Edge AI: A survey

May combine two claims.

Cloud computing has key problems with security, service speed, slow connections, and low bandwidth/high latency and jitter when mobile devices offload processing to cloud services.

“However, this technology faces key issues: security, speed of services and slow connections, which are often combined as low bandwidth/high latency and jitter as mobile devices offload computational and processing capacity to cloud computing services [3,4].”

• cites [3] Chen, 2016, Efficient multi-user computation offloading for mobile-edge cloud computing, IEEE/ACM Trans. Netw. sourcePartly supported

  The Introduction (p. 1) says remote-cloud offloading can cause long network latency that harms interactive response, and explicitly mentions delay and jitter. It also says wireless interference can reduce offloading data rates. The cited paper does not identify security as an issue, so it supports only part of the attributed list. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [4] Singh, 2022, To offload or not? an analysis of big data offloading strategies from edge to cloud sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

MEC aims to provide mobile and cloud computing services close to mobile users to decrease latency, achieve high throughput, and provide direct access to real-time network information.

“The concept of MEC is simply to provide mobile and cloud computing services within close proximity of the mobile user, i.e. the provision of computing power in a delocalized manner close to mobile users (smartphones, tablets, etc.), aiming to decrease latency, achieve as high throughput as possible and provide direct access to real-time network information [16].”

• cites [16] Iftikhar, 2023, Hunterplus: ai based energy-efficient task scheduling for cloud–fog computing environments, Internet Things sourcePartly supported

  The System Architecture section describes smartphones and tablets as gateway devices and says fog resources offer lower latency than geographically distant cloud resources. This supports the proximity and latency aspects of the claim, but our copy does not define MEC or state that it aims for high throughput or direct access to real-time network information (searched for MEC and real-time network information). (LLM, conf=0.83)

  Experimental — accuracy not yet measured

May combine two claims.

Most survey articles on Edge AI were published before 2021 except [23], but [23] is only an overview or visionary work offering future directions in many relevant areas, not an in-depth study.

“Most of the survey articles on Edge AI have been published before 2021 except [23], but this [23] work is just an overview on Edge AI or a visionary work which offers only future directions in many relevant areas but not in-depth.”

• cites [23] Ding, 2022 sourcePartly supported

  The Introduction and Section 5 describe the paper as an envisioned Edge AI roadmap spanning 5G, future cloud, and AI/ML, supporting its characterization as a broad, forward-looking work. But the claim that it offers only future directions and no in-depth discussion goes beyond the source: Section 3 says it will examine technological challenges in detail. The paper also does not establish when most other Edge AI surveys were published. (LLM, conf=0.85)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Emerging technologies such as SDN and NFV enable network reliability and robustness and improve integration with existing IoT environments.

“With emerging technologies, such as software-defined networking (SDN) and Network Function Virtualization (NFV), this enables reliability and robustness of the network and improves integration with existing IoT environments [59].”

• cites [59] Nabavi, 2022, Tractor: traffic-aware and power-efficient virtual machine placement in edge-cloud data centers using artificial bee colony optimization, Int. J. Commun. Syst. sourceCould not verify

  The Future Research Directions section discusses IoT and attributes system reliability to a blockchain-based framework, while section 2.3 discusses IoT–edge–cloud integration through FogBus. Searches for SDN, NFV, and the claimed reliability and integration benefits found no passage connecting those benefits to SDN or NFV. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Cloudlets are dedicated either to small-enterprise scenarios or can be expanded to publicly funded schemes with free access and open source software.

“while cloudlets are dedicated either (as originally envisioned) to small-enterprise scenarios or can be expanded to publicly funded schemes with free access and open source software [48].”

• cites [48] Golec, 2022, Aiblock: blockchain based lightweight framework for serverless computing using ai sourceCould not verify

  The Introduction discusses cloud and serverless computing, while the Conclusions describe AIBLOCK; neither addresses cloudlets or their intended users. Searches for cloudlets, small-enterprise use, publicly funded access, and open-source schemes found no supporting passage in our copy. (LLM, conf=0.78)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

In 2017, Mobile Edge Computing was retitled Multi Access Edge Computing to recognize the large IoT component in proof-of-concept studies and to harmonize Edge Computing systems and applications in highly heterogeneous wireless networks.

“In 2017, Mobile Edge Computing was retitled Multi Access Edge Computing to recognize the large IoT component in proof-of-concept studies and to harmonize Edge Computing systems and applications in highly heterogeneous wireless networks [49].”

• cites [49] Malik, 2021, Effort: energy efficient framework for offload communication in mobile cloud computing, Software Pract. Ex. sourceCould not verify

  The Introduction discusses mobile cloud computing and wireless offloading, while the Abstract mentions IoT devices. Neither section addresses a 2017 renaming of Mobile Edge Computing, proof-of-concept studies, or the stated reasons for a name change. Searches for the new name and those reasons found no supporting passage in our copy. (LLM, conf=0.78)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Fog Computing is the optimal solution for IoT applications to minimize the time required for time-critical processing and Big Data analytics.

“Fog Computing as the optimal solution for IoT applications to minimize the time required for time-critical processing and Big Data analytics;”

• cites [47] Nabavi, 2023, Seagull optimization algorithm based multi-objective vm placement in edge-cloud data centers, Internet Things and Cyber-Phys. Syst. sourceNot checked — no full text of the source
• cites [48] Golec, 2022, Aiblock: blockchain based lightweight framework for serverless computing using ai sourceCould not verify

  The Introduction (page unavailable) discusses processing large amounts of IoT data using cloud and serverless computing, not fog computing. Searches for fog computing, time-critical processing, and Big Data analytics found no passage calling fog computing optimal or saying it minimizes processing time for those tasks. (LLM, conf=0.77)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Nearly every industry today has tasks that might be better served by using edge AI.

“Nearly every single industry today has tasks that might be better served by using edge AI [22].”

• cites [22] Kalapothas, 2022, Efficient edge-ai application deployment for fpgas, Information sourcePartly supported

  The Introduction says AI and machine learning are being deployed in various sectors, including healthcare, education and safety, and identifies tasks requiring processing at the edge. The Discussion also describes demanding edge-AI tasks. These passages support applications across multiple fields, but not the claim that nearly every industry has tasks better served by edge AI. (LLM, conf=0.83)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Micro Data Centres are optimal for industrial applications in specific locations or at temporary sites.

“Micro Data Centres as the optimum for industrial applications in specific locations or at temporary sites”

• cites [47] Nabavi, 2023, Seagull optimization algorithm based multi-objective vm placement in edge-cloud data centers, Internet Things and Cyber-Phys. Syst. sourceNot checked — no full text of the source
• cites [48] Golec, 2022, Aiblock: blockchain based lightweight framework for serverless computing using ai sourceCould not verify

  The Introduction discusses cloud and serverless computing for IoT, while the Hardware Instruments & Software Components section describes a serverless datacenter used for COVID-19 detection. Searches for micro data centres and temporary industrial sites found no passage assessing whether micro data centres are optimal for those applications. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Mobile Edge Computing is centred on applications intended for large numbers of mobile devices.

“Mobile Edge Computing as centred on applications intended for large numbers of mobile devices,”

• cites [47] Nabavi, 2023, Seagull optimization algorithm based multi-objective vm placement in edge-cloud data centers, Internet Things and Cyber-Phys. Syst. sourceNot checked — no full text of the source
• cites [48] Golec, 2022, Aiblock: blockchain based lightweight framework for serverless computing using ai sourceCould not verify

  The Introduction (page unavailable) discusses the growing number of IoT devices and serverless cloud computing, but does not characterize Mobile Edge Computing. Searches for Mobile Edge Computing and applications for large numbers of mobile devices found no passage addressing that claim. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

May combine two claims.

