{"id":14159,"date":"2019-08-30T21:44:53","date_gmt":"2019-08-30T21:44:53","guid":{"rendered":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/?p=14159"},"modified":"2022-10-14T18:27:51","modified_gmt":"2022-10-14T22:27:51","slug":"what-is-edge-computing","status":"publish","type":"post","link":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/iot\/what-is-edge-computing.html","title":{"rendered":"What is Edge Computing?"},"content":{"rendered":"\n

The Linux Foundation just released <\/a>V2.0 of the Open Glossary of Edge Computing<\/a>. We found it worth sharing and hope it is useful to you.<\/p>\n\n\n\n

3G, 4G, 5G<\/strong><\/p>\n\n\n\n

3rd, 4th, and 5th generation cellular technologies,\nrespectively. In simple terms, 3G represents the introduction of the smartphone\nalong with their mobile web browsers; 4G, the current generation cellular technology,\ndelivers true broadband internet access to mobile devices; the coming 5G\ncellular technologies will deliver massive bandwidth and reduced latency to\ncellular systems, supporting a range of devices from smartphones to autonomous\nvehicles and large-scale IoT. Infrastructure edge computing is considered a key\nbuilding block for 5G. <\/p>\n\n\n\n

Access Edge Layer<\/strong><\/p>\n\n\n\n

The sublayer of infrastructure edge closest to the end user\nor device, zero or one hops from the last mile network. For example, an edge\ndata center deployed at a cellular network site. The Access Edge Layer functions\nas the front line of the infrastructure edge and may connect to an aggregation\nedge layer higher in the hierarchy. See also: Aggregation Edge Layer<\/p>\n\n\n\n

Access Network<\/strong><\/p>\n\n\n\n

A network that connects subscribers and devices to their\nlocal service provider. It is contrasted with the core network which connects\nservice providers to one another. The access network connects directly to the\ninfrastructure edge. See also: Infrastructure Edge <\/p>\n\n\n\n

Aggregation Edge Layer<\/strong><\/p>\n\n\n\n

The layer of infrastructure edge one hop away from the\naccess layer. Can exist as either a medium scale data center in a single\nlocation or may be formed from multiple interconnected micro data centers to form\na hierarchical topology with the access edge to allow for greater\ncollaboration, workload failover and scalability than access edge alone. See\nalso: Access Layer Edge <\/p>\n\n\n\n

Base Station<\/strong><\/p>\n\n\n\n

A network element in the RAN which is responsible for the\ntransmission and reception of radio signals in one or more cells to or from\nuser equipment. A base station can have an integrated antenna or may be connected\nto an antenna array by feeder cables. Uses specialized digital signal\nprocessing and network function hardware. In modern RAN architectures, the base\nstation may be split into multiple functional blocks operating in software for\nflexibility, cost and performance. See also: Baseband Unit (BBU), Cloud RAN\n(C-RAN) <\/p>\n\n\n\n

Baseband Unit (BBU)<\/strong><\/p>\n\n\n\n

A component of the Base Station which is responsible for\nbaseband radio signal processing. Uses specialized hardware for digital signal\nprocessing. In a C-RAN architecture, the functions of the BBU may be operated\nin software as a VNF. See also: Base Station <\/p>\n\n\n\n

Central Office (CO)<\/strong><\/p>\n\n\n\n

An aggregation point for telecommunications infrastructure\nwithin a defined geographical area where telephone companies historically\nlocated their switching equipment. Physically designed to house telecommunications\ninfrastructure equipment but typically not suitable to house compute, data\nstorage and network resources on the scale of an edge data center due to their\ninadequate flooring, as well as their heating, cooling, ventilation, fire\nsuppression and power delivery systems. See also: Central Office Re-architected\nas a Data Center (CORD) <\/p>\n\n\n\n

Central Office Re-architected as Data Center (CORD)<\/strong><\/p>\n\n\n\n

An initiative to deploy data center-level compute and data\nstorage capability within the CO. Although this is often logical topologically,\nCO facilities are typically not physically suited to house compute, data storage\nand network resources on the scale of an edge data center due to their\ninadequate flooring, as well as their heating, cooling, ventilation, fire\nsuppression and power delivery systems. See also: Central Office (CO) <\/p>\n\n\n\n

Centralized Data Center<\/strong><\/p>\n\n\n\n

A large, often hyperscale physical structure and logical\nentity which houses large compute, data storage and network resources which are\ntypically used by many tenants concurrently due to their scale. Located a\nsignificant geographical distance from the majority of their users and often\nused for cloud computing. See also: Cloud Computing <\/p>\n\n\n\n

