With the rapid increase in the number of users and changing pattern of network usage, cellular networks will continue to be challenged meeting bandwidth and latency requirements. A significant contributor to latency and overhead is cellular network's complex control-plane. We propose CleanG, a new packet core architecture and significantly more efficient control-plane protocol, that exploits the capabilities of modern-day Network Function Virtualization (NFV) platforms. CleanG is a single component NFV-based architecture. With the elastic scalability offered by NFV, the data and control sub-components of the core can scale, adapting to workload demand. CleanG eliminates the use of GTP tunnels for data packets and the associated complex protocol for coordination across multiple, distributed components for setting up and managing them. We carefully examine the use of each protocol message exchange (and the component fields of those messages) in developing a substantially simplified protocol, while retaining similar essential functionality for security, mobility, and air-interface resource management. We have implemented CleanG on the OpenNetVM platform and perform an apples-to-apples comparison with the existing 3GPP LTE architecture and an architecture that separates the control and user plane (the CUPS-based architecture like the 5G architecture). Measurements on our testbed show that CleanG substantially reduces both control and data plane latency, and significantly increases system capacity.

With the rapid increase in the number of users and changing pattern of network usage, cellular networks will continue to be challenged meeting bandwidth and latency requirements. A significant contributor to latency and overhead in cellular networks is the complex control-plane involving many message exchanges across multiple components in the packet core, base station, and user equipment. We propose CleanG, a new packet core architecture and significantly more efficient control-plane protocol, that exploits the capabilities of modern-day Network Function Virtualization (NFV) platforms. In CleanG, we have consolidated the core components into a set of virtual network functions on an NFV platform. With the elastic scalability offered by NFV, the data and control sub-components of the core functions can scale, adapting to workload demand. CleanG eliminates the use of GPRS Tunneling Protocol (GTP) Tunnels for data packets and the associated complex protocol for coordination across multiple, distributed components for setting up and managing them, as specified in the 3rd Generation Partnership Project (3GPP) architecture and protocol standard, while retaining similar essential functionality for security, mobility, and air-interface resource management. Measurements on our testbed show that CleanG substantially reduces both control and data plane latency, and significantly increases system capacity.

Network function virtualization (NFV) implements network middleboxes in software, enabling them to be more flexible and dynamic. NFV resource allocation methods can exploit the capabilities of virtualization to dynamically instantiate network functions (NFs) to adapt to traffic demand and network conditions. Deploying NFs requires decisions for NF placement, and routing of flows through these NFs in accordance with the sequence of NFs required to process each flow. The challenges in developing an NFV resource allocation scheme include the need to manage the dependency between flow-level (routing) and network-level (placement) decisions and to efficiently utilize resources that may be distributed network-wide, while fulfilling the performance requirements of flows. We propose a scalable resource allocation scheme, called ClusPR, that addresses these challenges. By elegantly capturing the dependency between flow routing and NF placement, ClusPR strikes a balance between multiple objectives including minimizing path stretch, balancing the load among NF instances, while maximizing the total network utilization by accommodating the maximum number of flows possible. ClusPR addresses the offline problem of NFV resource allocation. To address the online problem of dynamically placing and routing flows upon their arrival, we propose iClusPR. iClusPR is an online algorithm that performs dynamic scaling by adjusting the number of NF instances based on the traffic demand and the network state. Our experiments show that ClusPR achieves the near-optimal solution for a practical large-sized network in reasonable time. Compared to the state-of-the-art approaches, ClusPR decreases the average normalized delay by a factor of 1.2 - 1.6 × and the worst-case delay by more than 10 ×, with the same or slightly better network utilization and balances the load among NF instances. Furthermore, the performance of iClusPR, the online version, is comparable to the offline ClusPR algorithm.

As demand for wireless mobile connectivity continues to explode, cellular network infrastructure capacity requirements continue to grow. While 5G tries to address capacity requirements at the radio layer, the load on the cellular core network infrastructure (called Enhanced Packet Core (EPC)) stresses the network infrastructure. Our work examines the architecture, protocols of current cellular infrastructures and the workload on the EPC. We study the challenges in dimensioning capacity and review the design alternatives to support the significant scale up desired, even for the near future. We breakdown the workload on the network infrastructure into its components-signaling event transactions; database or lookup transactions and packet processing. We quantitatively show the control plane and data plane load on the various components of the EPC and estimate how future 5G cellular network workloads will scale. This analysis helps us to understand the scalability challenges for future 5G EPC network components. Other efforts to scale the 5G cellular network take a system view where the control plane is separated from the data path and is terminated on a centralized SDN controller. The SDN controller configures the data path on a widely distributed switching infrastructure. Our analysis of the workload informs us on the feasibility of various design alternatives and motivates our efforts to develop our clean-slate approach, called CleanG.

The integration of network function virtualization (NFV) and software defined networks (SDN) seeks to create a more flexible and dynamic software-based network environment. The line between entities involved in forwarding and those involved in more complex middle box functionality in the network is blurred by the use of high-performance virtualized platforms capable of performing these functions. A key problem is how and where network functions should be placed in the network and how traffic is routed through them. An efficient placement and appropriate routing increases system capacity while also minimizing the delay seen by flows. In this paper, we formulate the problem of network function placement and routing as a mixed integer linear programming problem. This formulation not only determines the placement of services and routing of the flows, but also seeks to minimize the resource utilization. We develop heuristics to solve the problem incrementally, allowing us to support a large number of flows and to solve the problem for incoming flows without impacting existing flows.