Edge computing: Difference between revisions
From HandWiki
imported>BotanyGa correction |
update |
||
| Line 1: | Line 1: | ||
{{Short description|Distributed computing paradigm}} | {{Short description|Distributed computing paradigm}} | ||
'''Edge computing''' is a [[Distributed computing|distributed computing]] | '''Edge computing''' is a [[Distributed computing|distributed computing]] model that brings computation and data storage closer to the sources of data. More broadly, it refers to any design that pushes computation physically closer to a user, so as to reduce the [[Engineering:Latency (engineering)|latency]] compared to when an application runs on a centralized [[Data center|data center]].<ref>{{cite web |last1=Gill |first1=Bob |last2=Smith |first2=David |date=2018-09-14 |df=mdy |title=The Edge Completes the Cloud: A Gartner Trend Insight Report |url=https://emtemp.gcom.cloud/ngw/globalassets/en/doc/documents/3889058-the-edge-completes-the-cloud-a-gartner-trend-insight-report.pdf |url-status=live |archive-url=https://web.archive.org/web/20201218093221/https://emtemp.gcom.cloud/ngw/globalassets/en/doc/documents/3889058-the-edge-completes-the-cloud-a-gartner-trend-insight-report.pdf |archive-date=2020-12-18 |access-date=2026-02-02 |work=Gartner |publisher=[[Company:Gartner|Gartner]] |id=G00360847}}</ref> | ||
The | The term began being used in the 1990s to describe [[Content delivery network|content delivery network]]s—these were used to deliver website and video content from servers located near users.<ref name=":02">{{cite journal |last1=Dilley |first1=John |last2=Maggs |first2=Bruce |last3=Parikh |first3=Jay |last4=Prokop |first4=Harald |last5=Sitaraman |first5=Ramesh |last6=Weihl |first6=Bill |date=2002-10-31 |df=mdy |url=https://people.cs.umass.edu/~ramesh/Site/PUBLICATIONS_files/DMPPSW02.pdf |title=Globally Distributed Content Delivery |url-status=live |archive-url=https://web.archive.org/web/20170809231307/http://people.cs.umass.edu/~ramesh/Site/PUBLICATIONS_files/DMPPSW02.pdf |archive-date=2017-08-09 |access-date=2026-02-02 |work=[[IEEE Internet Computing]] |volume=6 |issue=5 |pages=50-58 |issn=1089-7801 |doi=10.1109/MIC.2002.1036038}}</ref> In the early 2000s, these systems expanded their scope to hosting other applications,<ref name=":1">{{cite journal |last1=Nygren |first1=Erik |last2=Sitaraman |first2=Ramesh K. |last3=Sun |first3=Jennifer |date=2010-08-17 |df=mdy |title=The Akamai Network: A Platform for High-Performance Internet Applications |work=ACM SIGOPS Operating Systems Review |url=https://www.akamai.com/site/en/documents/research-paper/the-akamai-network-a-platform-for-high-performance-internet-applications-technical-publication.pdf |url-status=live |volume=44 |issue=3 |pages=2–19 |issn=0163-5980 |doi=10.1145/1842733.1842736 |s2cid=207181702 |archive-url=https://web.archive.org/web/20120913205810/http://www.akamai.com/dl/technical_publications/network_overview_osr.pdf |archive-date=2012-09-13 |access-date=2026-02-02 |quote=See Section 6.2: Distributing Applications to the Edge}}</ref> leading to early edge computing services.<ref>{{cite conference |last1=Davis |first1=Andy |last2=Parikh |first2=Jay |last3=Weihl |first3=William E. |book-title=Proceedings of the 13th international World Wide Web conference on Alternate track papers & posters |conference=WWW Alt. '04 |title=EdgeComputing: Extending Enterprise Applications to the Edge of the Internet |s2cid=578337 |date=2004 |page=180-187 |doi=10.1145/1013367.1013397 |isbn=1-58113-912-8 |url=https://www.akamai.com/site/en/documents/research-paper/edgecomputing-extending-enterprise-applications-to-the-edge-of-the-internet-technical-publication.pdf |access-date=2026-02-02}}</ref> These services could do things like find dealers, manage shopping carts, gather real-time data, and place ads. | ||
[[Internet of things]] (IoT) is | The [[Internet of things]] (IoT), where devices are connected to the [[Internet]], is often linked with edge computing.<ref>{{cite web |last1=Gill |first1=Bob |date=2021-11-03 |df=mdy |title=2021 Strategic Roadmap for Edge Computing |url=https://www.gartner.com/doc/reprints?id=1-24JFAZOO&ct=201104&st=sb |url-status=dead |archive-url=https://web.archive.org/web/20210330133536/https://www.gartner.com/doc/reprints?id=1-24JFAZOO&ct=201104&st=sb |archive-date=2021-03-30 |access-date=2021-07-11 |work=www.gartner.com |publisher=[[Company:Gartner|Gartner]] |id=G00723410}}{{dead|date=February 2025}}</ref> | ||
[[File:Edge computing infrastructure.png|thumb|The edge computing infrastructure]] | [[File:Edge computing infrastructure.png|thumb|The edge computing infrastructure]] | ||
==Definition== | ==Definition== | ||
Edge computing involves running computer programs that deliver quick responses [[Locality of reference|close to where requests are made]]. Karim Arabi, during an IEEE DAC 2014 keynote<ref>{{Cite web |url=http://www2.dac.com/events/videoarchive.aspx?confid=170&filter=keynote&id=170-103--0&#video |title=IEEE DAC 2014 Keynote: Mobile Computing Opportunities, Challenges and Technology Drivers |access-date=2019-03-25 |archive-date=2020-07-30 |archive-url=https://web.archive.org/web/20200730234708/http://www2.dac.com/events/videoarchive.aspx?confid=170&filter=keynote&id=170-103--0&#video |url-status=dead }}</ref> and later at an MIT MTL Seminar in 2015, described edge computing as computing that occurs outside [[Cloud computing|the cloud]], at the network's edge, particularly for applications needing immediate data processing.<ref>[https://www.mtl.mit.edu/seminars/trends-opportunities-and-challenges-driving-architecture-and-design-next-generation-mobile-computing-and-iot-devices MIT MTL Seminar: Trends, Opportunities and Challenges Driving Architecture and Design of Next Generation Mobile Computing and IoT Devices] </ref> | |||