Cloud Computing<\/strong><\/p>\n\n\n\n

A system to provide on-demand access to a shared pool of\ncomputing resources, including network servers, storage, and computation\nservices. Typically utilises a small number of large centralized data centers\nand regional data centers today. See also: Centralized Data Center <\/p>\n\n\n\n

Cloud Native Network Function (CNF)<\/strong><\/p>\n\n\n\n

A Virtualized Network Function (VNF) built and deployed\nusing cloud native technologies. These technologies include containers, service\nmeshes, microservices, immutable infrastructure and declarative APIs that allow\ndeployment in public, private and hybrid cloud environments through loosely\ncoupled and automated systems. See also: Virtualized Network Function (VNF) <\/p>\n\n\n\n

Cloud Node<\/strong><\/p>\n\n\n\n

A compute node, such as an individual server or other set of\ncomputing resources, operated as part of a cloud computing infrastructure.\nTypically resides within a centralized data center. See also: Edge Node <\/p>\n\n\n\n

Cloud RAN (C-RAN)<\/strong><\/p>\n\n\n\n

An evolution of the RAN that allows the functionality of the\nwireless base station to be split into two components: A Remote Radio Head\n(RRH) and a centralized BBU. Rather than requiring a BBU to be located with\neach cellular radio antenna, C-RAN allows the BBUs to operate at some distance\nfrom the tower, at an aggregation point, often referred to as a DAS hub.\nCo-locating multiple BBUs in an aggregation facility creates infrastructure\nefficiencies and allows for a more graceful evolution to Cloud RAN. In a C-RAN\narchitecture, tasks performed by a legacy base station are often performed as\nVNFs operating on infrastructure edge micro data centers on general-purpose\ncompute hardware. These tasks must be performed at high levels of performance\nand with as little latency as possible, requiring the use of infrastructure\nedge computing at the cellular network site to support them. See also:\nInfrastructure Edge <\/p>\n\n\n\n

Cloud Service Provider (CSP)<\/strong><\/p>\n\n\n\n

An organization which operates typically large-scale cloud\nresources comprised of centralized and regional data centers. Most frequently\nused in the context of the public cloud. May also be referred to as a Cloud\nService Operator (CSO). See also: Cloud Computing <\/p>\n\n\n\n

Cloudlet<\/strong><\/p>\n\n\n\n

In academic circles, this term refers to a mobility-enhanced\npublic or private cloud at the infrastructure edge, as popularized by Mahadev\nSatyanarayanan of Carnegie Mellon university. In the context of CDNs such as\nAkamai, cloudlet refers to the practice of deploying self-serviceable\napplications at CDN nodes. See also: Edge Cloud <\/p>\n\n\n\n

Co-Location<\/strong><\/p>\n\n\n\n

The process of deploying compute, data storage and network\ninfrastructure owned or operated by different parties in the same physical\nlocation, such as within the same physical structure. Distinct from Shared\nInfrastructure as co-location does not require infrastructure such as an edge\ndata center to have multiple tenants or users. See also: Shared Infrastructure <\/p>\n\n\n\n

Computational Offloading<\/strong><\/p>\n\n\n\n

An edge computing use case where tasks are offloaded from an\nedge device to the infrastructure edge for remote processing. Computational\noffloading seeks, for example, performance improvements and energy savings for\nmobile devices by offloading computation to the infrastructure edge with the\ngoal of minimizing task execution latency and mobile device energy consumption.\nComputational offloading also enables new classes of mobile applications that\nwould require computational power and storage capacity that exceeds what the\ndevice alone is capable of employing (e.g., untethered Virtual Reality). In\nother cases, workloads may be offloaded from a centralized to an edge data\ncenter for performance. See also: Traffic Offloading <\/p>\n\n\n\n

Content Delivery Network (CDN)<\/strong><\/p>\n\n\n\n

A distributed system positioned throughout the network that\npositions popular content such as streaming video at locations closer to the\nuser than are possible with a traditional centralized data center. Unlike a\ndata center, a CDN node will typically contain data storage without dense\ncompute resources. When using infrastructure edge computing CDN nodes operate\nin software at edge data centers. See also: Edge Data Center, Traffic\nOffloading <\/p>\n\n\n\n

Core Network<\/strong><\/p>\n\n\n\n

The layer of the service provider network which connects the\naccess network and the devices connected to it to other network operators and\nservice providers, such that data can be transmitted to and from the internet\nor to and from other networks. May be multiple hops away from infrastructure\nedge computing resources. See also: Access Network <\/p>\n\n\n\n

