Glossary of Terms

A


ASIC – Application-Specific Integrated Circuit. A device (chip) designed and used by a single company in a specific system.  An integrated circuit (IC) customized for a particular use. A microchip designed for a special application, such as a particular type of transmission protocol

Avionics – electronic and computational systems used on aircraft, satellites, and spacecraft, responsible for communication, navigation, and many other important functions during flight. There are specific international standards for the avionics industry to ensure the highest possible level of safety. FPGA-based SoCs can be used in avionics switches to accelerate communication between sensors aboard an aircraft


B


BFD – Bidirectional Forwarding Detection. A protocol defined in IETF RFC 5880 for detecting and responding to network faults. BFD provides a consistent failure detection method for network administrators at a uniform rather than variable rate, which makes profiling, planning, and reconvergence simpler and more predictable.  The goal of BFD is to provide low-overhead, short-duration detection of failures in the path between adjacent forwarding engines, including the interfaces, data link(s), and, to the extent possible, the forwarding engines themselves. See also:  OAM

BNG – Broadband Network Gateway. Another term for Broadband Remote Access Server. See also: BRAS

BRAS – Broadband Remote Access Servers. Routers at the edge of a service provider’s core network that route traffic to and from subscribers. BRAS serve as the main gateway element and must support massive throughput and be scalable enough to handle the significant growth in data usage that is expected from ultrabroadband services. It is at the BRAS that an ISP can inject policy management and IP Quality of Service (QoS). Traditional BRAS has fixed control plane and data plane capacity, which eliminates the flexibility to independently scale each of these elements to match actual network conditions. A virtual BRAS solution delivers a separate, software-based control plane, as well as a data plane that can scale independently and benefits from hardware acceleration.  BRAS functionality is defined in Broadband Forum TR-092. See also:  BNG


C


Carrier-Grade – a standard of quality within the telecommunications industry that ensures reliability. The expectation is that Carrier-Grade systems or components will provide or even exceed “five nines” (99.999%) high availability and provide redundancy to achieve fault recovery within 50 milliseconds

CSR/CSG/CSGR (Cell Site Router/Cell Site Gateway/Cell Site Gateway Router) – network equipment that provides Ethernet/IP connectivity between cellular sites, such as base stations and towers and the mobile backhaul network. By connecting mobile users to the internet, CSRs enable more advanced communication, including 4G and now 5G. Node at the cell site that presents the transport network interface to the base station equipment. In some cases, the CSR from a transport carrier provides transport services for base stations from several mobile operators

Central Office – a location in a telecommunications network where service providers/cable operators connect access/aggregation networks to metro/core networks. It therefore is the ideal point for enabling access to network services. Traditionally, Central Offices have used proprietary equipment that is hard to upgrade and ill-suited for next generation multi-access edge computing (MEC) services, but many providers are now virtualizing their Central Offices to overcome these concerns.  See also:  CORD, Virtual Central Office

CLI – Command Line Interface. A means of user interaction with a computer program in which the user provides commands to the program via successive lines of text, as opposed to a graphic user interface (GUI)

Control Plane – the part of the network that is responsible for network-wide logic. Traditionally, both the control plane and the data plane are implemented in the hardware, but with SDN,  the control plane has been decoupled from the data plane, and is handled through software implementation.  See also:  Data Plane

CORD (Central Office Re-architected as a Data Center) – a project of the Open Networking Foundation (ONF) that aims to bring data center economies and cloud agility to service provider Central Offices, improving the service offerings for their residential, enterprise, and mobile customers. The reference architecture of CORD combines commercial off-the-shelf servers and switches with open source software.  See also:  Virtual Central Office

COTS – Commercial Off-The-Shelf. A category of networking hardware and software that are pre-packaged or generic, as opposed to custom-developed components. COTS products tend to be less expensive, require less development time, and are less vendor-dependent than in-house or custom-made solutions


D


Datapath – a physical set of functional units that carry out data processing

Data Plane – the part of the network that determines and carries out the processes and functions (on a per-router basis) that process and forward data packets from one interface to another. With SDN, the data plane has been physically separated from the control plane that determines the network-wide logic.  See also:  Control Plane

DDoS – Distributed Denial of Service. A type of security attack on a computer system that floods the network with unwanted or superfluous requests for service from various distributed sources, overloading the system and disrupting service because the host is unavailable to its intended users

DDR – Double Data Rate. A specification for synchronous dynamic remote access memory (SDRAM) interface, with successively higher bandwidths and faster data access rates as the technology has advanced (DDR2, DDR3, DDR4)

DPDK – Data Plane Development Kit. A set of open source data plane libraries and NIC drivers for accelerating packet processing. DPDK provides a framework for programming in various environments in order to enable faster development of high-speed networking applications. DPDK is a very active open source project with a large community behind it.

