Utilisateur:Karolinasea/Brouillon

Une page de Wikipédia, l'encyclopédie libre.

The COVID-19 pandemic gave rise to numerous ways to carry out digital contact tracing, particularly via mobile devices and smartphone applications. These new tools have been developed to accelerate the contact tracing effort, which done traditionally is slow, tedious and demands extensive resources. Digital contact tracing uses diverse technologies and protocols. The selection of technologies and type of network protocols influences the accuracy and the efficiency of digital contact tracing tools.

Tools for digital contact tracing[modifier | modifier le code]

For digital contact tracing to be efficient and accurate, technologies must be able to either detect proximity between people or locate people who came into close contact with each other, for a certain amount of time.[1], [2] In some cases it needs to be able to estimate the distance between people.[1], [2] There are several existing tools that are able to do that.

Global Positioning System (GPS)[modifier | modifier le code]

GPS technology records the coordinates of a device (latitude and longitude) and links each coordinate to a timestamp. [3], [4]

GPS can be used for digital contact tracing to retrace all the movements of an infected individual and notify people, who were in the same locations at the same times of the infected person. [4], [5] As of 2021, GPS is used by several COVID-19 tracking mobile applications such as Bulgaria's VirusSafe and India's Aarogya Setu. [6], [7], [8], [9]

Advantages[modifier | modifier le code]

GPS data provides insights into the movements of individuals. [10], [11], [4] When it is possible to trace, where an infected individual has been, then it is easier to locate most of the people they may have come into contact with.[12] This can help detecting entire areas with high infection rates. [13], [12] Moreover, numerous smartphone applications already use GPS to offer location-based services. [3] Since GPS is already available in most smartphones, it makes it a good candidate to be used for digital contact tracing. [14], [9], [15]

Disadvantages[modifier | modifier le code]

GPS works well in outdoor environments, that lack disturbances such as buildings. However GPS signals are weaker indoors, which is a problem when it comes to COVID-19 contact tracing, since the majority of infections happens inside. [16], [17], [5], [18], [19], [11], [20], [21], [22] Additionally, GPS signal doesn't provide enough precisions to determine the exact position and distance between people. [23], [24], [25], [26] For example, it hardly distinguishes people separated by walls or ceilings, which means it can lead to false positives. [23], [24], [25], [26]

GPS uses a massive amount a battery. [27], [28], [26] For effective digital contact tracing, it needs to be enabled at all time to be as exact as possible. This might cause a low user adoption rate since people may not want to use a tool that drains their device battery too quickly. [26], [27], [29]

In a privacy perspective, GPS is not ideal. [13], [30], [31] Indeed tracing the locations of people at specific times can expose their identity. [32] For example, GPS data can give away home or work addresses. [23], [33], [11], [27], [34], [35]

In a security aspect, GPS is exposed to spoofing attacks. [36], [27] This could not only lead to false positives, but also specific businesses or areas could be targeted to become considered hotspots of COVID-19 infections, to the benefit of other businesses. [36], [27] Since GPS data is not encrypted it can easily be spied on and manipulated. [36], [37] GPS is also vulnerable to jamming attacks, which can lead to false negatives. [34], [38]

Bluetooth Low Energy (BLE)[modifier | modifier le code]

Bluetooth Low Energy (BLE) distinguishes itself from classic Bluetooth. Classic Bluetooth is more of a connection-oriented protocol, used to pair devices together, whereas BLE is rather used for constant scanning and detecting other BLE devices. [39] BLE is a protocol more commonly used with Internet of Things. [40] BLE uses a method called advertising and discovery, to discover devices and be discovered by other devices around. [41], [42] Unlike GPS which records the geolocation of the users at a specific point in time, BLE equipments record which other BLE-enabled devices were close by at some point. [39], [41], [43], [44] With BLE it is possible to estimate the distance between devices by inspecting the Received signal strength indication (RSSI): The stronger the signal, the closer the device. [41], [43], [36], [45]

Advantages[modifier | modifier le code]

BLE is a widespread technology. [45], [46] It first came out in 2010 and since 2015, it has been consistently included in most smartphones and Internet of Things equipments. [45] This omnipresence makes it convenient to use for digital contact tracing. [28], [47], [41], [45]