Edge computing is often equated with [[Fog computing|fog computing]], particularly in smaller setups.<ref>{{Cite web|date=2017-03-02|title=What is fog and edge computing?|url=https://www.capgemini.com/2017/03/what-is-fog-and-edge-computing/|access-date=2021-07-06|website=Capgemini Worldwide|language=en-US|archive-date=2021-07-09|archive-url=https://web.archive.org/web/20210709185638/https://www.capgemini.com/2017/03/what-is-fog-and-edge-computing/|url-status=dead}}</ref> However, in larger deployments, such as smart cities, fog computing serves as a distinct layer between edge computing and cloud computing, with each layer having its own responsibilities.<ref>{{Cite book |last1=Dolui |first1=Koustabh |last2=Datta |first2=Soumya Kanti |title=2017 Global Internet of Things Summit (GIoTS) |chapter=Comparison of edge computing implementations: Fog computing, cloudlet and mobile edge computing |date=June 2017 |pages=1–6 |doi=10.1109/GIOTS.2017.8016213|isbn=978-1-5090-5873-0 |s2cid=11600169 }}</ref><ref>{{Cite web |date=2021-11-27 |title=Difference Between Edge Computing and Fog Computing |url=https://www.geeksforgeeks.org/difference-between-edge-computing-and-fog-computing/ |access-date=2022-09-11 |website=GeeksforGeeks |language=en-us}}</ref> | |||
The | "The State of the Edge" report explains that edge computing focuses on servers located close to the end-users.<ref>{{cite web|url=https://stateoftheedge.com/reports/data-at-the-edge-2019/|title=Data at the Edge Report|publisher=[[Company:Seagate Technology|Seagate Technology]]}}</ref> Alex Reznik, Chair of the ETSI MEC ISG standards committee, defines 'edge' loosely as anything that's not a traditional data center.<ref>{{Cite web |first=Alex |last=Reznik |url=https://www.etsi.org/newsroom/blogs/entry/what-is-edge |work=ETSI - ETSI Blog - etsi.org |title=What is Edge? |access-date=2019-02-19 |date=2018-05-14 |quote=What is 'Edge'? The best that I can do is this: it’s anything that's not a 'data center cloud'. }}</ref> | ||
In [[Software:Cloud gaming|cloud gaming]], edge nodes, known as "gamelets", are typically within one or two network hops from the client, ensuring quick response times for real-time games.<ref name="Anand 14–20">{{Cite book|last1=Anand|first1=B.|last2=Edwin|first2=A. J. Hao|title=2014 Seventh International Conference on Mobile Computing and Ubiquitous Networking (ICMU) |chapter=Gamelets — Multiplayer mobile games with distributed micro-clouds |s2cid=10374389|date=January 2014|pages=14–20|doi=10.1109/ICMU.2014.6799051|isbn=978-1-4799-2231-4 |url=http://scholarbank.nus.edu.sg/handle/10635/78158 }}</ref> | |||
Edge computing might use [[Virtualization|virtualization]] technology to simplify deploying and managing various applications on edge servers.<ref>{{Cite web |title=Edge virtualization manages the data deluge, but can be complex {{!}} TechTarget |url=https://www.techtarget.com/searchitoperations/feature/Edge-virtualization-manages-the-data-deluge-but-can-be-complex |access-date=2022-12-13 |website=IT Operations |language=en}}</ref> | |||
Edge computing | |||
== Concept == | == Concept == | ||
In 2018, the world's data was expected to grow 61 percent to 175 [[Byte#Multiple-byte units|zettabyte]]s by 2025.<ref>{{Cite web|last=Patrizio|first=Andy|date=2018-12-03|title=IDC: Expect 175 zettabytes of data worldwide by 2025|url=https://www.networkworld.com/article/3325397/idc-expect-175-zettabytes-of-data-worldwide-by-2025.html|access-date=2021-07-09|website=Network World|language=en}}</ref> According to research firm Gartner, around 10 percent of enterprise-generated data is created and processed outside a traditional centralized data center or cloud. By 2025, the firm predicts that this figure will reach 75 percent.<ref>{{Cite web|title=What We Do and How We Got Here|url=https://www.gartner.com/en/about|access-date=2021-12-21|website=Gartner|language=en}}</ref> The increase in [[Internet of things|IoT]] devices at the edge of the network is producing a massive amount of data — storing and using all that data in cloud data centers pushes network bandwidth requirements to the limit.<ref>{{Cite thesis |degree=Doctoral |last1=Ivkovic|first1=Jovan|date=2016-07-11|title=The Methods and Procedures for Accelerating Operations and Queries in Large Database Systems and Data Warehouse (Big Data Systems) |url=https://nardus.mpn.gov.rs/bitstream/handle/123456789/8683/Disertacija13336.pdf?sequence=5&isAllowed=y|website=National Repository of Dissertations in Serbia|language=sr, en-US}}</ref> Despite the improvements in [[Telecommunications network|network]] technology, data centers cannot guarantee acceptable transfer rates and response times, which often is a critical requirement for many applications.<ref name="shi-edge">{{cite journal |last1=Shi |first1=Weisong |last2=Cao |first2=Jie |last3=Zhang |first3=Quan |last4=Li |first4=Youhuizi |last5=Xu |first5=Lanyu |s2cid=4237186 |title=Edge Computing: Vision and Challenges |journal=IEEE Internet of Things Journal |date=October 2016 |volume=3 |issue=5 |pages=637–646 |doi=10.1109/JIOT.2016.2579198 |bibcode=2016IITJ....3..637S }}</ref> Furthermore, devices at the edge constantly consume data coming from the cloud, forcing companies to decentralize data storage and service provisioning, leveraging physical proximity to the end user. | |||