DPU – Distribution Point Unit. A piece of hardware that receives fiber from the central network and distributes the bandwidth via G.fast to multiple locations closer to the end user.  See also:  G.fast


E


Edge Cloud – an ecosystem of data centers or data center capabilities located in close distance to end-users, enabling data to be processed closer to its source, which reduces latency and enables real-time, interactive applications

Edge Computing –the processing and analyzing of data at the network edge, closest to the point of its collection. Data processing starts at the point of collection and is distributed accordingly, such that only the data that needs to be forwarded is sent onward to the cloud

EFM Bonding – Ethernet in the First Mile. The “first mile” is the last segment of the network toward reaching the user (known as the “last mile” from the service provider’s perspective). By bonding multiple copper pairs, networks can carry far more bandwidth in a single, high-capacity link over the existing copper infrastructure, which is generally found in the first mile where fiber does not exist

EPC – Evolved Packet Core. A network framework that converges voice and data over a broadband IP architecture, enabling the 4G LTE mobile network


F


Flow Classification – The automated application of a policy to sequences of user data traffic, categorising them in order to assign them a priority level within the network traffic

Flow Processor – A next-generation network processor that handles more than just data transport, incorporating traffic management, classification, and services integration. A flow processor should offer programmability, flexibility, and scalability in addition to fast network performance

FPGA – Field-Programmable Gate Array. An integrated circuit that contains arrays of logic blocks that are designed to be programmed by a customer or a designer after manufacturing to perform complex functions. In other words, FPGAs are programmable hardware, combining the performance and efficiency of hardware with the flexibility of software

Fully Programmable – a technology that integrates the software programmability of a processor with the hardware programmability of an FPGA


G


G.fast – A network transport protocol designed for delivering data at speeds of up to 1Gbps over distances of up to 500m. This is especially useful for “last mile” deployments within the network.  See also: DPU

GPU – Graphics Processing Unit. A specialized electronic circuit chip designed to accelerate the creation of images for output to a display device and often are used in personal computers, mobile phones, and game consoles. Because of their highly parallel structure, they are more efficient than general-purpose central processing units (CPUs) for processing large blocks of data. This has led some to try using GPUs for networking functions as well.  See also: NPU


H


Header Compression – a mechanism that compresses the IP header in a packet before the packet is transmitted in order to reduce network overhead and consumed bandwidth, while also speeding up the transmission of packets. A mechanism on the receiving end decompresses the header

Hierarchical QoS – Hierarchical Quality of Service. A feature that improves network efficiency by classifying, prioritising, and scheduling user data traffic, and by allocating network resources to services based on that prioritization

Host Bypass – A unique feature within Ethernity’s VPN Gateway for isolating traffic, packet editing, and encryption exclusively to the FPGA, entirely bypassing the CPU, which is vulnerable to breaches. The host can be authorized to receive or monitor traffic as needed, but by default, the FPGA handles all networking, flow monitoring, and security functionalities. This enhances the security within the network


I


IPSec – Internet Protocol Security. An IETF standard suite of protocols that provide data authentication, integrity, and confidentiality, and that define the encryption/decryption of packets and secure key exchange/ key management

IoT – Internet of Things. By extending internet connectivity and wireless sensors into physical devices and everyday objects, those objects can communicate and interact with each other. This enables better monitoring, remote control, automation, and machine learning. Objects within the IoT are often known as “smart”.