Due to hardware limitations of wireless devices, offloading must be performed to increase functionality, but overloaded offloads can degrade performance.

“However, due to hardware limitations of wireless devices' computation, offloading must be performed to increase functionality, but overloaded offloads can degrade performance [85].”

• cites [85] Singh, 2019, The advantage of computation offloading in multi-access edge computing sourceCould not verify

  The available indexed passages list the paper’s title and keywords, including computation offloading, mobile devices, and task completion time (Abstract; page unavailable). Searches for device hardware limitations and performance degradation from overloaded offloading found no passage addressing either assertion. (LLM, conf=0.72)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

Edge applications are leading the future generation of AI computing, which will positively affect everyday life at home, work, classroom, and travel.

“In reality, edge applications are leading the charge for the future generation of AI computing, which will have positive effects on our everyday lives in many contexts, including at home, in the workplace, in the classroom, and when travelling [23].”

• cites [23] Ding, 2022 sourcePartly supported

  The Introduction describes Edge AI as a fast-evolving field already entering mainstream services, including connected vehicles and smart factories. Section 2.4 anticipates benefits for everyday life but says they still need evaluation. The paper does not establish that edge applications are leading the next generation of AI computing or support the full list of settings, particularly homes and classrooms. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

The key objective of Edge Computing is to put resources close to users and data sources to help overcome cloud computing's high latency, jitter, and narrow bandwidth.

“The key objective of Edge Computing is to put resources within close proximity to the users and sources of data and information to help overcome cloud computing's recognized weaknesses of high latency, jitter and narrow bandwidth [11,52].”

• cites [11] Satyanarayanan, 2009, The case for vm-based cloudlets in mobile computing, IEEE pervasive Comput. sourceNot checked — no full text of the source
• cites [52] T.H. Luan, L. Gao, Z. Li, Y. Xiang, G. Wei, L. Sun, Fog computing: focusing on mobile users at the edge, arXiv preprint arXiv:1502.01815 (2015), 1-11. sourcePartly supported

  The Introduction (p. 2) describes placing fog resources near mobile users to replace a long, thin cloud connection, and “Why Fog?” (p. 4) reports reduced latency and higher data rates. The paper treats fog as an edge-computing platform (Computing, p. 7), but these passages focus on mobile users rather than data sources and do not identify jitter as a cloud weakness. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

IDC predicts that by 2025 there will be 150 billion intelligent edge devices in use.

“International Data Corporation (IDC) predicts that by 2025, there will be 150 billion intelligent edge devices in use [72].”

• cites [72] McEnroe, 2022 sourceCould not verify

  The Edge AI section discusses industry interest and applications, but gives no IDC forecast. Searches for “150 billion,” “2025,” and “International Data Corporation/IDC” found no passage in our copy stating the claimed prediction; no page numbers were available for the retrieved passages. (LLM, conf=0.74)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

MEC and fog computing have subsequently been completely mixed and matched by various authors.

“This was unfortunate because confusion with Fog Computing was made possible and MEC and Fog Computing have subsequently been completely “mixed and matched” by various authors [52,112,135,136].”

• cites [52] T.H. Luan, L. Gao, Z. Li, Y. Xiang, G. Wei, L. Sun, Fog computing: focusing on mobile users at the edge, arXiv preprint arXiv:1502.01815 (2015), 1-11. sourcePartly supported

  The Introduction (p. 2) describes mobile edge computing as a system similar to fog computing and says fog computing lacks a standardized definition. It does not say that various authors subsequently used the two terms interchangeably or “mixed and matched” them completely. (LLM, conf=0.83)

  Experimental — accuracy not yet measured
• cites [112] Natraj, 2016, Fog computing” focusing on users at the edge of internet of things, Int. J. Eng. Res. SpecialNot checked yet

  Not checked in the free run.

• cites [135] Orsini, 2015, Computing at the mobile edge: designing elastic android applications for computation offloading sourceNot checked — no full text of the source
• cites [136] Vallati, 2016, Mobile-edge computing come home connecting things in future smart homes using lte device-to-device communications, IEEE Consumer Electron. Magaz. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Fog computing is distinguished from cloudlets by its explicit linkage to fixed devices with connectivity and intelligent data analyzing and processing capabilities.

“This explicit linkage to fixed devices with connectivity and intelligent data analysing and data processing capabilities clearly distinguished FC from cloudlets [7].”

• cites [7] Aslanpour, 2020 sourcePartly supported

  Section 2.2 (page unavailable) describes fog computing as edge computing in which IoT devices can gain computational capabilities. Section 3.0.1 (page unavailable) also describes edge nodes with communication and computation capabilities. Neither passage says fog computing is explicitly linked to fixed devices or distinguishes it from cloudlets; searches for “cloudlet” and the claimed distinction found no such comparison. (LLM, conf=0.78)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Dense geographical distribution with a local focus can analyze big data faster.

“The great value of this approach was noted as the dense geographical distribution with a local focus can analyse big data faster [105].”

• cites [105] Saharan, 2015, Fog in comparison to cloud: a survey, Int. J. Comput. Appl. sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

Open Edge Computing concluded that the aim should be the development of an open platform for edge computing aligned with the IT and telecommunications industries.

“In 2016, Open Edge Computing discussed gaps in the current raft of technologies and concluded that the aim should be the development of an open platform for Edge Computing aligned with the IT and telecommunications industries [55].”

• cites [55] Gill, 2019, Router: fog enabled cloud based intelligent resource management approach for smart home iot devices, J. Syst. Software sourceCould not verify

  The Introduction (page unavailable) discusses extending cloud computing to the network edge, while Related Work (page unavailable) addresses fog-computing resource management. Neither those passages nor searches for Open Edge Computing, an open platform, and telecommunications-industry alignment address the attributed 2016 conclusion. (LLM, conf=0.75)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Trust in the security of personal and socially sensitive data would increase if management of such data were ensured at the edge rather than in distant data centers.

“These authors also asserted that trust in the security of personal and socially sensitive data would be increased if the management of such sensitive data could be ensured at the edge rather than being centred in distant data centres [114].”

• cites [114] Golec, 2020, Biosec: a biometric authentication framework for secure and private communication among edge devices in iot and industry 4.0, IEEE Consumer Electron. Magaz. sourceCould not verify

  The Introduction and BioSec System sections discuss protecting personal biometric data through authentication and encryption, including storage in a server database. Searches for trust and for edge-based management versus distant data centers found no passage asserting that managing sensitive data at the edge would increase trust. (LLM, conf=0.74)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Nokia's MEC video surveillance application enables security personnel to analyze unusual activities from anywhere at any time.

“One example of a MEC application from Nokia is video surveillance which enables security personnel the ability to analyse unusual activities from anywhere at any time [131].”

• cites [131] Kumar, 2022, Securing the future internet of things with post-quantum cryptography, Secur. Priv. sourceCould not verify

  The Introduction (p. 1) describes IoT security and post-quantum cryptography, not Nokia’s MEC applications. Searches for Nokia, video surveillance, and remote analysis of unusual activities found no passage addressing the claim in our copy. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Mobile edge cloud computing is conceptually indistinguishable from mobile edge computing.

“A variant term -“mobile edge cloud computing” -has also been used, but this is conceptually indistinguishable from MEC: “mobile-edge cloud computing can provide cloud-computing capabilities at the edge of pervasive radio access networks in close proximity to mobile users” [126].”