In a similar way, the aim of edge computing is to move the computation away from data centers towards the edge of the network, exploiting smart objects, [[Engineering:Smartphone|mobile phones]], or [[Engineering:Gateway (telecommunications)|network gateways]] to perform tasks and provide services on behalf of the cloud.<ref>{{cite journal |last1=Merenda |first1=Massimo |last2=Porcaro |first2=Carlo |last3=Iero |first3=Demetrio |title=Edge Machine Learning for AI-Enabled IoT Devices: A Review |journal= Sensors|date=29 April 2020 |volume=20 |issue=9 |page=2533 |doi=10.3390/s20092533 |pmid=32365645 |pmc=7273223 |bibcode=2020Senso..20.2533M |doi-access=free }}</ref> By moving [[Service (systems architecture)|services]] to the edge, it is possible to provide content [[Cache (computing)|caching]], service delivery, persistent data storage, and IoT management resulting in better response times and transfer rates. At the same time, distributing the logic to different network nodes introduces new issues and challenges.<ref>{{cite web | url=https://aristeksystems.com/blog/iot-2020-whats-next/| title=IoT management | accessdate=2020-04-08}}</ref> | In a similar way, the aim of edge computing is to move the computation away from data centers towards the edge of the network, exploiting smart objects, [[Engineering:Smartphone|mobile phones]], or [[Engineering:Gateway (telecommunications)|network gateways]] to perform tasks and provide services on behalf of the cloud.<ref>{{cite journal |last1=Merenda |first1=Massimo |last2=Porcaro |first2=Carlo |last3=Iero |first3=Demetrio |title=Edge Machine Learning for AI-Enabled IoT Devices: A Review |journal= Sensors|date=29 April 2020 |volume=20 |issue=9 |page=2533 |doi=10.3390/s20092533 |pmid=32365645 |pmc=7273223 |bibcode=2020Senso..20.2533M |doi-access=free }}</ref> By moving [[Service (systems architecture)|services]] to the edge, it is possible to provide content [[Cache (computing)|caching]], service delivery, persistent data storage, and IoT management resulting in better response times and transfer rates. At the same time, distributing the logic to different network nodes introduces new issues and challenges.<ref>{{cite web | url=https://aristeksystems.com/blog/iot-2020-whats-next/| title=IoT management | accessdate=2020-04-08}}</ref> | ||
=== Privacy and security === | === Privacy and security === | ||
The distributed nature of this paradigm introduces a shift in security schemes used in [[Cloud computing|cloud computing]]. In edge computing, data may travel between different distributed nodes connected | The distributed nature of this paradigm introduces a shift in security schemes used in [[Cloud computing|cloud computing]]. In edge computing, data may travel between different distributed nodes connected via the internet, and thus requires special encryption mechanisms independent of the cloud. This approach minimizes latency, reduces bandwidth consumption, and enhances real-time responsiveness for applications. Edge nodes may also be resource-constrained devices, limiting the choice in terms of security methods. Moreover, a shift from centralized top-down infrastructure to a decentralized trust model is required.<ref name="lopez-edge">{{cite journal |last1=Garcia Lopez |first1=Pedro |last2=Montresor |first2=Alberto |last3=Epema |first3=Dick |last4=Datta |first4=Anwitaman |last5=Higashino |first5=Teruo |last6=Iamnitchi |first6=Adriana |last7=Barcellos |first7=Marinho |last8=Felber |first8=Pascal |last9=Riviere |first9=Etienne |title=Edge-centric Computing |journal=ACM SIGCOMM Computer Communication Review |date=30 September 2015 |volume=45 |issue=5 |pages=37–42 |doi=10.1145/2831347.2831354 |doi-access=free |hdl=11572/114780 |hdl-access=free }}</ref> | ||
On the other hand, by keeping and processing data at the edge, it is possible to increase privacy by minimizing the transmission of sensitive information to the cloud. Furthermore, the ownership of collected data shifts from service providers to end-users.<ref name="brand">[https://medium.com/@aronbrand/edge-computing-alexa-and-the-future-of-enterprise-it-51c13268a365 3 Advantages of Edge Computing]. Aron Brand. Medium.com. Sep 20, 2019</ref> | On the other hand, by keeping and processing data at the edge, it is possible to increase privacy by minimizing the transmission of sensitive information to the cloud. Furthermore, the ownership of collected data shifts from service providers to end-users.<ref name="brand">[https://medium.com/@aronbrand/edge-computing-alexa-and-the-future-of-enterprise-it-51c13268a365 3 Advantages of Edge Computing]. Aron Brand. Medium.com. Sep 20, 2019</ref> | ||
| Line 37: | Line 36: | ||
=== Speed === | === Speed === | ||