Customer-Premises Equipment (CPE)<\/strong><\/p>\n\n\n\n

The local piece of equipment such as a cable network modem\nwhich allows the subscriber to a network service to connect to the access\nnetwork of the service provider. Typically one hop away from infrastructure edge\ncomputing resources. See also: Access Network <\/p>\n\n\n\n

Data Center<\/strong><\/p>\n\n\n\n

A purpose-designed structure that is intended to house\nmultiple high-performance compute and data storage nodes such that a large\namount of compute, data storage and network resources are present at a single\nlocation. This often entails specialized rack and enclosure systems,\npurpose-built flooring, as well as suitable heating, cooling, ventilation,\nsecurity, fire suppression and power delivery systems. May also refer to a\ncompute and data storage node in some contexts. Varies in scale between a\ncentralized data center, regional data center and edge data center. See also:\nCentralized Data Center <\/p>\n\n\n\n

Data Gravity<\/strong><\/p>\n\n\n\n

The concept that data is not free to move over a network and\nthat the cost and difficulty of doing so increases as both the volume of data\nand the distance between network endpoints grows, and that pplications will\ngravitate to where their data is located. Observed with applications requiring\nlarge-scale data ingest. See also: Edge-Native Application <\/p>\n\n\n\n

Data Ingest<\/strong><\/p>\n\n\n\n

The process of taking in a large amount of data for storage\nand subsequent processing. An example is an edge data center storing much\nfootage for a video surveillance network which it must then process to identify\npersons of interest. See also: Edge-Native Application <\/p>\n\n\n\n

Data Reduction<\/strong><\/p>\n\n\n\n

The process of using an intermediate point between the\nproducer and the ultimate recipient of data to intelligently reduce the volume\nof data transmitted, without losing the meaning of the data. An example is a\nsmart data deduplication system. See also: Edge-Native Application <\/p>\n\n\n\n

Data Sovereignty<\/strong><\/p>\n\n\n\n

The concept that data is subject to the laws and regulations\nof the country, state, industry it is in, or the applicable legal framework\ngoverning its use and movement. See also: Edge-Native Application <\/p>\n\n\n\n

Decision Support<\/strong><\/p>\n\n\n\n

The use of intelligent analysis of raw data to produce a\nrecommendation which is meaningful to a human operator. An example is\nprocessing masses of sensor data from IoT devices within the infrastructure edge\nto produce a single statement that is interpreted by and meaningful to a human\noperator or higher automated system. See also: Edge-Native Application <\/p>\n\n\n\n

Device Edge<\/strong><\/p>\n\n\n\n

Edge computing capabilities on the device or user side of\nthe last mile network. Often depends on a gateway or similar device in the\nfield to collect and process data from devices. May also use limited spare compute\nand data storage capability from user devices such as smartphones, laptops and\nsensors to process edge computing workloads. Distinct from infrastructure edge\nas it uses device resources. See also: Infrastructure Edge <\/p>\n\n\n\n

Device Edge Cloud<\/strong><\/p>\n\n\n\n

An extension of the edge cloud concept where certain\nworkloads can be operated on resources available at the device edge. Typically\ndoes not provide cloud-like elastically-allocated resources, but may be optimal\nfor zero-latency workloads. See also: Edge Cloud <\/p>\n\n\n\n

Distributed Antenna System (DAS) Hub<\/strong><\/p>\n\n\n\n

A location which serves as an aggregation point for many\npieces of radio communications equipment, typically in support of cellular\nnetworks. May contain or be directly attached to an edge data center deployed\nat the infrastructure edge. See also: Edge Data Center <\/p>\n\n\n\n

Edge Availability Zones<\/strong><\/p>\n\n\n\n

Isolated locations within the infrastructure edge, from\nwhich application workloads and other edge cloud services operate. An\napplication may intend to be run on a specific edge availability zone for performance,\nreliability and data sovereignty reasons, or be dynamically allocated to it by\na workload orchestration system. Edge availability zones are linked to others\nfor failover. See also: Location Awareness <\/p>\n\n\n\n

Edge Cloud<\/strong><\/p>\n\n\n\n

Cloud-like capabilities located at the infrastructure edge,\nincluding from the user perspective access to elastically-allocated compute,\ndata storage and network resources. Often operated as a seamless extension of a\ncentralized public or private cloud, constructed from micro data centers\ndeployed at the infrastructure edge. See also: Cloud Computing <\/p>\n\n\n\n