• cites [126] Chen, 2016, Efficient multi-user computation offloading for mobile-edge cloud computing, IEEE/ACM Trans. Netw. sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

Analyzing data close to the device that collected it can make the difference between averting disaster and a cascading system failure.

“Analysing data close to the device that collected the data can make the difference between averting disaster and a cascading system failure”

• cites [6] Aslanpour, 2021, Serverless Edge Computing: Vision and Challenges sourcePartly supported

  Section 2.2 says computing closer to data sources helps meet low-latency and real-time requirements, and §4.5 says mission-critical IoT applications may fail to react promptly without rapid processing. Section 5.10 warns that unhandled function or edge-node failures can cause a disaster, but it does not say that analyzing data near its collecting device averts a cascading system failure. Page numbers were not available in our copy. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Micro data centers form more of a hardware solution for edge computing than a novel IT scenario.

“In contrast to other edge computing models, mDCs form more of a hardware solution for Edge Computing than a novel IT scenario [16].”

• cites [16] Iftikhar, 2023, Hunterplus: ai based energy-efficient task scheduling for cloud–fog computing environments, Internet Things sourceCould not verify

  The Introduction discusses cloud–fog computing and hardware costs, while §3.1 describes a fog datacenter with edge and cloud layers. Neither passage characterizes micro data centers as a hardware solution rather than a novel IT scenario; searches for that distinction and for mDCs found no passage addressing it. Page numbers were not available in our copy. (LLM, conf=0.75)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Vehicle edge computing places edge functionality on a mobile node via a gateway to the intra-vehicle network.

“This application considers a “Vehicle Edge Computing” where a gateway to the intra-vehicle network as the edge device and hence placing the edge functionality on a mobile node.”

• cites [35] Ghafouri, 2022, Mobile-kube: mobility-aware and energy-efficient service orchestration on kubernetes edge servers sourceCould not verify

  The Introduction and II. Motivating Scenario describe mobile users connecting to edge servers near base stations, not a gateway to an intra-vehicle network acting as an edge device. Searches for “Vehicle Edge Computing” and vehicle gateways found no passage addressing that arrangement in our copy. (LLM, conf=0.72)

  Experimental — accuracy not yet measured

May combine two claims.

MEC can move mobile networks into real-time location-based gaming by bringing cloud computing into the radio access network, allowing computationally demanding AR applications to run close to the end user and avoiding high latency, high real-time responses, and backhaul bottlenecks.

“MEC can shift mobile networks into real-time location-based gaming, bringing cloud computing capabilities inside a radio access network and this allows delivery of computationally demanding AR applications running close to the end user by avoiding high latency, high real-time responses and backhaul network bottlenecks [61].”

• cites [61] Golec, 2021, Ifaasbus: a security-and privacy-based lightweight framework for serverless computing using iot and machine learning, IEEE Trans. Ind. Inf. sourceCould not verify

  Split into 2 atoms — 0 confirmed, 0 partially supported, 0 disputed, 0 misattributed, 2 unverifiable. Expand the claim to see per-atom verdicts.

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Edge AI has a wide range of applications across various industries.

“Edge AI has a wide range of applications in various industries [54].”

• cites [54] Iftikhar, 2022, Fogdlearner: a deep learning-based cardiac health diagnosis framework using fog computing, Austral. Comput. Sci. Week sourcePartly supported

  The Abstract describes an AI-based cardiac diagnosis system using fog computing, which supports healthcare as one application. Its reference to IoT adoption in various domains does not establish that Edge AI has applications across various industries; the Introduction likewise focuses on healthcare. The indexed passages provide no page numbers. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

In smart parking, data from sensors is transferred to edge devices efficiently and reliably.

“For example, in smart parking scenarios where data is coming from sensors and transferred to edge devices efficiently and reliably [66].”

• cites [66] Singh, 2022, Machine learning for cloud, fog, edge and serverless computing environments: comparisons, performance evaluation benchmark and future directions, Int. J. Grid Util. Comput. sourcePartly supported

  Section 5.1 (page unavailable) describes sensor data being forwarded to gateway devices such as tablets, laptops and mobile phones; the Introduction identifies mobile phones and personal computers as potential edge devices. The copy does not establish this transfer in smart parking or say it is reliable, and section 7.1 treats reliability as a parameter for future work. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Edge AI can power smart home devices such as voice assistants, thermostats, and security cameras.

“Smart Home: Edge AI can be used to power smart home devices such as voice assistants, thermostats, and security cameras [161].”

• cites [161] Ke, 2020, A smart, efficient, and reliable parking surveillance system with edge artificial intelligence on iot devices, IEEE Trans. Intell. Transport. Syst. sourcePartly supported

  The paper’s abstract (p. 1) describes AI processing video locally on IoT devices for parking surveillance, supporting the general idea of edge AI in a camera-based system. It does not address smart homes, voice assistants, thermostats, or home security cameras; its tested setting is a parking garage. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Edge AI processes and analyses data locally, in real time, at the edge of the network rather than transmitting it to a remote server.

“This means that data is processed and analysed locally, in real-time, at the edge of the network, rather than being transmitted to a remote server for processing [185].”

• cites [185] Shafique, 2021, Towards energy-efficient and secure edge ai: a cross-layer framework iccad special session paper sourcePartly supported

  Section II (p. 2) describes efficient, real-time data processing at the edge, and the Key Challenges section (p. 1) says Edge AI systems are expected to provide real-time responses. The copy does not explicitly say that data is processed locally instead of being transmitted to a remote server, so that contrast is not fully supported. (LLM, conf=0.82)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Various companies are working on a solution to provide city-centre parking availability via an up-to-date heat map.

“Currently, various companies are working on a solution to provide a service to car drivers requiring parking spaces in city centres by providing an up-to-date “heat map” of availability [153]”

• cites [153] Carnelli, 2017, Parkus: a novel vehicle parking detection system sourcePartly supported

  The Introduction (p. 1) describes real-time parking information and an envisaged city road map showing where drivers should search. The Conclusions and Future Work (p. 6) say visualising city parking occupancy remains a future feature. Neither passage establishes that various companies are currently developing an up-to-date availability heat map. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

May combine two claims.

Dynamic load balancing and developing resource allocation platforms are central to resource management.

“Central to resource management are dynamic load balancing [166] and developing resource allocation platforms [16].”

• cites [166] Ningning, 2016, Fog computing dynamic load balancing mechanism based on graph repartitioning, China Commun. sourceNot checked — no full text of the source
• cites [16] Iftikhar, 2023, Hunterplus: ai based energy-efficient task scheduling for cloud–fog computing environments, Internet Things sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

Edge AI has the potential to revolutionize industries by enabling intelligent and autonomous devices that make real-time decisions based on sensor data.

“Edge AI has the potential to revolutionize industries by enabling intelligent and autonomous devices that can make real-time decisions based on sensor data [164].”

• cites [164] Surianarayanan, 2023, A survey on optimization techniques for edge artificial intelligence (ai), Sensors sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

There are two main types of simulation packages: continuous simulation and discrete event simulation.

“There are two main types of simulation packages available: continuous simulation and discrete event simulation [55].”

• cites [55] Gill, 2019, Router: fog enabled cloud based intelligent resource management approach for smart home iot devices, J. Syst. Software sourceCould not verify

  Section 4 (Performance Evaluation) describes using CloudSim’s event-simulation functions, but does not classify simulation packages into continuous and discrete-event types. Searches for the claimed two-type classification and its terms found no passage in our copy that addresses it; no page number is available for the passage. (LLM, conf=0.74)

  Experimental — accuracy not yet measured

May combine two claims.