BLE provides fairly accurate proximity estimations, particularly indoors with precisions around 1-2 meters, even though the measurements of signal strength, used to determine proximity between devices, must be adjusted, because the transmission power varies amongst different types of smartphones. [21], [48], [36], [16], [35], [49], [45] BLE works accurately both in indoor and outdoor settings. [50] BLE signals are greatly diminished when encountering a wall or other types of obstacles, which is important to get accurate data on actual physical close contact and not get false positives when people are near in terms of distance, but are actually separated by a wall or a ceiling for instance. [34]

Another main advantage of BLE, is that it does not need to be connected to any network to function (it doesn't require an Internet connection for example). [48], [50] Neither does it require devices to be paired to exchange enough information for contact tracing. [41] This connectionless aspect is also interesting to build digital tools that respect users' privacies. [44] Since the data exchanged by BLE-enabled devices is stored in the device itself, it is easier for users to control their data. [51], [52] To top that off, BLE records which devices are nearby and not actual geolocation of a device, so it is more private. [48], [45], [44] However, the BLE data stays locally on the device as long as the digital contact tracing protocol does not need it. [45] Therefore, privacy also depends on how BLE data is processed, but the main idea is that the BLE data itself does not need to go through satellites like GPS data or routers like wifi data for instances [52]

Finally, one of the main features of BLE is its low battery usage. Unlike GPS, BLE was developed to use as little power as possible, thus making it ideal for tools that need to constantly have BLE activated on a device. [53], [28], [35]

Disadvantages[modifier | modifier le code]

Even though signal strength decreases when passing through objects and walls, the signal can still go through and potentially create false positive cases. [46] For example if two persons are close by each other, but with an obstacle separating them such as a thin wall. [46] False negatives can also happen in the case of two persons speaking face to face, but with their phones in their back pocket, in this example the BLE signals might not detect that the two persons are in contact. [54], [55] Some research also showed that the received signal strength does not decrease significantly enough even with a long separation between smartphone devices, which could create false positive, since COVID-19 get contracted by close contact (1-2 meters), therefore this needs to be taken into account and if possible adjusted when using BLE in digital contact tracing. [56] Moreover, BLE signal strength used to measure proximity, does not have the same reduction rate outdoors and indoors. [57], [43] Indoors it is more likely to be obstructed or disrupted, making it more challenging to estimate accurately the distance between devices. [57], [43], [46]

In iOS devices, BLE scanning cannot be active when the application using it is opened in the background and therefore it cannot detect or be detected, while in background state. [58], [35], [59] This isn't an issue on Android devices. [35], [58]

Specific attacks can target BLE-enabled systems. [60] The majority of these BLE attacks can be mitigated or countered by adding other features to the digital tracing protocol that uses BLE. [60] However, jamming attacks (see radio jamming) are impossible to get around. In this context jamming attacks consists of jamming the BLE signal spectrum range so that no BLE signal can be emitted. [60]

Wi-Fi[modifier | modifier le code]

Wi-Fi is one of the most prevalent technology used in wireless communications. [61] For this reason it can easily be considered for digital contact tracing. [28]

Advantages[modifier | modifier le code]

Wi-Fi can not only gather a complete list of surrounding devices quickly and work well inside buildings, but also give a reasonably precise location and distance between devices. [16], [21], [28], [23] Besides, Wi-Fi is not power-hungry, which makes it a good candidate to be used in contact tracing mobile applications. [16] Wi-Fi based digital contact tracing has also been found more accurate in areas where people are connected to the same Wi-Fi network for long periods of time (home, work, school etc.), which are the environments where people are most likely to contract the virus. [62]

Disadvantages[modifier | modifier le code]

One of the main problem is that users need to be connected on the same Wi-Fi network and Wi-Fi networks are not always available everywhere.

[16], [50], [50] Another specificity is that contact tracing requires to identify devices of people who were physically closed to one another. [63], [28] So, it assumes that one device belongs to one individual, however one user may have several devices connected to a same Wi-Fi network, this could announce several cases of COVID-19 infections instead of one. [4]

Additionally, Wi-Fi networks are mainly found inside, so it could not gather contacts efficiently in outside settings. [23], [50] Lastly, mobile devices connected to the same Wi-Fi network can be vulnerable to distributed denial-of-service attacks. [50]

Cellular network[modifier | modifier le code]

Advantages[modifier | modifier le code]