Edge Computing<\/strong><\/p>\n\n\n\n

The delivery of computing capabilities to the logical\nextremes of a network in order to improve the performance, operating cost and\nreliability of applications and services. By shortening the distance between\ndevices and the cloud resources that serve them, and also reducing network\nhops, edge computing mitigates the latency and bandwidth constraints of today’s\nInternet, ushering in new classes of applications. In practical terms, this\nmeans distributing new resources and software stacks along the path between\ntoday’s centralized data centers and the increasingly large number of devices\nin the field, concentrated, in particular, but not exclusively, in close\nproximity to the last mile network, on both the infrastructure and device sides.\nSee also: Infrastructure Edge, Device Edge, Last Mile <\/p>\n\n\n\n

Edge Data Center<\/strong><\/p>\n\n\n\n

A data center which is capable of being deployed as close as\npossible to the edge of the network, in comparison to traditional centralized\ndata centers. Capable of performing the same functions as centralized data\ncenters although at smaller scale individually. Because of the unique\nconstraints created by highly-distributed physical locations, edge data centers\noften adopt autonomic operation, multitenancy, distributed and local resiliency\nand open standards. Edge refers to the location at which these data centers are\ntypically deployed. Their scale can be defined as micro, ranging from 50 to 150\nkW of capacity. Multiple edge data centers may interconnect to provide capacity\nenhancement, failure mitigation and workload migration within the local area,\noperating as a virtual data center. See also: Virtual Data Center <\/p>\n\n\n\n

Edge Node<\/strong><\/p>\n\n\n\n

A compute node, such as an individual server or other set of\ncomputing resources, operated as part of an edge computing infrastructure.\nTypically resides within an edge data center operating at the infrastructure\nedge, and is therefore physically closer to its intended users than a cloud\nnode in a centralized data center. See also: Cloud Node <\/p>\n\n\n\n

Edge-Enhanced Application<\/strong><\/p>\n\n\n\n

An application which is capable of operating in a\ncentralized data center, but which gains performance, typically in terms of\nlatency, or functionality advantages when operated using edge computing. These applications\nmay be adapted from existing applications which operate in a centralized data\ncenter, or may require no changes. See also: Edge-Native Application <\/p>\n\n\n\n

Edge-Native Application<\/strong><\/p>\n\n\n\n

An application which is impractical or undesirable to\noperate in a centralized data center. This can be due to a range of factors\nfrom a requirement for low latency and the movement of large volumes of data,\nthe local creation and consumption of data, regulatory constraints, and other\nfactors. These applications are typically developed for and operate on the edge\ndata centers at the infrastructure edge. May use the infrastructure edge to provide\nlarge-scale data ingest, data reduction, real-time decision support, or to solve\ndata sovereignty issues. See also: Edge-Enhanced Application <\/p>\n\n\n\n

Fog Computing<\/strong><\/p>\n\n\n\n

A distributed computing concept where compute and data\nstorage resource, as well as applications and their data, are positioned in the\nmost optimal place between the user and Cloud with the goal of improving performance\nand redundancy. Fog computing workloads may be run across the gradient of\ncompute and data storage resource from Cloud to the infrastructure edge. The\nterm fog computing was originally coined by Cisco. Can utilize centralized,\nregional and edge data centers. See also: Workload Orchestration <\/p>\n\n\n\n

Infrastructure Edge<\/strong><\/p>\n\n\n\n

Edge computing capability, typically in the form of one or\nmore edge data centers, which is deployed on the operator side of the last mile\nnetwork. Compute, data storage and network resources positioned at the infrastructure\nedge allow for cloud-like capabilities similar to those found in centralized\ndata centers such as the elastic allocation of resources, but with lower\nlatency and lower data transport costs due to a higher degree of locality to\nuser than with a centralized or regional data center. See also: Device Edge <\/p>\n\n\n\n

Interconnection<\/strong><\/p>\n\n\n\n

The linkage, often via fiber optic cable, that connects one\nparty’s network to another, such as at an internet peering point, in a meet-me\nroom or in a carrier hotel. The term may also refer to connectivity between two\ndata centers or between tenants within a data center, such as at an edge meet\nme room. See also: Meet Me Room <\/p>\n\n\n\n

Internet Edge<\/strong><\/p>\n\n\n\n

A sub-layer within the infrastructure edge where the interconnection\nbetween the infrastructure edge and the internet occurs. Contains the edge meet\nme room and other equipment used to provide this high-performance level of\ninterconnectivity. See also: Interconnection <\/p>\n\n\n\n