Edge computing can increase confidentiality and security by lowering the amount of information transferred over the Internet and by offering more control over processes and storage of confidential documents.

“Better Privacy/Security: Edge computing can assist in increasing confidentiality and security by lowering the quantity of information that has to be transferred over the Internet and by offering more command over the processes and storage of confidential documents [164].”

• cites [164] Surianarayanan, 2023, A survey on optimization techniques for edge artificial intelligence (ai), Sensors sourcePartly supported

  The Introduction (page unavailable) says edge computing moves processing and storage away from centralized servers, can prevent data transfer to centralized infrastructure, and can keep financial data at the edge for security and privacy. It does not specifically discuss greater control over the processing and storage of confidential documents. (LLM, conf=0.82)

  Experimental — accuracy not yet measured

May combine two claims.

Client Edge Computing, including CCTV cameras, mobile phones, and smart sensors, can perform some processing rather than sending all data to the edge network, and this model can help service providers.

“Client Edge Computing, including CCTV cameras, mobile phones and smart sensors, can perform some processing rather than sending all data to the edge network; therefore, this model can be helpful to service providers [49].”

• cites [49] Malik, 2021, Effort: energy efficient framework for offload communication in mobile cloud computing, Software Pract. Ex. sourceCould not verify

  The Introduction discusses smartphones and decisions about what data to offload, while the Discussion describes offloading communication from mobile devices to the cloud. Neither passage addresses Client Edge Computing, local processing by CCTV cameras or smart sensors instead of sending all data to an edge network, or benefits to service providers. Searches for those terms and for local processing returned no passage establishing the attributed claim. (LLM, conf=0.73)

  Experimental — accuracy not yet measured

May combine two claims.

Cloudlets initiated developments in edge computing but have proved marginal in commercial practice, though they could find implementations in municipality-centred schemes based on free access and open-source software, unlike subscription or fee-based services in other edge computing paradigms.

“Although cloudlets initiated these developments in edge computing, they have proved marginal in commercial practice but could find implementations in municipality-centred schemes based on free access and open-source software, in contrast to the subscription or fee-based services in other edge computing paradigms [66].”

• cites [66] Singh, 2022, Machine learning for cloud, fog, edge and serverless computing environments: comparisons, performance evaluation benchmark and future directions, Int. J. Grid Util. Comput. sourceCould not verify

  Split into 2 atoms — 0 confirmed, 0 partially supported, 0 disputed, 0 misattributed, 2 unverifiable. Expand the claim to see per-atom verdicts.

  Experimental — accuracy not yet measured

May combine two claims.

mDCs have an already defined segment in the Industrial IoT but could readily be deployed to support FC, MEC, and even Cloudlets, although Cloudlets would require public funding if open-access platforms are implemented.

“In contrast, mDCs have an already defined segment in the Industrial IoT but could readily be deployed to support FC, MEC and even Cloudlets (although CL would require some form of public funding if open-access platforms are to be implemented) [35].”

• cites [35] Ghafouri, 2022, Mobile-kube: mobility-aware and energy-efficient service orchestration on kubernetes edge servers sourceCould not verify

  The Introduction (page unavailable) discusses MEC, fog computing and cloudlets, but does not address mobile data centers (mDCs) or their use in Industrial IoT. Searches for mDC deployment and public funding for open-access cloudlets found no passage supporting those claims. (LLM, conf=0.77)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

With the right combination of hardware and software, Edge AI can enable AI algorithms to run on edge devices, bringing intelligence to the network's edge and enabling new applications and use cases.

“With the right combination of hardware and software, Edge AI technology can enable AI algorithms to run on edge devices, bringing intelligence to the network's edge and enabling new applications and use cases [32].”

• cites [32] Letaief, 2021, Edge artificial intelligence for 6g: vision, enabling technologies, and applications, IEEE J. Sel. Area. Commun. sourceSupported

Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing), S4 (based).

ICN defines data chunks so the network interprets and treats content semantically without needing deep packet inspection or the application layer.

“Data”

• cites [88] Gill, 2021, A manifesto for modern fog and edge computing: vision, new paradigms, opportunities, and future directions sourceCould not verify

  Section 5.9 (p. 11) discusses processing user data in fog and edge computing, and Section 5.13 (p. 12) discusses software-defined networking. Neither addresses ICN-defined data chunks, semantic interpretation of content, or avoiding deep packet inspection and the application layer. Searches for ICN, information-centric networking, and deep packet inspection found no passage supporting the claim. (LLM, conf=0.74)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

The evolving discussion on Edge Computing has identified four distinct paradigms differentiated by intended markets and business models.

“The evolving discussion on Edge Computing has identified four distinct paradigms differentiated on the basis of intended markets and business models [59].”

• cites [59] Nabavi, 2022, Tractor: traffic-aware and power-efficient virtual machine placement in edge-cloud data centers using artificial bee colony optimization, Int. J. Commun. Syst. sourceCould not verify

  The Introduction discusses cloud-computing services, and the Performance Evaluation discusses edge-cloud data-center topologies, but neither identifies four edge-computing paradigms. Searches for the claimed paradigms and their differentiation by intended markets or business models found no supporting passage in our copy; page numbers were not available. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Moving resources closer to the fixed edge represented by surveillance cameras, traffic flow enhancement systems, and smart meters is a feature of fog computing.

“Sectors: Moving resources closer to the”

• cites [13] Singh, 2021, Fog computing: a taxonomy, systematic review, current trends and research challenges, J. Parallel Distr. Comput. sourcePartly supported

  Section A. Context says fog computing places storage, computing, and networking functions near connected devices, supporting the claim’s central point. The paper also discusses video surveillance (Development) and fog-based traffic-signal control (Platforms), but the passages do not establish the claimed ‘fixed edge’ category or identify smart meters as an example. (LLM, conf=0.83)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Edge AI technology is rapidly evolving, with new techniques and frameworks being developed to improve performance, efficiency, and security of Edge AI solutions.

“The technology is rapidly evolving, with new techniques and frameworks being developed to improve the performance, efficiency, and security of Edge AI solutions [62].”

• cites [62] Mazzia, 2020, Real-time apple detection system using embedded systems with hardware accelerators: an edge ai application, IEEE Access sourcePartly supported

  The ‘AI on the Edge and Related Work’ section (p. 2) describes hardware innovations aimed at improving Edge AI performance and power efficiency. The ‘Methodology and Architecture Framework’ section (p. 4) describes successive, faster and more accurate detection models suitable for edge devices. These passages support the claim about evolving techniques and efficiency, but our copy does not address improvements to Edge AI security. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Vehicular Edge Computing introduces a computation offloading mechanism where embedded devices in cars offload data to MEC servers.

“Vehicular Edge Computing introduces a computation offloading mechanism with embedded devices in cars offloading data to MEC servers [46].”

• cites [46] Iftikhar, 2022, Fog computing based router-distributor application for sustainable smart home sourceCould not verify

  The Introduction and simulation-results sections discuss smart-home devices sending data to a home router or cloud facility. Searches for vehicular edge computing, cars, and MEC servers found no passage addressing the claimed vehicle-to-MEC offloading mechanism; the smart-home example does not establish that specific claim. (LLM, conf=0.75)

  Experimental — accuracy not yet measured

May combine two claims.

Mobile MEC users form a transient population with varying requirements.