The main advantage of using cellular network for contact tracing is that it does not necessitate users to download an application and can be done using the already existing infrastructure, which makes it cheap and easy to implement. [22], [21], [16] Another point is that it does not significantly impact battery power, and since it is activated most of the time on smartphones anyways, it will not significantly make a difference in battery drainage. [16]

Disadvantages[modifier | modifier le code]

The precisions of locations retrieved with cellular network are not sufficient to track cover-19 infections: the signals can locate a phone with around 140 meters in cities and in rural environment the precision can be up to one kilometer. [16], [21] Using cellular network also raises huge privacy concerns with individuals who may not accept to be tracked, it is then necessary to have a way to opt-out. [16] And finally, cellular network might not be available everywhere, particularly in remote areas, and may not work sufficiently indoors. [16]

Architectures[modifier | modifier le code]

To build digital contact tracing solutions, not only tools are needed but also a general approach to deal with data, data processing and notifications for individuals who have been in contact with an infected person. Until 2021, there are three different types of approaches that have stood out: centralized, decentralized and hybrid [20]

Centralized[modifier | modifier le code]

In a centralized architecture, the contact tracing data gathered on mobile devices, be it BLE identifiers, geolocation data, or any other type, is uploaded to central servers that are managed by a central authority, generally a government or a health agency. [46], [64] On the servers, the data is processed, in order to determine who has been nearby an infected individual. [64], [28], [46], [28] Then a notification is sent to all the devices that have crossed path with an infected person. [46], [28], [64] Basically the determination of which contact is at risk and who needs to receive a notification is established on the central servers with the data transferred from each device. [65] The central servers handle the generation of encryption keys, the creation of unique identifiers for each device, the analysis of risky contacts and the alerts that need to be sent to the devices. [66] France's digital contact tracing application, TousAntiCovid and the Australian application COVIDSafe are both based on a centralized architecture. [67], [68]

Here is an example of a how a centralized architecture can do digital contact tracing based on a BLE mobile application: users register on the central server via the mobile application. The server generates ids. The ids are encrypted with a key, that only the central server knows and the encrypted ids are sent to the device at regular intervals (the ids need to change regularly for privacy reasons, so that it is not easy to trace one specific id to a device). Then the devices record the ids of other nearby devices and regularly upload them on the central servers. When someone declares a positive diagnosis on the application, the central servers are notified and they check which ids have been in contact with the ids that have been used by the device on which someone declared a positive test. Once the central servers gather all those ids, they can match them to the devices they belonged to since they have the encryption key, they will then send a notification all those devices that used the ids in the list st some point. [69]

Decentralized[modifier | modifier le code]

The main idea in the decentralized architecture is that data processing, determination of at risk contacts and notification is done on the users' devices instead of a central server and that no specific authority manages the data. [70], [31], [65] Unlike the centralized approach, in a decentralized system, the unique ids needed to link a device in case of a positive test, are generated on the device itself, either by the application or by the operating system. [31] The data processing, determining at risk contact and notifications, are also done on the devices themselves, so the devices behave as local servers. [28], [31], [65] This way the majority of the data stays on the devices and are not shared unless necessary, which means in case of a positive test. Indeed, in the case of a positive test, this system still needs central servers to update the ids that the COVID-19 positive user has used recently to a server. [28], [31], [65] The devices then regularly download the list of ids from the server and compare the list to the ids recently gathered by the device: on the device the data is processed and if it considers that the users is at risk of infection, it will create a notification. [28], [31], [51] Numerous digital contact tracing tools are based on a decentralized system, for example Irish contact tracing application: COVID tracker Ireland or Switzerland's application: SwissCovid. [71], [70], [51]

Here is an example of a how a decentralized architecture can do digital contact tracing based on a BLE mobile application: users download the application and the application regularly generates ids that are send through BLE to other nearby devices. The list of id gathered is stored on the device. At regular intervals, for example every 24 hours, the app downloads the list of ids from the server. Those downloaded ids belong to users who have tested positive. Then on the device, the application verifies if the ids gathered lately on the phone match any of the ones downloaded from a server. If there is any match, the application decides if the user is at risk and if they are, it will generate an alter on the phone. [72]