J


Jitter – The variance in time delay in milliseconds between data packets over a network. It is a disruption in the normal sequence of sending data packets. Jitter is an issue in latency-sensitive applications and can be detrimental to the user experience.  See also:  Latency


L


LAG – Link Aggregation Groups. Link aggregation allows a switch to treat multiple physical links between two endpoints as a single logical link. LAGs can be used to connect two switches when the traffic between them requires high bandwidth and reliability, or to provide a higher-bandwidth connection to a public network

Latency – The time it takes for a packet to travel between the sender and the recipient (in one-way latency) and back (in round trip latency). High latency creates bottlenecks in network communication and prevents data from taking full advantage of the network pipe, effectively decreasing the communication bandwidth

Libreswan – A well-respected open source IPSec VPN security implementation

Load Balancing – Efficient distribution of network traffic across multiple links to maximize speed and capacity utilization and ensures that no one link is overloaded, which could degrade performance


M


MEC – Multi-access Edge Computing. A network technology implemented in network edge nodes to enable distributed data processing closer to the end users, reducing network congestion and improving application performance. See also: Edge Computing

MEF – formerly Metro Ethernet Forum, now simply MEF or MEF Forum. A nonprofit international industry consortium that is composed of service providers, telecom carriers, network equipment vendors, and other networking companies that share an interest in Metro Ethernet, and that is dedicated to the adoption of Carrier Ethernet networks and services. The forum makes recommendations to existing standards bodies and creates specifications that are not being developed by those other bodies.

MPLS – Multi-Protocol Label Switching. A network routing technique that directs data from one node to the next based on short path labels rather than long network addresses. This avoids complex lookups routing tables and improves traffic flow speeds. It can be used with various network protocols but is most often associated with DSL


N


Network Appliance – A specialized device, often server-based, preloaded with appropriate hardware and/or software for a dedicated purpose on a network

Network Edge – the area of a service provider network closest to the end users. Depending on the use case, this could include the central office for telcos, the hub for cable operators, or C-RAN for mobile operators.  See also:  Central Office, Edge Cloud

NFV – Network Functions Virtualization. A concept that implements network functions (e.g., router, firewall) as a software package that is deployed on interconnected servers in the cloud for agility, optimization and cost-reduction purposes. NFV architecture comprises VNF (virtual networking functions), MANO (management and orchestration) and NFVI (infrastructure)

NFVI – NFV Infrastructure. NFVI enables the physical and virtual layers of the network, based on low-cost, standardized computing components:  servers – virtual or bare metal; and software – hypervisors, virtual machines, and virtual infrastructure managers

NIC – Network Interface Card (or Controller). Also called a Network Adapter. A plug-in card (or controller implemented on the motherboard) that enables a computer to transmit and receive data on a local network

NID – Network Interface Device. A device that serves as the demarcation point between a telecommunications carrier’s core network and the wires to the customer premises

NPU – Network Processing Unit. A specialized electronic circuit chip designed to handle specifically networking functions, whereas a CPU is designed to handle compute functions. An NPU is often software-programmable, but they have fallen out of favor in recent years.  See also: GPU

NVGRE – Network Virtualization using Generic Routing Encapsulation. A network technology principally backed by Microsoft that attempts to provide scalability to large cloud computing deployments by using an encapsulation to tunnel layer 2 packets over layer 3 networks. See also: Overlay Network, VxLAN


O


OAM – Operations, Administration, and Maintenance. A protocol for installing, monitoring, and troubleshooting Ethernet networks, which includes features for discovery, link monitoring, remote fault detection, and remote loopback.  See also: BFD

OEM/ODM – Original Equipment Manufacturer/Original Design Manufacturer. An OEM is a company responsible for designing and building a product according to its own specifications, which then sells the product to another company that distributes that product. An ODM is a company responsible for designing and building a product according to another company’s specifications

Offloading (or Protocol Offloading) – The process of relieving the CPU from the responsibility of handling the network stack and transferring that responsibility to the NIC. This is especially important for high-speed network interfaces, when the processing overhead on the CPU becomes significant and affects performance

OpenFlow – An open protocol that enables NICs to remotely administer the packet forwarding tables of layer 3 switches to determine the path of network packets across a network of switches. By separating the control from the actual forwarding, networks can achieve sophisticated traffic management. Also, OpenFlow allows switches from different vendors to be managed remotely using a single, open protocol.