“Mobile MEC users form a transient population with varying requirements [96].”

• cites [96] Lindsay, 2021, The evolution of distributed computing systems: from fundamental to new frontiers, Computing sourceCould not verify

  The “Cloud, Mobile & IoT” and “Fog and Edge Computing” sections discuss mobile devices and edge infrastructure, but do not describe MEC users as a transient population or say their requirements vary. Searches for those user characteristics found no passage addressing them in our copy. (LLM, conf=0.72)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Machine learning frameworks such as TensorFlow Lite and PyTorch are used to develop and run machine learning models on edge devices.

“Machine learning frameworks: ML frameworks, such as TensorFlow Lite and PyTorch, are used to develop and run machine learning models on edge devices [202].”

• cites [202] Liu, 2022, Bringing ai to edge: from deep learning’s perspective, Neurocomputing sourcePartly supported

  The Quantization section (page unavailable) names TensorFlow Lite among tools supporting efficient execution of quantized models for resource-constrained edge systems. It also identifies PyTorch as a framework supporting post-training quantization, but does not say PyTorch is used to develop or run models on edge devices. The cited passages therefore support the edge-use claim for TensorFlow Lite, but not its full application to both examples. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

May combine two claims.

Strategies for avoiding overloading and disaster management procedures for recovering non-functional base stations were central to Multi-access Edge Computing.

“Central to MEC were strategies for avoiding overloading and “disaster management” procedures for recovering non-functional base stations [96].”

• cites [96] Lindsay, 2021, The evolution of distributed computing systems: from fundamental to new frontiers, Computing sourceCould not verify

  The ‘Fog and Edge Computing (2010-present)’ section discusses edge infrastructure and decentralized workflows, but not Multi-access Edge Computing strategies for preventing overload or recovering base stations. Searches for disaster management, non-functional base stations, and MEC overload found no passage addressing those procedures. Page numbers were unavailable in our copy. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

May combine two claims.

Nascent commercial opportunities have been identified for mDCs and in proof of concept studies for MEC.

“Nascent commercial opportunities have been identified for mDCS and in proof of concept studies for MEC [49].”

• cites [49] Malik, 2021, Effort: energy efficient framework for offload communication in mobile cloud computing, Software Pract. Ex. sourceCould not verify

  The Introduction discusses projected business use and revenue for mobile cloud computing, while the Conclusions describe prospects for cloud computing generally. Neither section identifies commercial opportunities for mDCs or proof-of-concept studies for MEC; searches for those specific terms and ideas found no supporting passage in our copy. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Finding and accessing appropriate nodes in an Edge Computing network, discovery, and trust must be considered together, especially for mobile devices.

“Finding and accessing the appropriate nodes in an Edge Computing network and discovery and trust are features that must be considered together, especially for mobile devices [14].”

• cites [14] Garcia Lopez, 2015, Edge-centric computing: vision and challenges, Comput. Commun. Rev. sourceSupported

The quote only partly states this claim.

This improved security leads to improved uptime and dependability for AI applications essential to business operations.

“The upshot of this is improved uptime and dependability for AI applications that are essential to business operations [164].”

• cites [164] Surianarayanan, 2023, A survey on optimization techniques for edge artificial intelligence (ai), Sensors sourcePartly supported

  The Conclusions (page unavailable) say edge AI can deliver reliable functions with low latency, which supports the dependability aspect. The Introduction discusses data security and privacy, but the paper does not say that improved security increases uptime or dependability for business-essential AI applications. (LLM, conf=0.82)

  Experimental — accuracy not yet measured

May combine two claims.

By keeping data on-premises, Edge AI protects users' confidentiality, sending only the results of analyses and observations to the cloud.

“By keeping such data on-premises, Edge AI additionally protects users' confidentiality, sending just the results of their analyses and observations to the cloud [185].”

• cites [185] Shafique, 2021, Towards energy-efficient and secure edge ai: a cross-layer framework iccad special session paper sourceCould not verify

  The security discussion (Section III, pp. 3–5) addresses vulnerabilities and protection against attacks, but does not describe keeping users’ data on-premises or sending only analysis results to the cloud. Searches for local processing, confidentiality, and cloud transmission found no passage establishing that data-sharing arrangement. (LLM, conf=0.72)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Edge computing has already demonstrated active participation of different industrial interests, namely hardware and IT (mDCs and FC) and telecommunications (MEC).

“Industry players: Edge computing has already demonstrated the active participation of different industrial interests, namely hardware and IT (mDCs and FC) and telecommunications (MEC) [31].”

• cites [31] Moqurrab, 2022, A deep learning-based privacy-preserving model for smart healthcare in internet of medical things using fog computing, Wireless Pers. Commun. sourceCould not verify

  The Introduction (page unavailable) describes fog computing as an intermediary layer at the network edge, but does not discuss participation by hardware, IT, or telecommunications companies. Searches for industrial interests, mDCs, FC, and MEC found no passage in our copy supporting that claim. (LLM, conf=0.75)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Bodies such as ETSI have extended the use of the word 'mobile' to include immobile but rotatable devices such as surveillance cameras.

“In addition, bodies such as ETSI have extended the use of the word “mobile” to include devices such as surveillance cameras which are immobile (but rotatable) [13].”

• cites [13] Singh, 2021, Fog computing: a taxonomy, systematic review, current trends and research challenges, J. Parallel Distr. Comput. sourceCould not verify

  The Expectation/Interest section discusses mobility in fog computing, and the Development section mentions cameras on vehicles, but neither defines stationary, rotatable cameras as “mobile.” Searches for ETSI, surveillance cameras, and immobile or rotatable devices found no passage addressing the attributed definition. The retrieved passages have no page numbers. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Not yet classified.

In a MEC overload recovery scenario, as the number of relay devices and their data transfer rate increase, data throughput and the number of disconnected mobile users that can be serviced increase.

“As the number of relay devices increases and as their rate of data transfer increases, data throughput and the number of disconnected mobile users capable of being serviced increase greater the number of relay devices per disconnected user, the faster the recovery system works [16].”

• cites [16] Iftikhar, 2023, Hunterplus: ai based energy-efficient task scheduling for cloud–fog computing environments, Internet Things sourceCould not verify

  The Introduction discusses migrating jobs between cloud-fog servers, and §4.7 analyzes host overload and task allocation; neither addresses relay devices or disconnected mobile users. Searches for MEC overload recovery, relay counts, data-transfer rates, and throughput found no passage in our copy supporting the attributed relationship. Page numbers were not available for these passages. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

• AssumptionThe authors state this as an assumption: In the MEC overload recovery scenario, the disconnected mobile users are out of range of the recovery MEC server.
• AssumptionRests on an assumption: In the MEC overload recovery scenario, mobile devices directly connected to the recovery MEC server act as relay nodes and communicate with the disconnected mobile users.
• AssumptionRests on an assumption: In the MEC overload recovery scenario, the pathway between the recovery MEC server and the disconnected mobile users goes via mobile devices directly connected to the second server.

Classified as probably not a substantial claim. May combine two claims.

This unlocks new possibilities for real-time decision making as well as enhanced confidentiality and safety.

“This unlocks new possibilities for real-time making decisions as well as enhanced confidentiality and safety [81].”

• cites [81] Yang, 2022, Lead federated neuromorphic learning for wireless edge artificial intelligence, Nat. Commun. sourcePartly supported

  The Introduction and Discussion (page not available) say LFNL enhances privacy and reduces training latency; the Results also report resistance to model-poisoning attacks. Those findings support the confidentiality aspect, but the paper does not establish real-time decision-making or improved safety. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim.