Internet Exchange Point (IXP)<\/strong><\/p>\n\n\n\n

Places in which large network providers converge for the\ndirect exchange of traffic. A typical service provider will access tier 1\nglobal providers and their networks via IXPs, though they also serve as meet points\nfor like networks. IXPs are sometimes referred to as Carrier Hotels because of\nthe many different organizations available for traffic exchange and peering.\nThe internet edge may often connect to an IXP. See also: Internet Edge <\/p>\n\n\n\n

IP Aggregation<\/strong><\/p>\n\n\n\n

The use of compute, data storage and network resources at a\nlayer beyond the infrastructure edge to separate and route network data\nreceived from the cellular network RAN. Although it does not provide the\nimproved user experience of local breakout, IP aggregation can improve\nperformance and network utilization when compared to traditional cellular\nnetwork architectures. See also: Local Breakout <\/p>\n\n\n\n

Jitter<\/strong><\/p>\n\n\n\n

The variation in network data transmission latency observed\nover a period of time. Measured in terms of milliseconds as a range from the\nlowest to highest observed latency values which are recorded over the measurement\nperiod. A key metric for real-time applications such as VoIP, autonomous\ndriving and online gaming which assume little latency variation is present and\nare sensitive to changes in this metric. See also: Quality of Service (QoS) <\/p>\n\n\n\n

Last Mile<\/strong><\/p>\n\n\n\n

The segment of a telecommunications network that connects\nthe service provider to the customer. The type of connection and distance\nbetween the customer and the infrastructure determines the performance and\nservices available to the customer. The last mile is part of the access\nnetwork, and is also the network segment closest to the user that is within the\ncontrol of the service provider. Examples of this include cabling from a DOCSIS\nheadend site to a cable modem, or the wireless connection between a customer’s\nmobile device and a cellular network site. See also: Access Network <\/p>\n\n\n\n

Latency<\/strong><\/p>\n\n\n\n

In the context of network data transmission, the time taken\nby a unit of data (typically a frame or packet) to travel from its originating\ndevice to its intended destination. Measured in terms of milliseconds at single\nor repeated points in time between two or more endpoints. A key metric of\noptimizing the modern application user experience. Distinct from jitter which\nrefers to the variation of latency over time. Sometimes expressed as Round Trip\nTime (RTT). See also: Quality of Service (QoS) <\/p>\n\n\n\n

Latency Critical Application<\/strong><\/p>\n\n\n\n

An application that will fail to function or will function\ndestructively if latency exceeds certain thresholds. Latency critical\napplications are typically responsible for real-time tasks such as supporting\nan autonomous vehicle or controlling a machine-to-machine process. Unlike\nLatency Sensitive Applications, exceeding latency requirements will often\nresult in application failure. See also: Edge-Native Application <\/p>\n\n\n\n

Latency Sensitive Application<\/strong><\/p>\n\n\n\n

An application in which reduced latency improves\nperformance, but which can still function if latency is higher than desired.\nUnlike a Latency Critical Application, exceeding latency targets will typically\nnot result in application failure, though may result in a diminished user\nexperience. Examples include image processing and bulk data transfers. See\nalso: Edge-Enhanced Application <\/p>\n\n\n\n

Local Breakout<\/strong><\/p>\n\n\n\n

The use of compute, data storage and network resources at\nthe infrastructure edge to separate and route network data received from the\ncellular network RAN at the earliest point possible. If local breakout is not\nused, this data may be required to take a longer path to a local CO or other\naggregation point before it can be routed on to the internet or another network.\nImproves cellular network QoS for the user and network utilization for the\nnetwork operator. See also: Traffic Offloading, IP Aggregation <\/p>\n\n\n\n

Location Awareness<\/strong><\/p>\n\n\n\n

The ability of an application workload to know where it is\noperating, in terms of its physical and logical location as well as the type\nand quantity of compute, data storage and network resources which are |available\nto it. May also refer to the ability of an application workload to determine\nthe location of its user, allowing the workload to move to the nearest point of\nthe infrastructure edge. See also: Workload Orchestration <\/p>\n\n\n\n

Location-Based Node Selection<\/strong><\/p>\n\n\n\n

A method of selecting an optimal edge node on which to run a\nworkload based on the node’s physical location in relation to the device’s\nphysical location with the aim of improving application workload performance. A\npart of workload orchestration. See also: Workload Orchestration <\/p>\n\n\n\n