Other proposals have been made for fully decentralized systems using blockchain technology. [73], [74], [75] For example, IEEE members: Hao Xu, Lei Zhang, Oluwakayode Onireti, Yang Fang, William J. Buchanan and Muhammad Ali Imran, present a blockchain based contact tracing approach they named BeepTrace in their paper BeepTrace: Blockchain-Enabled Privacy-Preserving Contact Tracing for COVID-19 Pandemic and Beyond. [75] This contact tracing blockchain is based on a Directed acyclic graph consensus mechanism, which is the approach used to decide if a block can be added to the blockchain, because it allows branches to the blockchain, which means blocks are easily added even simultaneously and it does not consumes too much power. [76]

In their paper, the authors argue that using a blockchain system for digital contact tracing can solve the issues of users' privacy, facilitate the protection of the identities of COVID-19 positive individuals, and also be used internationally, without compatibility issues. [77] The fact that the blockchain is like a ledger that cannot be rewritten or misused easily, could also prevent attacks that the other systems are vulnerable to. [77] Yet to this day (January 2022) there are no blockchain digital contact tracing tools available to the public. [77] [78]

Hybrid[modifier | modifier le code]

In the hybrid approach, some tasks are done on central servers and others on the devices themselves. [79], [36], [80] For instance, the analysis of risky contacts (the matching of ids) and the sending of notifications can be done on central servers and the generation of ids and the management of those ids can be done on the devices, but there are other combinations possible. [13], [79], [36], [80]

Here is one example of how a hybrid architecture can operate for digital contact tracing, based on an smartphone application using BLE: devices using the application exchange their locally generated ids with nearby devices and store them locally. [80] When a user declares to have tested positive on the application, the list of encrypted ids gathered and stored on the phone is sent to the central server. [80] Any user who wants to know their risk can then also upload their data to the server. [80], [81] The server then will do the risk analysis and send a notification if needed. [80], [82]

Protocols[modifier | modifier le code]

There are several protocols that have been designed to build digital contact tracing tools. Some of those protocols are open source and currently used by a digital contact tracing tool available to the public. [83], [84]

Centralized protocols[modifier | modifier le code]

  • ROBERT protocol

ROBERT stands for ROBust and privacy-presERving proximity Tracing it is an open source protocol developed by the Inria Fraunhofer research labs. [83], [85], [86] It is based on a centralized model and uses BLE. [87], [88], [89], [86] It is mainly used in the French digital contact tracing app TousAntiCovid. [85], [86] The ROBERT protocol works like this: when a user first sign up on the application, a permanent user id is generated by the server as well as multiple Ephemeral Bluetooth Identifiers. [90] The server regenerates new Ephemeral Bluetooth Identifiers when needed. [91] The Ephemeral Bluetooth Identifiers are encrypted by the server (with its private key). [91] Once a user is registered on the application and bluetooth enabled on the device, it will start broadcasting BLE advertisement messages to devices around. This advertisement messages contain an Ephemeral Bluetooth Identifier and are stored on the device with the time of reception and the RSSI. [90] When a user tests positive for COVID-19 it uploads to the server the list of BLE advertisement messages received as well as the time they were received and the RSSI. [90] The application queries the server regularly to check if the user is at risk of exposure to COVID-19 or not. [90] When that happens the server verifies the number of times the Ephemeral Bluetooth Identifier have been uploaded by an infected user as well as other information linked to those messages and measure a risk score. [90] If the score is high enough to consider that the user is likely to have been infected the server sends a notification. [90]

Decentralized protocols[modifier | modifier le code]

Hybrid protocols[modifier | modifier le code]

  • ConTra Corona

ConTra Corona is a hybrid protocol developed in Germany by researchers from the FZI Research Center for Information Technology and Karlsruhe Institute of Technology. [85], [92] The protocol is based on BLE technology and uses three different separated servers: the submission server, the matching server, and the notification server. [92], [93] Here is how ConTra Corona works: when users first register they need to provide their name and phone number which will consist in their "real id". [94] Each day, the application generates a "warning id" that are derived from the user's "real id" by a commitment scheme, it also creates 96 "seed ids" for each "warning id". [94] The "seed ids" are encrypted with the submission server's public key. [94], [92] After that, the application sends the pair (encrypted seed id, pid), (pid is a pseudorandom identifier encrypted and hashed based on the seed id), to the submission server. [94], [92] The submission sever mix up all those pairs and upload them on the matching server, before deleting them. [94], [92] When a user tests positive for COVID-19, the medical staff generates a code and push this code to the matching server. Then the application of the infected user uploads on the matching server, the code generated by the medical staff and the pids that were received during the window of time when the user was infectious but not yet tested. [94] The matching server then matches the ids in the list with their seed id and sends the seed ids retrieved to the notification server [95] The notification server decrypts the seed ids into the corresponding warning id and make the list of decrypted warning ids public so that users can regularly download them and their application verifies if any of them have belonged to the device at some point, in which case a notification is created. [95], [96]