Implementing AI systems in situations with limited resources, such as edge computing and other resource-constrained settings, brings a mix of benefits and constraints.

“In summary, implementing AI systems in situations with limited resources, such as edge computing and other resource-constrained settings, brings a mix of benefits and constraints [76].”

• cites [76] Gia, 2019, Edge ai in smart farming iot: cnns at the edge and fog computing with lora sourcePartly supported

  The Introduction describes advantages of edge and fog computing, including energy efficiency, while noting connectivity and data-rate constraints in remote farming. The Edge layer section describes a CNN-based approach that reduces data sent over LoRa. These passages support a mix of benefits and constraints for this edge-AI application, but do not establish the claim for other resource-constrained settings generally. (LLM, conf=0.82)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

Hardware enterprises must be flexible on precise mechanisms for implementing Multi-Access Edge Computing because a wide range of applications will be recognized with different requirements and market sizes.

“In particular, the implementation of Multi Access Edge Computing has a clear business case for service providers to develop services and products for “vertical” markets, i.e.where an entire sector such as healthcare is considered as a whole; hardware enterprises must also be flexible on precise mechanisms for implementing Multi-Access Edge Computing because a wide range of applications will be rec”

• cites [223] RayalNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites [224] Singh, 2021, Towards multi-criteria heuristic optimization for computational offloading in multi-access edge computing sourceCould not verify

  The Introduction discusses computational offloading in MEC, and the Conclusions describe a flexible offloading method that accommodates users’ choices. Neither section addresses hardware enterprises, implementation mechanisms, or applications with different market sizes. Searches for those terms and for vertical markets found no supporting passage in our copy; page numbers were not provided. (LLM, conf=0.74)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Grey areas in Edge Computing definitions have allowed overlapping views that over-claim for particular paradigms; for example, one source defines Edge Computing too narrowly as connected embedded computing for the Industrial Internet of Things, while another brings all mobile users into Fog Computing.

“These grey areas have allowed overlapping views that over-claim for particular paradigms; for example [221], defines Edge Computing too narrowly: “connected embedded computing for the Industrial Internet of Things (IIoT)” while [52] brings all mobile users into FC.”

• cites [221] Steve Jennis, 2017Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites [52] T.H. Luan, L. Gao, Z. Li, Y. Xiang, G. Wei, L. Sun, Fog computing: focusing on mobile users at the edge, arXiv preprint arXiv:1502.01815 (2015), 1-11. sourceSupported

May combine two claims.

Financial investments required and opportunities are key factors in decision making and implementation routes for edge computing.

“The financial investments required are key items in decision making but, equally, opportunities are major factors in implementation routes for Edge Computing (Table 4).”

May combine two claims.

A single-device cloudlet with one mobile user achieved reductions in power usage by 50%, a nearly 10-fold reduction in delay time, and a 10-fold increase in throughput.

“In a separate practical test of cloudlet versus cloud, reductions in power usage by 50%, a nearly 10-fold reduction in delay time and an increased throughput of 10-fold (Table 3) were achieved using a single-device cloudlet with one mobile user [141].”

• cites [141] Jararweh, 2013, Resource efficient mobile computing using cloudlet infrastructure sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

ETSI-sponsored proofs of concept showed the viability of the mobile edge computing concept.

“Table 2 shows the ETSI-sponsored proofs of concept (PoCs) that showed the viability of the MEC concept.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The principal differentiators for applications are mobile devices versus fixed devices, means of access, application domains, and drivers.

“The principal differentiators for applications are mobile devices versus fixed devices, means of access, application domains and drivers (Table 5).”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

The renaming of MEC as Multi-access Edge Computing reflects aims for applications development in 2017 for non-mobile devices.

“The renaming of MEC as Multi-access Edge Computing reflects aims for applications development in 2017 for non-mobile devices; this is a crucial change of direction and its full implications will be discussed later in this survey.1”

Probably not a substantial claim: may not pass S3 (load-bearing).

With recent AI efficiency innovations, IoT device proliferation, and the rise of edge computing, the potential of edge AI has been unlocked.

“With the latest innovations in Artificial Intelligence (AI) efficiency, the proliferation of IoT devices, and the rise of edge computing, the potential of edge AI has now been unlocked [19].”

• cites [19] Shi, 2020, Communication-efficient edge ai: algorithms and systems, IEEE Commun. Surv. Tutor. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S4 (based).

Overlap between the four edge computing concepts in the literature has significantly degraded the boundaries between the different approaches.

“Overlap between the four concepts as they have been discussed in the literature has, however, resulted in significant degradation of the boundaries between the different approaches.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The proliferation of IoT-connected devices is largely responsible for the current era's unprecedented growth in data volume.

“The proliferation of IoT-connected devices is largely responsible for the current era's unprecedented growth in data volume [34].”

• cites [34] Raychaudhuri, 2022, Green internet of things using mobile cloud computing: architecture, applications, and future directions sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Edge AI could be provided with computational intelligence to develop intelligent systems that perform tasks typically requiring human-level intelligence, such as decision-making, problem-solving, pattern recognition, and learning.

“Furthermore, Edge AI could be provided with the computational intelligence to develop intelligent systems that can perform tasks that typically require human-level intelligence, such as decision-making, problem-solving, pattern recognition, and learning [79].”

• cites [79] Singh, 2017, The journey of qos-aware autonomic cloud computing, It Profession. sourceNot checked — no full text of the source

May combine two claims.

The aforementioned figures suggest AI computation on edge devices will grow exponentially, allowing consumers to engage with compiled code at the source safely and optimally.

“This suggests that AI computation on such devices will grow at an exponential rate, allowing consumers to engage with compiled code at the source in a safe and optimal manner [75].”

• cites [75] Lv, 2022, An edge-ai based forecasting approach for improving smart microgrid efficiency, IEEE Trans. Ind. Inf. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Computational offloading faces challenges, particularly how a wireless device decides which tasks to perform locally or hand over to the edge server, and what happens if the edge server overloads due to high demand.

“However, computational offloading still faces many challenges, in particular, how a wireless device will decide which tasks need to perform locally or which to hand over to the edge server.Another critical issue is on the edge server side: what happens if the edge server starts to overload due to the high demand from the wireless devices?”

May combine two claims.

Smart sensors make decisions, improving overall network performance and sending only useful data to the cloud.

“Smart sensors make decisions and this improves the performance of the overall network and sends only useful data to the cloud [56].”

• cites [56] Kaur, 2018, The future of cloud computing: opportunities, challenges and research trends sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

This case study overlaps Fog Computing and MEC.

“This case study overlaps FC, as discussed in (Sec.II-F) and (with images uploaded from mobile devices) MEC (Sec II-G).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Between 2009 and 2014, four Edge Computing paradigms-Cloudlets, Fog Computing, Mobile Edge Computing, and Micro Data Centres-emerged as practical contenders for implementation.

“Between 2009 and 2014, four paradigms for Edge Computing Cloudlets, Fog Computing, Mobile Edge Computing and Micro Data Centres emerged as practical contenders for implementation of the Edge Computing concept [44].”

• cites [44] Dhillon, 2022, Iotpulse: machine learning-based enterprise health information system to predict alcohol addiction in Punjab (India) using iot and fog computing, Enterprise Inf. Syst. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The original fixed-location cloudlet has become one component in an intervening layer between mobile users and the cloud, offering better performance for mobile users regardless of physical proximity to a closer server.