Meet Me Room<\/strong><\/p>\n\n\n\n

An area within an edge data center where tenants and\ntelecommunications providers can interconnect with each other and other edge\ndata centers in the same fashion as they would in a traditional meet me room\nenvironment, except at the edge. See also: Interconnection <\/p>\n\n\n\n

Micro Modular Data Center (MMDC)<\/strong><\/p>\n\n\n\n

A data center which applies the modular data center concept\nat a smaller scale, typically from 50 to 150 kW in capacity. Takes a number of\npossible forms including a rackmount cabinet which may be deployed indoors or\noutdoors as required. Like larger modular data centers, micro modular data\ncenters are capable of being combined with other data centers to increase\navailable resource in an area. See also: Edge Data Center <\/p>\n\n\n\n

Mobile Network Operator (MNO)<\/strong><\/p>\n\n\n\n

The operator of a cellular network, who is typically\nresponsible for the physical assets such as RAN equipment and network sites\nrequired for the network to be deployed and operate effectively. Distinct from\nMVNO as the MNO is responsible for physical network assets. May include those\nedge data centers deployed at the infrastructure edge positioned at or\nconnected to their cell sites under these assets. Typically also a service\nprovider providing access to other networks and the internet. See also: Mobile Virtual\nNetwork Operator (MVNO) <\/p>\n\n\n\n

Mobile Virtual Network Operator (MVNO)<\/strong><\/p>\n\n\n\n

A service provider similar to an MNO with the distinction\nthat the MVNO does not own or often operate their own cellular network\ninfrastructure. Although they will not own an edge data center deployed at the infrastructure\nedge connected to a cell site they may be using, the MVNO may be a tenant\nwithin that edge data center. See also: Mobile Network Operator (MNO) <\/p>\n\n\n\n

Modular Data Center (MDC)<\/strong><\/p>\n\n\n\n

A method of data center deployment which is designed for\nportability. High-performance compute, data storage and network capability is\ninstalled within a portable structure such as a shipping container which can\nthen be transported to where it is required. These data centers can be combined\nwith existing data centers or other modular data centers to increase the local\nresources available as required. See also: Micro Modular Data Center (MMDC) <\/p>\n\n\n\n

Multi-access Edge Computing (MEC)<\/strong><\/p>\n\n\n\n

An open application framework sponsored by ETSI to support\nthe development of services tightly coupled with the Radio Access Network\n(RAN). Formalized in 2014, MEC seeks to augment 4G and 5G wireless base\nstations with a standardized software platform, API and programming model for\nbuilding and deploying applications at the edge of the wireless networks. MEC\nallows for the deployment of services such as radio-aware video optimization,\nwhich utilizes caching, buffering and real-time transcoding to reduce congestion\nof the cellular network and improve the user experience. Originally known as Mobile\nEdge Computing, the ETSI working group renamed itself to Multi-Access Edge\nComputing in 2016 in order to acknowledge their ambition to expand MEC beyond\ncellular to include other access technologies. Utilizes edge data centers\ndeployed at the infrastructure edge. See also: Infrastructure Edge <\/p>\n\n\n\n

Network Function Virtualization (NFV)<\/strong><\/p>\n\n\n\n

The migration of network functions from embedded services\ninside proprietary hardware appliances to software-based VNFs running on\nstandard x86 and ARM servers using industry standard virtualization and cloud\ncomputing technologies. In many cases NFV processing and data storage will\noccur at the edge data centers that are connected directly to the local\ncellular site, within the infrastructure edge. See also: Virtualized Network\nFunction (VNF) <\/p>\n\n\n\n

Network Hop<\/strong><\/p>\n\n\n\n

A point at which the routing or switching of data in transit\nacross a network occurs; a decision point, typically at an aggregating device such\nas a router, as to the next immediate destination of that data. Reducing the\nnumber of network hops between user and application is one of the primary\nperformance goals of edge computing. See also: Edge Computing <\/p>\n\n\n\n

Northbound vs Southbound (and east\/west)<\/strong><\/p>\n\n\n\n

The direction in which data is transmitted when viewed in\nthe context of a hierarchy where the cloud is at the top, the infrastructure\nedge is in the middle, and the device edge is at the bottom. Northbound and\nsouthbound data transmission is defined as flowing to and from the cloud or\nedge data center accordingly. Eastbound and westbound data transmission is\ndefined as occurring between data centers at the same hierarchical layer, for\npurposes such as workload migration or data replication. This may occur between\ncentralized or between edge data centers. See also: Virtual Data Center <\/p>\n\n\n\n