Edge computing brings analytical computational resources close to the end users and therefore can increase the responsiveness and throughput of applications. A well-designed edge platform would significantly outperform a traditional cloud-based system. Some applications rely on short response times, making edge computing a significantly more feasible option than cloud computing. Examples range from IoT to autonomous driving,<ref>{{Cite journal|author1=Liu, S.|author2=Liu, L.|author3=Tang, B. Wu|author4=Wang, J.|author5=Shi, W.|title=Edge Computing for Autonomous Driving: Opportunities and Challenges | Edge computing brings analytical computational resources close to the end users and therefore can increase the responsiveness and throughput of applications. A well-designed edge platform would significantly outperform a traditional cloud-based system. Some applications rely on short response times, making edge computing a significantly more feasible option than cloud computing. Examples range from IoT to autonomous driving,<ref>{{Cite journal|author1=Liu, S.|author2=Liu, L.|author3=Tang, B. Wu|author4=Wang, J.|author5=Shi, W.|title=Edge Computing for Autonomous Driving: Opportunities and Challenges|journal=Proceedings of the IEEE|year=2019 |volume=107 |issue=8 |pages=1697–1716 |doi=10.1109/JPROC.2019.2915983 |s2cid=198311944 }}</ref> anything health or human / public safety relevant,<ref>{{Cite journal|last=Yu |display-authors=0 |first=W.|date=2018|title=A Survey on the Edge Computing for the Internet of Things|doi=10.1109/JIOT.2021.3072611 |arxiv=2104.01776 |s2cid=233025108 }}</ref> or involving human perception such as facial recognition, which typically takes a human between 370-620 ms to perform.<ref name=":2">{{Cite journal|last=Satyanarayanan|first=Mahadev|date=January 2017|title=The Emergence of Edge Computing|journal=Computer|volume=50|issue=1|pages=30–39|doi=10.1109/MC.2017.9|bibcode=2017Compr..50a..30S |s2cid=12563598 |issn=1558-0814}}</ref> Edge computing is more likely to be able to mimic the same perception [[Speed|speed]] as humans, which is useful in applications such as [[Augmented reality|augmented reality]], where the headset should preferably recognize who a person is at the same time as the wearer does. | ||
=== Efficiency === | === Efficiency === | ||
Due to the nearness of the analytical resources to the end users, sophisticated analytical tools and Artificial | Due to the nearness of the analytical resources to the end users, sophisticated analytical tools and [[Artificial intelligence|artificial intelligence]] tools can run on the edge of the system. This placement at the edge helps to increase operational efficiency and is responsible for many advantages to the system. In distributed AI systems on the edge, data compression is increasingly recognized as a foundational design layer to mitigate bandwidth constraints caused by the exchange of large models and high-resolution sensor streams.<ref>{{Cite web |date=2026-02-19 |title=Distributed AI and Edge Computing: Why Compression Is the Missing Layer |url=https://vanjasretenovic.com/blog/distributed-ai-edge-computing |access-date=2026-02-24 |website=vanjasretenovic.com |language=en}}</ref> | ||
Additionally, the usage of edge computing as an intermediate stage between client devices and the wider internet results in efficiency savings that can be demonstrated in the following example: A client device requires computationally intensive processing on video files to be performed on external servers. By using servers located on a local edge network to perform those computations, the video files only need to be transmitted in the local network. Avoiding transmission over the internet results in significant bandwidth savings and therefore increases efficiency.<ref name=":2" /> | Additionally, the usage of edge computing as an intermediate stage between client devices and the wider internet results in efficiency savings that can be demonstrated in the following example: A client device requires computationally intensive processing on video files to be performed on external servers. By using servers located on a local edge network to perform those computations, the video files only need to be transmitted in the local network. Avoiding transmission over the internet results in significant bandwidth savings and therefore increases efficiency.<ref name=":2" /> Another example is [[Speech recognition|voice recognition]]. If the recognition is performed locally, it is possible to send the recognized text to the cloud rather than audio recordings, significantly reducing the amount of required bandwidth.<ref name="brand"/> | ||
== Applications == | == Applications == | ||
Edge application services reduce the volumes of data that must be moved, the consequent traffic, and the distance that data must travel. That provides lower latency and reduces transmission costs. [[Computation offloading]] for real-time applications, such as facial recognition algorithms, showed considerable improvements in response times, as demonstrated in early research.<ref name="fog-yi">{{cite book |last1=Yi |first1=S. |last2=Hao |first2=Z. |last3=Qin |first3=Z. |last4=Li |first4=Q. |title=2015 Third IEEE Workshop on Hot Topics in Web Systems and Technologies (HotWeb) |chapter=Fog Computing: Platform and Applications |s2cid=6753944 |date=November | Edge application services reduce the volumes of data that must be moved, the consequent traffic, and the distance that data must travel. That provides lower latency and reduces transmission costs. [[Computation offloading]] for real-time applications, such as facial recognition algorithms, showed considerable improvements in response times, as demonstrated in early research.<ref name="fog-yi">{{cite book |last1=Yi |first1=S. |last2=Hao |first2=Z. |last3=Qin |first3=Z. |last4=Li |first4=Q. |title=2015 Third IEEE Workshop on Hot Topics in Web Systems and Technologies (HotWeb) |chapter=Fog Computing: Platform and Applications |s2cid=6753944 |date=November 2019 |pages=73–78 |doi=10.1109/HotWeb.2015.22 |isbn=978-1-4673-9688-2 }}</ref> Further research showed that using resource-rich machines called [[Cloudlet|cloudlet]]s or micro data centers near mobile users, which offer services typically found in the cloud, provided improvements in execution time when some of the tasks are offloaded to the edge node.<ref name="verbelen-cloudlets">{{cite book |last1=Verbelen |first1=Tim |last2=Simoens |first2=Pieter |last3=De Turck |first3=Filip |last4=Dhoedt |first4=Bart |title=Proceedings of the third ACM workshop on Mobile cloud computing and services |chapter=Cloudlets |s2cid=3249347 |date=2012 |pages=29–36 |doi=10.1145/2307849.2307858 |chapter-url=https://dl.acm.org/citation.cfm?id=2307858 |access-date=4 July 2019 |publisher=ACM|hdl=1854/LU-2984272 |isbn=9781450313193 |hdl-access=free }}</ref> On the other hand, offloading every task may result in a slowdown due to transfer times between device and nodes, so depending on the workload, an optimal configuration can be defined. | ||