  • DESIRE

DESIRE is an evolution of the centralized ROBERT protocol. [97] In DESIRE, the ephemeral user ids are generated on the device not on the server and they are encrypted with keys stored on the mobile device. [98] The devices exchange Private Encounter Token, which are derived from the ephemeral user ids that the application is using at the specific time, when encountering nearby devices. [99] When a user tests positive for COVID-19 the encountered Private Encounter Token recorded own the device are uploaded to the server, the server then adds them to a global list of uploaded Private Encounter Tokens. [100] The app regularly sends to the server its used Private Encounter Tokens to check if they are found in the list of at risk contact (tokens uploaded by infected individual). [100] The server then calculates a risk status and sends a notification if the user is at risk. [100]

References[modifier | modifier le code]

  1. ↑ a et b Castelluccia 2020, p. 2
  2. ↑ a et b Tedeschi 2020, p. 1
  3. ↑ a et b Chowdhury 2020, p. 225704
  4. ↑ a b c et d Min-Allah 2021, p. 7
  5. ↑ a et b Jiang 2020, p. 22
  6. ↑ Kouliaridis 2021, p. 7
  7. ↑ Gupta 2020, p. 3
  8. ↑ Reichert 2021, p. 9
  9. ↑ a et b Singh 2020, p. 1
  10. ↑ Gupta 2020, p. 5
  11. ↑ a b et c Leslie 2020, p. 2
  12. ↑ a et b Gupta 2020, p. 4
  13. ↑ a b et c Bedogni 2021, p. 2
  14. ↑ Dar 2020, p. 1
  15. ↑ Shahroz 2021, p. 3
  16. ↑ a b c d e f g h i et j Jiang 2020, p. 23-24
  17. ↑ Gupta 2020, p. 8
  18. ↑ Messai 2020, p. 2
  19. ↑ Kindt 2020, p. 1
  20. ↑ a et b Ahmed 2020, p. 124586
  21. ↑ a b c d et e Hernández-Orallo 2020, p. 99086
  22. ↑ a et b Legendre 2020, p. 6
  23. ↑ a b c d et e Reichert 2021, p. 6
  24. ↑ a et b Beskorovajnov 2021, p. 27
  25. ↑ a et b Patel 2020, p. 1
  26. ↑ a b c et d Min-Allah 2021, p. 8
  27. ↑ a b c d et e Legendre 2020, p. 14
  28. ↑ a b c d e f g h i j k et l Shahroz 2021, p. 2
  29. ↑ Bay 2020, p. 6
  30. ↑ Madoery 2021, p. 3
  31. ↑ a b c d e et f Legendre 2020, p. 16
  32. ↑ Loh 2020, p. 6
  33. ↑ Chowdhury 2020, p. 225724
  34. ↑ a b et c Dar 2020, p. 3
  35. ↑ a b c d et e Reichert 2021, p. 5
  36. ↑ a b c d e f et g Tedeschi 2020, p. 7