“In other words, the original cloudlet in a fixed location has become one of many components in an intervening layer between the mobile user and the cloud but offering better performance for the mobile user, whether or not the user is connected to a geographically closer server (i.e. the “logical proximity” effect).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Edge computing aims to move data processing from distant consolidated cloud data centres to servers usually physically closer to end users to support high QoE applications for heterogeneous mobile device users and fixed Internet-connected streaming devices via wireless networks.

“Edge computing proposes moving data processing capability away from distant consolidated data centres in the cloud to servers usually (but not exclusively) physically closer to the end user in order to support high QoE applications for heterogeneous mobile device users and fixed Internet-connected streaming devices via wireless networks [43].”

• cites [43] Carvalho, 2021, Edge computing: current trends, research challenges and future directions, Computing sourceNot checked — no full text of the source

Classified as probably not a substantial claim. May combine two claims.

A more rapid solution is to bring computation capability close, and Edge Computing can provide limited computation capability to make quick lower-latency decisions.

“A more rapid solution is to bring computation capability close and Edge Computing can provide limited computation capability to make quick (lower latency) decisions.”

Probably not a substantial claim: may not pass S3 (load-bearing).

IoT benefits from improved speed, reliability, and security brought by 5G/6G.

“IoT is also benefiting from the improved speed, reliability, and security that 5G/6G is bringing to the battlefield [36].”

• cites [36] Adhikari, 2021, A roadmap of next-generation wireless technology for 6g-enabled vehicular networks, IEEE Internet Things Magaz. sourceNot checked — no full text of the source

May combine two claims.

Cloudlets as a business concept have failed to gain traction, and the default offering from small enterprises has become free Wi-Fi for mobile devices.

“Cloudlets as a business concept have failed to gain traction and the default offering from small enterprises has become free Wi-Fi for mobile devices (including laptops and notebooks) [91][92][93].”

• cites [91][92][93]Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Existing survey articles [39][40][41] introduced detailed surveys on edge computing focusing only on IoT applications, but [41] also discussed resource management in edge computing.

“Existing survey articles [39][40][41] introduced detailed surveys on edge computing by focusing on only IoT applications but [41] has also discussed about resource management in edge computing.”

• cites [39][40][41]Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

When AI models are placed on the Edge instead of the cloud, Edge AI devices can analyse data faster, make decisions faster, increase data processing security, and improve user experience.

“When AI models are put on the Edge instead of in the cloud, we allow Edge AI devices to analyse data more quickly, make decisions more quickly, increase the security of data processing, and improve the user experience [71].”

• cites [71] Lin, 2019, Making knowledge tradable in edge-ai enabled iot: a consortium blockchain-based efficient and incentive approach, IEEE Trans. Ind. Inf. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Deploying an unspecified number of smaller cloudlet servers to provide diffuse coverage ensures high quality of experience.

“The implication of these findings was that an unspecified number of smaller cloudlet servers would be positioned to provide diffuse coverage to ensure a high QoE [5].”

• cites [5] Iftikhar, 2022Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Edge computing has roots in 1990s content delivery networks that used edge servers near customers to provide web and gameplay videos.

“Numerous experts and companies are discussing and adopting edge computing, which has its roots in the 1990s with the advent of content delivery networks that employ edge servers located near customers to provide web and gameplay videos [21].”

• cites [21] Wang, 2020 sourceNot checked — no full text of the source

May combine two claims.

Currently, intelligent transportation uses cloud computing where all processing is done at remote data centres, but cloud computing lacks key safety features in smart transportation.

“Currently, an intelligent transportation environment uses cloud computing, where all the processing is done at remote data centres, but cloud computing lacks key safety features in smart transportation [56].”

• cites [56] Kaur, 2018, The future of cloud computing: opportunities, challenges and research trends sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Smart cities and smart factories would not have been possible without edge computing, which has become essential to the Industrial Internet.

“Smart cities and smart factories would not have been possible without edge computing, which has emerged as an essential part of the Industrial Internet [50].”

• cites [50] Lovén, 2019, Edgeai: a vision for distributed, edge-native artificial intelligence in future 6g networks, 1st 6G wireless summit.Not checked — no full text of the source

May combine two claims.

An edge server allows a wireless device to transfer tasks for offloading, substantially reducing latency and improving bandwidth between wireless users and the edge server.

“The solution is an edge server: a wireless device can transfer a series of tasks to the edge server for offloading.Therefore, this substantially reduces the latency and improves the bandwidth between the wireless users and the edge server.”

May combine two claims.

Edge AI allows faster and more efficient data processing, as well as improved privacy and security.

“This allows for faster and more efficient data processing, as well as improved privacy and security [78].”

• cites [78] Doyle, 2022, Blockchainbus: a lightweight framework for secure virtual machine migration in cloud federations using blockchain, Secur. Priv. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S2 (checkable).

A new innovative study is required to analyse, update, and integrate existing research and explore prospective trends in edge computing, especially Edge AI.

“In order to analyse, update, and integrate the existing research and explore prospective trends and futuristic views in the field of edge computing, especially Edge AI, there is a requirement for a new innovative study as this area of edge computing is continuing to expand towards Edge AI.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Edge AI has enabled applications such as assisting radiologists in diagnosing diseases, driving automobiles on highways, and fertilizing plants.

“This has resulted in hitherto inconceivable applications for edge AI, such as assisting radiologists in diagnosing diseases, driving automobiles on the highway, and even fertilizing plants [20].”

• cites [20] Kamruzzaman, 2021, New opportunities, challenges, and applications of edge-ai for connected healthcare in smart cities sourceNot checked — no full text of the source

May combine two claims.

As of 2019, the worldwide edge computing industry was estimated at $3.5 billion, and by 2027 it might increase to $43.4 billion.

“As of 2019, the worldwide edge computing industry is estimated to be worth $3.5 billion.By 2027, that figure might have increased to $43.4 billion [74].”

• cites [74] Hu, 2021, Rim: offloading inference to the edge sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

M2M communication can be used when direct communication between devices or machines is required without human intervention.

“Machine-to-machine (M2M) approach could be considered when direct communication is required between devices or machines, without the need for human intervention [83].”

• cites [83] Agarwal, 2021, Edge ai driven technology advancements paving way towards new capabilities, Int. J. Innovat. Technol. Manag. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Gartner expects that by 2025 75% of enterprise-generated data will be created and processed outside conventional data centres or cloud due to IoT expansion.

“By 2025, Gartner expects 75% of enterprise-generated data will be created and processed outside of the conventional data centre or cloud, thanks to the expansion of the IoT [73].”

• cites [73] Foukalas, 2021, Edge artificial intelligence for industrial internet of things applications: an industrial edge intelligence solution, IEEE Indu. Electron. Magaz. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

The use case study is entirely within the definition of Multi-Access Edge Computing, which aims to converge telecommunication and IT services in multiple settings.

“The use case study was, however, entirely within the definition of Multi-Access Edge Computing, which aims to converge telecommunication and IT services in multiple settings [146].”

• cites [146] Taleb, 2017, On multi-access edge computing: a survey of the emerging 5g network edge architecture orchestration, IEEE Commun. Surv. Tutor. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Nokia's approach optimized bandwidth, reduced uplink congestion, decreased power requirements for uploading content, making smartphones more energy efficient and extending battery lifetimes.

“This approach optimized bandwidth and reduced uplink congestion as well as decreasing power requirements for uploading the content, thus making smartphones more energy efficient and with extended battery lifetimes.8”

• cites [8] Church, 2008Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Once the concept of a network of fixed or moving sensors streaming data to a fog computing server is lost, the paradigm is radically altered.