Offload Processing<\/strong><\/p>\n\n\n\n

The use of compute, data storage and network resources at\nthe infrastructure edge to process workloads which have been handed off to the\ninfrastructure edge by other layers of the edge cloud system. An example is an\nedge device offloading a complex processing task to the edge data centers within\nthe infrastructure edge to conserve its own battery life and limited resources.\nSee also: Workload Orchestration <\/p>\n\n\n\n

Over-the-Top Service Provider (OTT)<\/strong><\/p>\n\n\n\n

An application or service provider who does not own or\noperate the underlying network, and in some cases data center, infrastructure\nrequired to deliver their application or service to users. Streaming video services\nand MVNOs are examples of OTT service providers that are very common today.\nOften data center tenants. See also: Mobile Virtual Network Operator (MVNO) <\/p>\n\n\n\n

Point of Presence (PoP)<\/strong><\/p>\n\n\n\n

A point in their network infrastructure where a service\nprovider allows connectivity to their network by users or partners. In the\ncontext of edge computing, in many cases a PoP will be within an edge meet me\nroom if an IXP is not within the local area. The edge data center would connect\nto a PoP which then connects to an IXP. See also: Interconnection <\/p>\n\n\n\n

Quality of Experience (QoE)<\/strong><\/p>\n\n\n\n

The advanced use of QoS principles to perform more detailed\nand nuanced measurements of application and network performance with the goal\nof further improving the user experience of the application and network. Also\nrefers to systems which will proactively measure performance and adjust\nconfiguration or load balancing as required. Can therefore be considered a\ncomponent of workload orchestration, operating as a high-fidelity data source\nfor an intelligent orchestrator. See also: Workload Orchestration <\/p>\n\n\n\n

Quality of Service (QoS)<\/strong><\/p>\n\n\n\n

A measure of how well the network and data center\ninfrastructure is serving a particular application, often to a specific user.\nThroughput, latency and jitter are all key QoS measurement metrics which edge computing\nseeks to improve for many different types of application, from realtime to bulk\ndata transfer use cases. See also: Edge Computing <\/p>\n\n\n\n

Radio Access Network (RAN)<\/strong><\/p>\n\n\n\n

A wireless variant of the access network, typically\nreferring to a cellular network such as 3G, 4G or 5G. The 5G RAN will be\nsupported by compute, data storage and network resources at the infrastructure\nedge as it utilises NFV and C-RAN. See also: Cloud RAN (C-RAN) <\/p>\n\n\n\n

Regional Data Center<\/strong><\/p>\n\n\n\n

A data center positioned in scale between a centralized data\ncenter and an edge data center. Significantly physically further away from end\nusers than an edge data center, but closer to them than a centralized data\ncenter. Also referred to as a metropolitan data center in some contexts. Part\nof traditional cloud computing. See also: Cloud Computing <\/p>\n\n\n\n

Service Provider<\/strong><\/p>\n\n\n\n

An organization which provides customers with access to its\nnetwork, typically with the goal of providing that customer access to the\ninternet. A customer will usually connect to the access network of the service provider\nfrom their side of the last mile. See also: Access Network <\/p>\n\n\n\n

Shared Infrastructure<\/strong><\/p>\n\n\n\n

The use of a single piece of compute, data storage and\nnetwork resources by multiple parties, for example two organizations each using\nhalf of a single edge data center, unlike co-location where each party possesses\ntheir own infrastructure. See also: Co-Location <\/p>\n\n\n\n

Software Edge<\/strong><\/p>\n\n\n\n

From a software development and application deployment\nperspective, the point physically closest to the end user where application\nworkloads can be deployed. Depending on the application workload and the\ncurrent availability of computing resources, this point may be at the device\nedge, but will typically be within the infrastructure edge due to its\ncloud-like capability to provide elastic resources. See also: Infrastructure\nEdge <\/p>\n\n\n\n

Throughput<\/strong><\/p>\n\n\n\n

In the context of network data transmission, the amount of\ndata per second that is able to be transmitted between two or more endpoints.\nMeasured in terms of bits per second typically at megabit or gigabit scales as\nrequired. Although a minimum level of throughput is often required for\napplications to function, after this latency typically becomes the\napplication-limiting and user experience-damaging factor. See also: Quality of\nService (QoS) <\/p>\n\n\n\n

