Cellular service operators frequently tune the network configuration to optimize coverage, support seamless handovers, minimize channel interference, and improve the service performance experience to the end-users. Tuning such a complicated network is highly challenging because of the many configuration parameters, evolving complexity of cellular networks, and diverse requirements across voice, video, and data applications. Any misconfigurations or even poor settings can significantly negatively impact service quality. In this paper, we propose a new approach Aurora that derives best practices knowledge from exploration of the massive existing configuration in the network and uses conformity-based recommendation with performance-based filtering to improve cellular service. We implemented and evaluated Aurora using data from a very large LTE and 5G cellular service provider. Our operational experience over the last one year highlights the benefits of Aurora and exposes exciting research opportunities and challenges in configuration tuning and performance management.

In a virtual network environment, a substrate network provider allocates computing and networking resources for service providers who request virtual networks to be created for particular services, and it also has the capability to provide resilient virtual network management with redundant resources, such as dynamic virtual network restoration from failures. In this papers, we consider the situation where the substrate network provider desires to have standby virtual routers ready to serve virtual networks under node failures. Such a failure can affect one or more virtual routers in multiple virtual networks. The goal of this paper is to make the optimal selection of standby virtual routers so that virtual networks can be dynamically reconfigured back to their original topologies right after the failures. We present an optimization formulation and a heuristic for this problem. By considering a number of factors, we present numerical studies to show how the optimal selection was affected by those factors, and the proposed heuristic's performance was close to the optimization model when there were sufficient standby virtual routers for each virtual network and the substrate nodes have the capability to support multiple standby virtual routers to be in service concurrently.

Fault-tolerance in a virtualized networking environment (VNE) requires additional efforts to provide the survivability against failures on either virtual networks (VNs) or the underlying substrate network. In this dissertation, we design a software-defined resilient VNE using a hybrid scheme of protection and restoration, where for each VN, a set of redundant virtual routers (VRs) are reserved as shared standby virtual routers (S-VRs), and any S-VR can be activated to replace a failed VR in the existed VN dynamically after the failure is identified. We first introduce a dynamic reconfiguration scheme (DRS) for node failures in a VNE, and then propose a Mixed-Integer Linear Programming (MILP) model with dual goals to optimally select S-VRs to restore all VNs while load balancing, and a heuristic algorithm based on the model is also presented. By considering a number of factors, the results showed that the proposed heuristics performance was close to the optimization model when there were sufficient standby virtual routers for each virtual network and the substrate nodes had the capability to support multiple standby virtual routers to be in service simultaneously. Finally, we present a prototyping design and implementation on the realistic virtual network testbeds (i.e., GpENI and GENI).

Multipath routing gives traffic demands an opportunity to use multiple paths through a network. In a single-demand situation, its benefits are easy to see. In a multi-commodity case, when potentially all node-pairs (demands) generate traffic, they compete for the same network resources. In this work, we consider multipath routing in communication networks in a multi-commodity setting from a traffic engineering perspective. Based on a result from linear programming, we show that at an optimal solution, the number of demands that can have multiple paths with nonzero flows is of the order of the number of network links for three commonly used traffic engineering objectives. We introduce a multipath measure (MPM) and show that under certain traffic conditions and topological structures, the MPM is zero or close to zero, i.e., Multipath routing provides little or limited gain compared to single-path routing. For the all-pair traffic case, multipath routing is observed to be advantageous for small networks. When the number of nodes is about 25 or higher and all node pairs have traffic, this advantage drops as the number of nodes in a network increases. For the fat-tree data center topology, the benefit of multipath routing also drops as the number of pods increases. Our findings are somewhat against a common belief (expressed by the term "load sharing") that multipath routing is significantly better in effective distribution of traffic over the network resources.

With Cloud Computing technology occupying the majority of future Internet research and development work, research on deploying and extending existing capabilities onto a newly emerging infrastructure becomes more significant. For example, extending the virtual network provisioning capability onto a Geo-distributed programmable layer-2 networking environment (G-PLaNE) is a novel attempt and is different from in a single domain system. In this paper, we aim to illustrate how to construct the virtual networking environment upon our self-designed resource provisioning system consisting of multiple clusters through G-PLaNE. Experimenters and researchers are able to develop and explore their own mechanisms in our platform. Furthermore, a concrete example named Secure and Resilient Virtual Trust Routing (SeRViTR) is given to illustrate how this can be constructed over G-PLaNE.

A secure network is considered to be an important goal of the Future Internet; one way this can be embodied is by having flexible and robust routing functionalities with intrinsic security mechanisms. It is also desirable to provide user-centric or service-centric routing capabilities to achieve service-oriented traffic controls as well as trust and policy management for security. Based on these potential needs, a flexible, scalable, and robust routing framework that enables fine-grained flow control under fixed or dynamic policies called the Virtual Trusted Routing and Provisioning Domain (VTRouPD)[11] has been recently proposed. In this paper, we present a framework called the Secure and Resilient Virtual Trust Routing (SeRViTR) framework, which is a proof-of-concept model of VTRouPD at the implementation level. SeRViTR has particular entities that are designed for policy management and trust management between different VTRouPDs to enable a secure Internet. We define the roles of each entity within the SeRViTR framework as well as the messages exchanged between them. We also discuss how policy management and trust negotiation can be achieved. Moreover, we present validation on the functional implementation of several SeRViTR components to illustrate how to create virtual domains and change of trust levels between virtual domains.

Great Plains Environment for Network Innovation (GpENI) is an international testbed for future Internet research. A key component of GpENI is programmable network virtualization (GpENI-VINI). The scope of this paper is to present the framework, implementation and integration experience with network virtualization in GpENI. In particular, this is described through our experience of implementing and integrating the XORP (eXtensible Open Router Platform) routing platform into GpENI-VINI. Preliminary results on measurements and validation are presented.