“Once the concept of a network of fixed or moving sensors streaming data to an FC server is lost, the paradigm is radically altered, which is discussed next [9].”

• cites [9] Gill, 2019 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

If cloudlets are essentially free at the point of use, many applications using fixed sensors are eliminated.

“If, for example, cloudlets are essentially free at the point of use, this eliminates many applications using fixed sensors [150].”

• cites [150] Shaukat, 2016, Cloudlet deployment in local wireless networks: motivation, architectures, applications, and open challenges, J. Netw. Comput. Appl. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Standardization for micro-data centres can help reduce costs, simplify deployment, and improve overall efficiency.

“It can also help reduce costs, simplify deployment, and improve overall efficiency [30].”

• cites [30] Xu, 2022Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Mobile edge computing reconfigures devices already deployed at the mobile edge as mobile access points, adding computing and storage capabilities to act as MEC servers.

“As its core concept, MEC reconfigures the devices already deployed at the mobile edge as mobile access points, i.e. base stations forward traffic but also adds computing and storage capabilities to act as MEC servers [127].”

• cites [127] Beck, 2014, Mobile edge computing: a taxonomyNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Urgent computing in disasters and emergencies is a fog computing application that can significantly speed up response times and optimize responses.

“While augmented reality can be viewed as being more in the domain of MEC than of FC, urgent computing in disaster and emergencies is an FC application that can significantly speed up response times and optimize responses [115].”

• cites [115] Tran, 2016, A systematic literature review of fog computing, Norsk konferanse for organisasjoners bruk av ITNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Mobile edge computing is clearly differentiated from cloudlets and fog computing.

“this clearly differentiates MEC from Cloudlets (the business model based on free access) and FC (the focus on mobile devices in the IoT) while again aiming to bring cloud computing capabilities closer to the user base similar to all other Edge Computing paradigms [15].”

• cites [15] PatelNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

The sports stadium scenario does not convincingly address mobile device users because users are in a precisely defined location and not freely mobile.

“This scenario (in a sports stadium) does not convincingly address mobile device users because they were in a precisely defined location and were not “mobile” in the sense of freely moving in, for example, a city centre or a large airport.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Mobile edge computing provides IT and cloud-computing capabilities within the Radio Access Network in close proximity to mobile subscribers.

“Mobile-edge Computing provides IT and cloud-computing capabilities within the Radio Access Network (RAN) in close proximity to mobile subscribers”

• cites [15] PatelNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

An optimized cloudlet network demonstrated advantages in traffic volume and reduction of latency.

“While these authors did not explicitly compare MEC and Cloud performance parameters, the results demonstrated advantages in an optimized “cloudlet network” in terms of traffic volume and reduction of latency [144].”

• cites [144] Ceselli, 2015, Cloudlet network design optimization sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

The challenge for edge computing is to provide standardized means of access for users and novel service providers as well as an efficient interface to the cloud.

“The challenge for edge computing is to provide standardized means of access for users and novel service providers as well as an efficient interface to the cloud as shown in Fig. 6.”

Classified as probably not a substantial claim.

In FemtoCloud, scheduling computational events and tasks in a transient architecture of mobile processing capacity is a major problem.

“The authors concluded from a detailed mathematical analysis that the client service faced a major problem in scheduling computational events and tasks in a transient architecture of mobile processing capacity but the design architecture enables a useable configuration and users can be given incentives to participate if the concept were to be implemented [102].”

• cites [102] Habak, 2015, Femto clouds: leveraging mobile devices to provide cloud service at the edge sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Edge computing projects and proposals are distributed across academic, industry, and standardization activities, and standardization initiatives have been active since 2009.

“Edge Computing projects and proposals are distributed across academic, industry and standardization activities and standardization initiatives have been active since 2009 [34].”

• cites [34] Raychaudhuri, 2022, Green internet of things using mobile cloud computing: architecture, applications, and future directions sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Users can be given incentives to participate in FemtoCloud if the concept were implemented.

“The authors concluded from a detailed mathematical analysis that the client service faced a major problem in scheduling computational events and tasks in a transient architecture of mobile processing capacity but the design architecture enables a useable configuration and users can be given incentives to participate if the concept were to be implemented [102].”

• cites [102] Habak, 2015, Femto clouds: leveraging mobile devices to provide cloud service at the edge sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The offloads framework distributes work overload between mobile users, reduces energy consumption, and increases battery longevity.

“This approach distributes the work overload between mobile users and reduces energy consumption, and increases battery longevity [140].”

• cites [140] S. Singh et al., “An Iiot Based Secure and Sustainable Smart Supply Chain System Using Sensor Networks,” Transactions on Emerging Telecommunications Technologies, p. e4681.Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

The solution to the IoT-cloud relationship problem is a multi-tiered architecture with at least three tiers, with parts on the thing, the fog platform, and the cloud datacenter.

“Thus, the solution to this problem is a multi-tiered architecture (with at least 3 tiers) whereby an IoT application is deployed as follows: a part on the “thing” (e.g. a car), a 2nd part on the fog platform (e.g. a roadside cabinet or a router in a wireless access network or an LTE base station), and in the case of three tiers a final 3rd part in a datacentre of the main cloud (e.g.Amazon EC2)”

• cites [109] Klas, 2015Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

Standardization for micro-data centres is important to ensure consistency, reliability, and interoperability between different components and systems.

“Standardization for micro-data centres is important to ensure consistency, reliability, and interoperability between different components and systems.”

• cites [30] Xu, 2022Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The analysis demonstrates diversity of applications and different conceptual landscapes in edge computing.

“The above analysis emphasizes the diversity of applications being investigated for Edge Computing but demonstrates the different conceptual landscapes being considered [17].”

• cites [17] Chakraborty, 2023, Journey from cloud of things to fog of things: survey, new trends, and research directions, Software Pract. Ex. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Extending the definition of 'thing' to personal mobile devices, particularly smartphones and tablets, is a dilemma.

“The dilemma lies in extending the definition of “thing” to personal mobile devices, particularly smartphones and tablets.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The FemtoCloud design architecture enables a usable configuration.

“The authors concluded from a detailed mathematical analysis that the client service faced a major problem in scheduling computational events and tasks in a transient architecture of mobile processing capacity but the design architecture enables a useable configuration and users can be given incentives to participate if the concept were to be implemented [102].”

• cites [102] Habak, 2015, Femto clouds: leveraging mobile devices to provide cloud service at the edge sourceNot checked — no full text of the source

May combine two claims.

A critical difference between fog computing and MEC is that wireless IoT networks are the principal driver for fog computing, while low latency and resource efficiency in cellular networks are dominant considerations for MEC.

“Nevertheless, different drivers have been emphasized when these two edge computing paradigms are compared, and a critical difference between FC and MEC is that wireless IoT networks can be viewed as the principal driver for FC, whilst low latency and resource efficiency in cellular networks are dominant considerations for MEC [137].”

• cites [137] SchusterNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Micro data centers are compatible with fog computing, MEC and cloudlets but are highly suitable for the industrial internet of things.

“As such, mDCs are compatible with FC, MEC and cloudlets but are highly suitable for a completely different market sector, IIoT [138].”

• cites [138] S. S. Gill, “Quantum and Blockchain Based Serverless Edge Computing: A Vision, Model, New Trends and Future Directions,” Internet Technology Letters, p. e275.Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

There are many different offered definitions for the term fog computing.

“At present, there are many different offered definitions for the term FC [123,124].”