  37. ↑ Messai 2020, p. 3
  38. ↑ Tedeschi 2020, p. 3
  39. ↑ a et b Aftab 2017, p. 34
  40. ↑ Kindt 2020, p. 2
  41. ↑ a b c d et e Reichert 2021, p. 4
  42. ↑ Kindt 2020, p. 5
  43. ↑ a b c et d Hatke 2020, p. 2
  44. ↑ a b et c Min-Allah 2021, p. 2
  45. ↑ a b c d e f et g Chowdhury 2020, p. 225705
  46. ↑ a b c d e f et g Jiang 2020, p. 12
  47. ↑ Messai 2020, p. 1
  48. ↑ a b et c Legendre 2020, p. 17-18
  49. ↑ Tedeschi 2020, p. 2
  50. ↑ a b c d e et f Min-Allah 2021, p. 4
  51. ↑ a b et c Bedogni 2021, p. 3
  52. ↑ a et b Chowdhury 2020, p. 225727
  53. ↑ Bay 2020, p. 5
  54. ↑ Shahroz 2021, p. 5
  55. ↑ Meklenburg 2020, p. 2
  56. ↑ Pandl 2021, p. 1
  57. ↑ a et b Madoery 2021, p. 2
  58. ↑ a et b Loh 2020, p. 5-6
  59. ↑ Meklenburg 2020, p. 13
  60. ↑ a b et c Reichert 2021, p. 22-24
  61. ↑ « Discover Wi-Fi », sur https://www.wi-fi.org/ (consulté le )
  62. ↑ Loh 2020, p. 7
  63. ↑ Vaudenay 2020, p. 1
  64. ↑ a b et c Tedeschi 2020, p. 5
  65. ↑ a b c et d Shubina 2020, p. 4
  66. ↑ Ahmed 2020, p. 134583
  67. ↑ Leslie 2020, p. 1
  68. ↑ Ahmed 2020, p. 134590
  69. ↑ Ahmed 2020, p. 134579
  70. ↑ a et b Jiang 2020, p. 20
  71. ↑ Dar 2020, p. 12
  72. ↑ Ahmed 2020, p. 134580-134581
  73. ↑ Vaudenay 2020, p. 14
  74. ↑ Zhang 2021
  75. ↑ a et b Xu 2021
  76. ↑ Xu 2021, p. 3918
  77. ↑ a b et c Xu 2021, p. 3917
  78. ↑ Hasan 2021, p. 62957
  79. ↑ a et b Ahmed 2020, p. 134581
  80. ↑ a b c d e et f Jiang 2020, p. 15
  81. ↑ Ahmed 2020, p. 134582
  82. ↑ Ahmed 2020, p. 134578
  83. ↑ a et b Messai 2020, p. 4
  84. ↑ Shahroz 2021, p. 6
  85. ↑ a b et c Jiang 2020, p. 11
  86. ↑ a b et c Ahmed 2020, p. 134591
  87. ↑ Dar 2020, p. 10
  88. ↑ Beskorovajnov 2021, p. 2
  89. ↑ Vaudenay 2020, p. 4
  90. ↑ a b c d e et f Castelluccia 2020, p. 5
  91. ↑ a et b Castelluccia 2020, p. 7
  92. ↑ a b c d et e Ahmed 2020, p. 134594
  93. ↑ Beskorovajnov 2021, p. 4
  94. ↑ a b c d e et f Beskorovajnov 2021, p. 13
  95. ↑ a et b Beskorovajnov 2021, p. 16
  96. ↑ Ahmed 2020, p. 134595
  97. ↑ Castelluccia 2020, p. 1
  98. ↑ Castelluccia 2020, p. 6
  99. ↑ Castelluccia 2020, p. 8
  100. ↑ a b et c Castelluccia 2020, p. 4