An IoT-based power grid system enables communication of electricity and data to monitor and control the power grid,<ref>{{Cite journal |last1=Minh |first1=Quy Nguyen |last2=Nguyen |first2=Van-Hau |last3=Quy |first3=Vu Khanh |last4=Ngoc |first4=Le Anh |last5=Chehri |first5=Abdellah |last6=Jeon |first6=Gwanggil |date=2022 |title=Edge Computing for IoT-Enabled Smart Grid: The Future of Energy |journal=Energies |language=en |volume=15 |issue=17 |pages=6140 |doi=10.3390/en15176140 |issn=1996-1073|doi-access=free }}</ref> which makes [[Physics:Energy management|energy management]] more efficient. | |||
Other notable applications include [[Engineering:Connected car|connected car]]s, | Other notable applications include [[Engineering:Connected car|connected car]]s, [[Engineering:Self-driving car|self-driving car]]s,<ref>[https://www.wired.com/story/its-time-to-think-beyond-cloud-computing/ It's Time to Think Beyond Cloud Computing] Published by wired.com retrieved April 10, 2019</ref> [[Social:Smart city|smart cities]],<ref name="taleb-mobile">{{cite journal |last1=Taleb |first1=Tarik |last2=Dutta |first2=Sunny |last3=Ksentini |first3=Adlen |last4=Iqbal |first4=Muddesar |last5=Flinck |first5=Hannu |s2cid=11163718 |title=Mobile Edge Computing Potential in Making Cities Smarter |journal=IEEE Communications Magazine |date=March 2017 |volume=55 |issue=3 |pages=38–43 |doi=10.1109/MCOM.2017.1600249CM |bibcode=2017IComM..55c..38T |url=http://researchopen.lsbu.ac.uk/378/ |access-date=5 July 2014|doi-access=free }}</ref> [[Industry 4.0]], [[Engineering:Home automation|home automation]],<ref name="chakraborty-home">{{cite book |last1=Chakraborty |first1=T. |last2=Datta |first2=S. K. |title=2017 IEEE International Symposium on Consumer Electronics (ISCE) |chapter=Home automation using edge computing and Internet of Things |date=November 2017 |pages=47–49 |doi=10.1109/ISCE.2017.8355544 |isbn=978-1-5386-2189-9 |s2cid=19156163 }}</ref> [[Engineering:Missile|missiles]],<ref>{{Cite web |last=Velayanikal |first=Malavika |date=2021-02-15 |title=Guided missiles homing in with Indian deep tech |url=https://www.livemint.com/news/business-of-life/guided-missiles-homing-in-with-indian-deep-tech-11613314273154.html |access-date=2021-02-19 |website=Mint |language=en}}</ref>, [[Astronomy:Satellite|satellite]] systems <ref>[https://www.satellitetoday.com/in-space-services/2023/05/12/size-of-the-prize-how-will-edge-computing-in-space-drive-value-creation/ Size of the Prize: How Will Edge Computing in Space Drive Value Creation?] Published by Via Satellite retrieved August 18, 2023</ref>, as well as frameworks such as [[Software:MediaPipe|MediaPipe]]. The growing field of '''edge artificial intelligence''' (edge AI or edge intelligence, sometimes referred to as "local AI" or "on-device AI") implements artificial intelligence in an edge computing environment, on the device or close to where data is collected.<ref name=AIOnTheEdge>{{cite journal | last1 = Su | first1 = Weixing | last2 = Li | first2 = Linfeng | last3 = Liu | first3 = Fang | last4 = He | first4 = Maowei | last5 = Liang | first5 = Xiaodan | title = AI on the edge: a comprehensive review | journal = Artificial Intelligence Review | ||
| year = 2022 | volume = 55 | issue = 8 | pages = 6125–6183 | doi = 10.1007/s10462-022-10141-4 | url = https://doi.org/10.1007/s10462-022-10141-4 | access-date = 4 November 2025| url-access = subscription }}</ref><ref name=EdgeIntelligence>{{cite journal | last1 = Zhou | first1 = Zhi | last2 = Chen | first2 = Xu | last3 = Li | first3 = En | last4 = Zeng | first4 = Liekang | last5 = Luo | first5 = Ke | last6 = Zhang | first6 = Junshan | title = Edge Intelligence: Paving the Last Mile of Artificial Intelligence with Edge Computing | journal = Proceedings of the IEEE | year = 2019 | volume = 107 | issue = 8 | pages = 1738–1762 | doi = 10.1109/JPROC.2019.2921977 | arxiv = 1905.10083 | url = https://arxiv.org/abs/1905.10083 | access-date = 4 November 2025}}</ref> | |||
==See also== | ==See also== | ||
Latest revision as of 06:33, 16 April 2026
Short description: Distributed computing paradigm
Edge computing is a distributed computing model that brings computation and data storage closer to the sources of data. More broadly, it refers to any design that pushes computation physically closer to a user, so as to reduce the latency compared to when an application runs on a centralized data center.[1]
The term began being used in the 1990s to describe content delivery networks—these were used to deliver website and video content from servers located near users.[2] In the early 2000s, these systems expanded their scope to hosting other applications,[3] leading to early edge computing services.[4] These services could do things like find dealers, manage shopping carts, gather real-time data, and place ads.