Overlay Network – A telecommunications network that is virtually built on top of another physical network and is supported by the underlying network’s infrastructure. Most overlay networks use encapsulation to transport packets, which are then decapsulated at the receiving endpoint. This enables the network services to be decoupled from the infrastructure.  See also: NVGRE, VxLAN

OVS – Open vSwitch. An open-source software implementation of a multi-layer network switch to provide a switching stack for hardware virtualization environments. OVS adds flexibility by supporting multiple network protocols and standards


P


PCIe – Peripheral Component Interconnect Express. A high-speed serial computer expansion bus standard, designed to replace the older PCI, PCI-X and AGP bus standards. It is the common motherboard interface for servers’ network cards, graphics cards, and other expansion cards. Format specifications are maintained and developed by the PCI-SIG (PCI Special Interest Group. Servers currently support PCIe Gen 3. PCI-SIG has released specifications for PCIe Gen 4 and are currently working on specifications for PCIe Gen 5


Q


QSFP – Quad Small Form-factor Pluggable. See SFP


R


RAN (or C-RAN) – Radio Access Network (or Cloud-RAN). A cellular network architecture that consists of many cellular base stations that provide coverage for a given geographic area. Each base station processes and transmits a signal to forward data payloads to and from a mobile terminal and the core network via backhaul. C-RAN makes use of open platforms and cloud-based virtualization technology to dynamically share resources and achieve lower costs, higher reliability, lower latency, and higher bandwidth radio access networks


S


SDN – Software-Defined Networking. A concept that centralizes network control by separating the network’s forwarding (data) plane from the control plane, which is implemented through SDN controller application software.   See also:  Control Plane, Data Plane

SD-WAN – Software-Defined Wide Area Network. An application for extending networks over large distances and connecting remote branch offices to data centers and each other. SD-WAN replaces traditional branch routers with virtualization appliances that control policies and offer network overlay, which simplifies the setup process for branch personnel. It helps solve the issues with network congestion, jitter, and packet loss that often occur when networks are extended over greater distances.

SFP – Small Form-factor Pluggable. A modular interface a modular slot for a variable, media-specific transceiver that enables the connection of optical or copper cables. See also: QSFP

The following table shows the possible throughput on SFP ports and later improvements on the technology:

Interface Maximum Throughput Lanes
SFP 1Gbps 1 x 1Gbps
SFP+ 10Gbps 1 x 10Gbps
SFP28 25Gbps 1 x 25Gbps
QSFP 40Gbps 4 x 10Gbps
QSFP28 100Gbps 4 x 25Gbps

 

SmartNIC – A network interface card (network adapter) that provides programmability. This offers the ability to upgrade in the future while the NIC is already deployed in the field. SmartNICs can be based on various implementations, including FPGA-based, NPU/GPU-based, and Multi-core, each with varying degrees of flexibility, programmability, and cost.  See also: FPGA, GPU, NIC, NPU

SoC – System-on-Chip . A SoC is a silicon chip that contains one or more processor cores, as well as on-chip memory, hardware accelerator functions, peripheral functions, and others networking functions


U


uCPE – Universal Customer Premises Equipment. A single, universal platform that leverages a generic server architecture with no proprietary extensions or specialized hardware, on which service providers can simplify customer site deployments by replacing the many dedicated appliances with virtual network functions (VNFs). This extends cloud properties and technologies to the access portion of the network.  See also:  VNF


V


vCO (Virtual Central Office) – a project of OPNFV that offers a Central Office reference architecture using OpenStack and the OpenDaylight controller to support BGP networks. When combined with NFV and access infrastructure, vCO supports efficient delivery of residential, business, and mobile services. See also:  CORD

VNF – Tasks formerly carried out by proprietary, dedicated hardware devices that have been virtualized such that they run on software o on commodity hardware.  See also: NFV

VPN – Virtual Private Network. A technology that uses tunneling protocols to create  a secure connection over untrusted networks

VPN Gateway – A network appliance that connects two or more network nodes via a VPN infrastructure. It is designed to bridge the communication between two or more remote devices, sites, or networks by connecting multiple VPNs together.  See also: Network Appliance, VPN

VxLAN – Virtual Extensible Local Area Network. A network technology that attempts to provide scalability to large cloud computing deployments by using a Virtual LAN (VLAN) encapsulation to tunnel layer 2 packets over layer 3 networks. See also: Overlay Network


W


White Box – A generic equipment without a brand name


X


XGS-PON – 10 Gigabit-capable Symmetric Passive Optical Network. A next-generation standard that improves on the original G-PON network architecture and that is capable of delivering rates of up to 10Gbps over optical fiber that is shared by many subscribers. Most frequently found in the access network reaching to the user premises, XGS-PON achieves high performance while cost-effectively centralizing most of the telecommunications equipment and often displacing copper phone lines that connect premises to the phone exchange