Traffic Offloading<\/strong><\/p>\n\n\n\n

The use of compute, data storage and network resources at the\ninfrastructure edge to route network data in preference of another network\npath. This may be seen when the infrastructure edge is providing local\nbreakout, and therefore provides a superior performance network path. See also:\nLocal Breakout <\/p>\n\n\n\n

Vehicle 2 Infrastructure (V2I)<\/strong><\/p>\n\n\n\n

The collection of technologies used to allow a connected or\nautonomous vehicle to connect to its supporting infrastructure such as an\nmachine vision and route finding application operating in an edge data center\nat the infrastructure edge. Typically uses newer cellular communications\ntechnologies such as 5G as its access network. See also: Access Network <\/p>\n\n\n\n

Virtual Data Center<\/strong><\/p>\n\n\n\n

A virtual entity constructed from multiple physical edge\ndata centers such that they can be considered externally as one. Within the\nvirtual data center, workloads can be intelligently placed within specific edge\ndata centers or availability zones as required based on load balancing,\nfailover or operator preference. In such a configuration, edge data centers are\ninterconnected by low latency networking and are designed to create a redundant\nand resilient edge computing infrastructure. See also: Edge Data Center <\/p>\n\n\n\n

Virtualized Network Function (VNF)<\/strong><\/p>\n\n\n\n

A software-based network function operating on\ngeneral-purpose compute resources which is used by NFV in place of dedicated\nphysical equipment. In many cases, several VNFs will operate on an edge data\ncenter at the infrastructure edge. Workload Orchestration An intelligent system\nwhich dynamically determines the optimal location, time and priority for application\nworkloads to be processed on the range of compute, data storage and network\nresources from the centralized and regional data centers to the resources\navailable at both the infrastructure edge and device edge. Workloads may be\ntagged with specific performance and cost requirements which determines where\nthey are to be operated as resources that meet them are available for use. See\nalso: Network Functions Virtualization (NFV) <\/p>\n\n\n\n

Workload Orchestration<\/strong><\/p>\n\n\n\n

The process of determining where and how an application\nworkload should be processed on the n-tier gradient of compute, data storage\nand network resources provided by the edge cloud. In most cases, workload\norchestration will be performed by an automated system which takes into account\nthe performance, time and cost requirements of an application operator at that\npoint.me layer. <\/p>\n\n\n\n

xHaul (“crosshaul”)<\/strong>The high-speed\ninterconnection of two or more pieces of network or data center infrastructure.\nBackhaul and fronthaul are the most common examples of xhaul today. Other\nexamples include the connectivity between infrastructure edge sites, and\nbetween the infrastructure edge and any other significant network or data\ncenter infrastructure within the local area at the same layer. <\/p>\n\n\n\n

Learn more about Edge, IoT and Smart Buildings at IoT Evolution<\/a> Expo, part\nof the ITEXPO <\/a>TECHSUPERSHOW, Feb\n12-14, 2020 in Fort Lauderdale, Florida. <\/p>\n\n\n\n

Learn more about Edge, IoT and Smart Buildings at IoT Evolution<\/a> Expo, part of the ITEXPO <\/a>TECHSUPERSHOW, Feb 12-14, 2020 in Fort Lauderdale, Florida.<\/p>\n\n\n\n

\"\"<\/figure>\n","protected":false},"excerpt":{"rendered":"

The Linux Foundation just released V2.0 of the Open Glossary of Edge Computing. We found it worth sharing and hope it is useful to you. 3G, 4G, 5G 3rd, 4th, and 5th generation cellular technologies, respectively. In simple terms, 3G represents the introduction of the smartphone along with their mobile web browsers; 4G, the current<\/p>\n","protected":false},"author":44,"featured_media":14161,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[185],"tags":[2330,2010,1798,789,2329],"post_mailing_queue_ids":[],"_links":{"self":[{"href":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/wp-json\/wp\/v2\/posts\/14159"}],"collection":[{"href":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/wp-json\/wp\/v2\/users\/44"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/wp-json\/wp\/v2\/comments?post=14159"}],"version-history":[{"count":2,"href":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/wp-json\/wp\/v2\/posts\/14159\/revisions"}],"predecessor-version":[{"id":14162,"href":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/wp-json\/wp\/v2\/posts\/14159\/revisions\/14162"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/wp-json\/wp\/v2\/media\/14161"}],"wp:attachment":[{"href":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/wp-json\/wp\/v2\/media?parent=14159"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/wp-json\/wp\/v2\/categories?post=14159"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.tmcnet.com\/blog\/rich-tehrani\/wp-json\/wp\/v2\/tags?post=14159"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}