The Internet of things (IoT), where devices are connected to the Internet, is often linked with edge computing.[5]
Definition
Edge computing involves running computer programs that deliver quick responses close to where requests are made. Karim Arabi, during an IEEE DAC 2014 keynote[6] and later at an MIT MTL Seminar in 2015, described edge computing as computing that occurs outside the cloud, at the network's edge, particularly for applications needing immediate data processing.[7] Edge computing is often equated with fog computing, particularly in smaller setups.[8] However, in larger deployments, such as smart cities, fog computing serves as a distinct layer between edge computing and cloud computing, with each layer having its own responsibilities.[9][10]
"The State of the Edge" report explains that edge computing focuses on servers located close to the end-users.[11] Alex Reznik, Chair of the ETSI MEC ISG standards committee, defines 'edge' loosely as anything that's not a traditional data center.[12]
In cloud gaming, edge nodes, known as "gamelets", are typically within one or two network hops from the client, ensuring quick response times for real-time games.[13]
Edge computing might use virtualization technology to simplify deploying and managing various applications on edge servers.[14]
Concept
In 2018, the world's data was expected to grow 61 percent to 175 zettabytes by 2025.[15] According to research firm Gartner, around 10 percent of enterprise-generated data is created and processed outside a traditional centralized data center or cloud. By 2025, the firm predicts that this figure will reach 75 percent.[16] The increase in IoT devices at the edge of the network is producing a massive amount of data — storing and using all that data in cloud data centers pushes network bandwidth requirements to the limit.[17] Despite the improvements in network technology, data centers cannot guarantee acceptable transfer rates and response times, which often is a critical requirement for many applications.[18] Furthermore, devices at the edge constantly consume data coming from the cloud, forcing companies to decentralize data storage and service provisioning, leveraging physical proximity to the end user.
In a similar way, the aim of edge computing is to move the computation away from data centers towards the edge of the network, exploiting smart objects, mobile phones, or network gateways to perform tasks and provide services on behalf of the cloud.[19] By moving services to the edge, it is possible to provide content caching, service delivery, persistent data storage, and IoT management resulting in better response times and transfer rates. At the same time, distributing the logic to different network nodes introduces new issues and challenges.[20]
Privacy and security
The distributed nature of this paradigm introduces a shift in security schemes used in cloud computing. In edge computing, data may travel between different distributed nodes connected via the internet, and thus requires special encryption mechanisms independent of the cloud. This approach minimizes latency, reduces bandwidth consumption, and enhances real-time responsiveness for applications. Edge nodes may also be resource-constrained devices, limiting the choice in terms of security methods. Moreover, a shift from centralized top-down infrastructure to a decentralized trust model is required.[21] On the other hand, by keeping and processing data at the edge, it is possible to increase privacy by minimizing the transmission of sensitive information to the cloud. Furthermore, the ownership of collected data shifts from service providers to end-users.[22]
Scalability
Scalability in a distributed network must face different issues. First, it must take into account the heterogeneity of the devices, having different performance and energy constraints, the highly dynamic condition, and the reliability of the connections compared to more robust infrastructure of cloud data centers. Moreover, security requirements may introduce further latency in the communication between nodes, which may slow down the scaling process.[18]
The state-of-the-art scheduling technique can increase the effective utilization of edge resources and scales the edge server by assigning minimum edge resources to each offloaded task.[23]
Reliability
Management of failovers is crucial in order to keep a service alive. If a single node goes down and is unreachable, users should still be able to access a service without interruptions. Moreover, edge computing systems must provide actions to recover from a failure and alert the user about the incident. To this aim, each device must maintain the network topology of the entire distributed system, so that detection of errors and recovery become easily applicable. Other factors that may influence this aspect are the connection technologies in use, which may provide different levels of reliability, and the accuracy of the data produced at the edge that could be unreliable due to particular environment conditions.[18] As an example, an edge computing device, such as a voice assistant, may continue to provide service to local users even during cloud service or internet outages.[22]
Speed
Edge computing brings analytical computational resources close to the end users and therefore can increase the responsiveness and throughput of applications. A well-designed edge platform would significantly outperform a traditional cloud-based system. Some applications rely on short response times, making edge computing a significantly more feasible option than cloud computing. Examples range from IoT to autonomous driving,[24] anything health or human / public safety relevant,[25] or involving human perception such as facial recognition, which typically takes a human between 370-620 ms to perform.[26] Edge computing is more likely to be able to mimic the same perception speed as humans, which is useful in applications such as augmented reality, where the headset should preferably recognize who a person is at the same time as the wearer does.
Efficiency
Due to the nearness of the analytical resources to the end users, sophisticated analytical tools and artificial intelligence tools can run on the edge of the system. This placement at the edge helps to increase operational efficiency and is responsible for many advantages to the system. In distributed AI systems on the edge, data compression is increasingly recognized as a foundational design layer to mitigate bandwidth constraints caused by the exchange of large models and high-resolution sensor streams.[27]
Additionally, the usage of edge computing as an intermediate stage between client devices and the wider internet results in efficiency savings that can be demonstrated in the following example: A client device requires computationally intensive processing on video files to be performed on external servers. By using servers located on a local edge network to perform those computations, the video files only need to be transmitted in the local network. Avoiding transmission over the internet results in significant bandwidth savings and therefore increases efficiency.[26] Another example is voice recognition. If the recognition is performed locally, it is possible to send the recognized text to the cloud rather than audio recordings, significantly reducing the amount of required bandwidth.[22]
Applications
Edge application services reduce the volumes of data that must be moved, the consequent traffic, and the distance that data must travel. That provides lower latency and reduces transmission costs. Computation offloading for real-time applications, such as facial recognition algorithms, showed considerable improvements in response times, as demonstrated in early research.[28] Further research showed that using resource-rich machines called cloudlets or micro data centers near mobile users, which offer services typically found in the cloud, provided improvements in execution time when some of the tasks are offloaded to the edge node.[29] On the other hand, offloading every task may result in a slowdown due to transfer times between device and nodes, so depending on the workload, an optimal configuration can be defined.
An IoT-based power grid system enables communication of electricity and data to monitor and control the power grid,[30] which makes energy management more efficient.
Other notable applications include connected cars, self-driving cars,[31] smart cities,[32] Industry 4.0, home automation,[33] missiles,[34], satellite systems [35], as well as frameworks such as MediaPipe. The growing field of edge artificial intelligence (edge AI or edge intelligence, sometimes referred to as "local AI" or "on-device AI") implements artificial intelligence in an edge computing environment, on the device or close to where data is collected.[36][37]
See also
- Content delivery network
- Dew computing
- Edge data integration
- Edge device
- Fat client
- Heterogeneous computing
- Mobile edge computing
- Personal computer
- Serverless architecture
- Smart camera
- Ubiquitous computing
References
- ↑ Gill, Bob; Smith, David (2018-09-14). "The Edge Completes the Cloud: A Gartner Trend Insight Report". Gartner. Gartner. https://emtemp.gcom.cloud/ngw/globalassets/en/doc/documents/3889058-the-edge-completes-the-cloud-a-gartner-trend-insight-report.pdf.
- ↑ Dilley, John; Maggs, Bruce; Parikh, Jay; Prokop, Harald; Sitaraman, Ramesh; Weihl, Bill (2002-10-31). "Globally Distributed Content Delivery". IEEE Internet Computing 6 (5): 50-58. doi:10.1109/MIC.2002.1036038. ISSN 1089-7801. https://people.cs.umass.edu/~ramesh/Site/PUBLICATIONS_files/DMPPSW02.pdf. Retrieved 2026-02-02.
- ↑ Nygren, Erik; Sitaraman, Ramesh K.; Sun, Jennifer (2010-08-17). "The Akamai Network: A Platform for High-Performance Internet Applications". ACM SIGOPS Operating Systems Review 44 (3): 2–19. doi:10.1145/1842733.1842736. ISSN 0163-5980. https://www.akamai.com/site/en/documents/research-paper/the-akamai-network-a-platform-for-high-performance-internet-applications-technical-publication.pdf. Retrieved 2026-02-02. "See Section 6.2: Distributing Applications to the Edge".
- ↑ Davis, Andy; Parikh, Jay; Weihl, William E. (2004). "EdgeComputing: Extending Enterprise Applications to the Edge of the Internet". WWW Alt. '04. p. 180-187. doi:10.1145/1013367.1013397. ISBN 1-58113-912-8. https://www.akamai.com/site/en/documents/research-paper/edgecomputing-extending-enterprise-applications-to-the-edge-of-the-internet-technical-publication.pdf. Retrieved 2026-02-02.
- ↑ Gill, Bob (2021-11-03). "2021 Strategic Roadmap for Edge Computing". www.gartner.com. Gartner. https://www.gartner.com/doc/reprints?id=1-24JFAZOO&ct=201104&st=sb.[|permanent dead link|dead link}}]
- ↑ "IEEE DAC 2014 Keynote: Mobile Computing Opportunities, Challenges and Technology Drivers". http://www2.dac.com/events/videoarchive.aspx?confid=170&filter=keynote&id=170-103--0&#video.
- ↑ MIT MTL Seminar: Trends, Opportunities and Challenges Driving Architecture and Design of Next Generation Mobile Computing and IoT Devices
- ↑ "What is fog and edge computing?" (in en-US). 2017-03-02. https://www.capgemini.com/2017/03/what-is-fog-and-edge-computing/.
- ↑ Dolui, Koustabh; Datta, Soumya Kanti (June 2017). "Comparison of edge computing implementations: Fog computing, cloudlet and mobile edge computing". 2017 Global Internet of Things Summit (GIoTS). pp. 1–6. doi:10.1109/GIOTS.2017.8016213. ISBN 978-1-5090-5873-0.
- ↑ "Difference Between Edge Computing and Fog Computing" (in en-us). 2021-11-27. https://www.geeksforgeeks.org/difference-between-edge-computing-and-fog-computing/.
- ↑ "Data at the Edge Report". Seagate Technology. https://stateoftheedge.com/reports/data-at-the-edge-2019/.
- ↑ Reznik, Alex (2018-05-14). "What is Edge?". ETSI - ETSI Blog - etsi.org. https://www.etsi.org/newsroom/blogs/entry/what-is-edge. "What is 'Edge'? The best that I can do is this: it’s anything that's not a 'data center cloud'."
- ↑ Anand, B.; Edwin, A. J. Hao (January 2014). "Gamelets — Multiplayer mobile games with distributed micro-clouds". 2014 Seventh International Conference on Mobile Computing and Ubiquitous Networking (ICMU). pp. 14–20. doi:10.1109/ICMU.2014.6799051. ISBN 978-1-4799-2231-4. http://scholarbank.nus.edu.sg/handle/10635/78158.
- ↑ "Edge virtualization manages the data deluge, but can be complex | TechTarget" (in en). https://www.techtarget.com/searchitoperations/feature/Edge-virtualization-manages-the-data-deluge-but-can-be-complex.
- ↑ Patrizio, Andy (2018-12-03). "IDC: Expect 175 zettabytes of data worldwide by 2025" (in en). https://www.networkworld.com/article/3325397/idc-expect-175-zettabytes-of-data-worldwide-by-2025.html.
- ↑ "What We Do and How We Got Here" (in en). https://www.gartner.com/en/about.
- ↑ Ivkovic, Jovan (2016-07-11). The Methods and Procedures for Accelerating Operations and Queries in Large Database Systems and Data Warehouse (Big Data Systems) (PDF). National Repository of Dissertations in Serbia (Doctoral thesis) (in српски / srpski and English).
- ↑ 18.0 18.1 18.2 Shi, Weisong; Cao, Jie; Zhang, Quan; Li, Youhuizi; Xu, Lanyu (October 2016). "Edge Computing: Vision and Challenges". IEEE Internet of Things Journal 3 (5): 637–646. doi:10.1109/JIOT.2016.2579198. Bibcode: 2016IITJ....3..637S.
- ↑ Merenda, Massimo; Porcaro, Carlo; Iero, Demetrio (29 April 2020). "Edge Machine Learning for AI-Enabled IoT Devices: A Review". Sensors 20 (9): 2533. doi:10.3390/s20092533. PMID 32365645. Bibcode: 2020Senso..20.2533M.
- ↑ "IoT management". https://aristeksystems.com/blog/iot-2020-whats-next/. Retrieved 2020-04-08.
- ↑ Garcia Lopez, Pedro; Montresor, Alberto; Epema, Dick; Datta, Anwitaman; Higashino, Teruo; Iamnitchi, Adriana; Barcellos, Marinho; Felber, Pascal et al. (30 September 2015). "Edge-centric Computing". ACM SIGCOMM Computer Communication Review 45 (5): 37–42. doi:10.1145/2831347.2831354.
- ↑ 22.0 22.1 22.2 3 Advantages of Edge Computing. Aron Brand. Medium.com. Sep 20, 2019
- ↑ Babar, Mohammad; Sohail Khan, Muhammad (July 2021). "ScalEdge: A framework for scalable edge computing in Internet of things–based smart systems" (in en). International Journal of Distributed Sensor Networks 17 (7): 155014772110353. doi:10.1177/15501477211035332. ISSN 1550-1477.
- ↑ Liu, S.; Liu, L.; Tang, B. Wu; Wang, J.; Shi, W. (2019). "Edge Computing for Autonomous Driving: Opportunities and Challenges". Proceedings of the IEEE 107 (8): 1697–1716. doi:10.1109/JPROC.2019.2915983.
- ↑ Yu, W. (2018). A Survey on the Edge Computing for the Internet of Things. doi:10.1109/JIOT.2021.3072611.
- ↑ 26.0 26.1 Satyanarayanan, Mahadev (January 2017). "The Emergence of Edge Computing". Computer 50 (1): 30–39. doi:10.1109/MC.2017.9. ISSN 1558-0814. Bibcode: 2017Compr..50a..30S.
- ↑ "Distributed AI and Edge Computing: Why Compression Is the Missing Layer" (in en). 2026-02-19. https://vanjasretenovic.com/blog/distributed-ai-edge-computing.
- ↑ Yi, S.; Hao, Z.; Qin, Z.; Li, Q. (November 2019). "Fog Computing: Platform and Applications". 2015 Third IEEE Workshop on Hot Topics in Web Systems and Technologies (HotWeb). pp. 73–78. doi:10.1109/HotWeb.2015.22. ISBN 978-1-4673-9688-2.
- ↑ Verbelen, Tim; Simoens, Pieter; De Turck, Filip; Dhoedt, Bart (2012). "Cloudlets". Proceedings of the third ACM workshop on Mobile cloud computing and services. ACM. pp. 29–36. doi:10.1145/2307849.2307858. ISBN 9781450313193. https://dl.acm.org/citation.cfm?id=2307858. Retrieved 4 July 2019.
- ↑ Minh, Quy Nguyen; Nguyen, Van-Hau; Quy, Vu Khanh; Ngoc, Le Anh; Chehri, Abdellah; Jeon, Gwanggil (2022). "Edge Computing for IoT-Enabled Smart Grid: The Future of Energy" (in en). Energies 15 (17): 6140. doi:10.3390/en15176140. ISSN 1996-1073.
- ↑ It's Time to Think Beyond Cloud Computing Published by wired.com retrieved April 10, 2019
- ↑ Taleb, Tarik; Dutta, Sunny; Ksentini, Adlen; Iqbal, Muddesar; Flinck, Hannu (March 2017). "Mobile Edge Computing Potential in Making Cities Smarter". IEEE Communications Magazine 55 (3): 38–43. doi:10.1109/MCOM.2017.1600249CM. Bibcode: 2017IComM..55c..38T. http://researchopen.lsbu.ac.uk/378/. Retrieved 5 July 2014.
- ↑ Chakraborty, T.; Datta, S. K. (November 2017). "Home automation using edge computing and Internet of Things". 2017 IEEE International Symposium on Consumer Electronics (ISCE). pp. 47–49. doi:10.1109/ISCE.2017.8355544. ISBN 978-1-5386-2189-9.
- ↑ Velayanikal, Malavika (2021-02-15). "Guided missiles homing in with Indian deep tech" (in en). https://www.livemint.com/news/business-of-life/guided-missiles-homing-in-with-indian-deep-tech-11613314273154.html.
- ↑ Size of the Prize: How Will Edge Computing in Space Drive Value Creation? Published by Via Satellite retrieved August 18, 2023
- ↑ Su, Weixing; Li, Linfeng; Liu, Fang; He, Maowei; Liang, Xiaodan (2022). "AI on the edge: a comprehensive review". Artificial Intelligence Review 55 (8): 6125–6183. doi:10.1007/s10462-022-10141-4. https://doi.org/10.1007/s10462-022-10141-4. Retrieved 4 November 2025.
- ↑ Zhou, Zhi; Chen, Xu; Li, En; Zeng, Liekang; Luo, Ke; Zhang, Junshan (2019). "Edge Intelligence: Paving the Last Mile of Artificial Intelligence with Edge Computing". Proceedings of the IEEE 107 (8): 1738–1762. doi:10.1109/JPROC.2019.2921977. https://arxiv.org/abs/1905.10083. Retrieved 4 November 2025.
 |  |