Bibliography[modifier | modifier le code]

  • (en) Castelluccia, Bielova, Boutet, Cunche et al., « ROBERT: ROBust and privacy-presERving proximity Tracing », hal-02611265,‎ (DOI white paper)
  • (en) Min-Allah, Alahmed, Albreek, Alghamdi et al., « A survey of COVID-19 contact-tracing apps », Computers in Biology and Medicine,‎ (DOI 10.1016/j.compbiomed.2021.104787)
  • (en) Tedeschi, Bakiras et Di Pietro, « SpreadMeNot: A Provably Secure and Privacy-Preserving Contact Tracing Protocol », arXiv preprint arXiv:2011.07306,‎
  • (en) Chowdhury, Ferdous, Chowdhury et Muthukkumarasamy, « COVID-19 Contact Tracing: Challenges and Future Directions », IEEE Access, vol. 8,‎ (DOI 10.1109/ACCESS.2020.3036718)
  • (en) Kouliaridis, Kambourakis, Chatzoglou, Geneiatakis et al., « Dissecting contact tracing apps in the Android platform », Plos one, vol. 15,‎ (DOI 10.1371/journal.pone.0251867)
  • (en) Gupta, Bedi, Goyal, Wadhera et al., « Analysis of COVID-19 Tracking Tool in India: Case Study of Aarogya Setu Mobile Application », Digital Government: Research and Practice, vol. 1,‎ (DOI 10.1145/3416088)
  • (en) Reichert, Brack et Scheuermann, « A Survey of Automatic Contact Tracing Approaches Using Bluetooth Low Energy », ACM Transactions on Computing for Healthcare, vol. 2,‎ (DOI https://doi.org/10.1145/3444847)
  • (en) Singh P., Singh A., Cojocaru, Vepakomma et al., « PPContactTracing: A Privacy-Preserving Contact Tracing Protocol for COVID-19 Pandemic », arXiv preprint arXiv:2008.06648,‎ (DOI white paper)
  • (en) Leslie, « COVID-19 Fight Enlists Digital Technology: Contact Tracing Apps », Engineering (Beijing), vol. 6,‎ (DOI 10.1016/j.eng.2020.09.001)
  • (en) Dar, Lone, Zahoor, Khan et al., « Applicability of mobile contact tracing in fighting pandemic (COVID-19): Issues, challenges and solutions », Computer Science Review, vol. 38,‎ (DOI 10.1016/j.cosrev.2020.100307)
  • (en) Shahroz, Ahmad, Younis, Ahmad et al., « COVID-19 digital contact tracing applications and techniques: A review post initial deployments », Transportation Engineering,‎ (DOI 10.1016/j.treng.2021.100072)
  • (en) Jiang, Zhang, Zhang, Yu et al., « A Survey on Contact Tracing: the Latest Advancements and Challenges », Computer Science, Business, ArXiv,‎ (DOI white paper)
  • (en) Bedogni, Rumi et Salim, « Modelling Memory for Individual Re-identification in Decentralised Mobile Contact Tracing Applications », Association for Computing Machinery, vol. 5,‎ (DOI 10.1145/3448088)
  • (en) Messai et Seba, « Short Paper: Privacy Comparison of Contact Tracing Mobile Applications for COVID-19 », arXiv:2010.03232v1,‎ (DOI white paper)
  • (en) Kindt, Chakraborty T. et Chakraborty S., « How Reliable is Smartphone-based Electronic Contact Tracing for COVID-19? », Communications of the ACM, vol. 65,‎ (DOI 10.1145/3471933)
  • (en) Ahmed, Michelin, Xue, Ruj et al., « A Survey of COVID-19 Contact Tracing Apps », IEEE Access,‎ (DOI 10.1109/ACCESS.2020.3010226)
  • (en) Legendre, Humbert, Mermoud et Lenders, « Contact tracing: An overview of technologies and cyber risks », arXiv preprint arXiv:2007.02806,‎ (DOI white paper)
  • (en) Beskorovajnov, Dörre, Hartung, Koch et al., « ConTra Corona: Contact Tracing against the Coronavirus by Bridging the Centralized–Decentralized Divide for Stronger Privacy », Springer-Verlag,‎ (DOI 10.1007/978-3-030-92075-3_23)
  • (en) Patel, Jain, Menon et Kodeboyina, « Comparative Study of Privacy Preserving-Contact Tracing on Digital Platforms », 2020 International Conference on Computational Intelligence (ICCI),‎ (DOI 10.1109/ICCI51257.2020.9247782)
  • (en) Bay, Kek, Tan et Hau, « BlueTrace: A privacy-preserving protocol for community-driven contact tracing across borders », Government Technology Agency-Singapore, Tech. Rep,‎ (DOI white paper)
  • (en) Madoery, Detke, Comerci et al., « Feature selection for proximity estimation in COVID-19 contact tracing apps based on Bluetooth Low Energy (BLE) », Pervasive and Mobile Computing,‎ (DOI https://doi.org/10.1016/j.pmcj.2021.101474)
  • (en) Loh, « Flipping the perspective in contact tracing », arXiv preprint arXiv:2010.03806,‎ (DOI white paper)
  • (en) Hernández-Orallo, Manzoni, Calafate et al., « Evaluating How Smartphone Contact Tracing Technology Can Reduce the Spread of Infectious Diseases: The Case of COVID-19 », IEEE Access,‎ (DOI 10.1109/ACCESS.2020.2998042)
  • (en) Hatke, Montanari, Wentz et al., « Using Bluetooth Low Energy (BLE) Signal Strength Estimation to Facilitate ContactTracing forCOVID-19 », arXiv preprint arXiv:2006.15711,‎ (DOI white paper)
  • (en) Muhammad Aftab, Building Bluetooth Low Energy Systems, Packt Publishing, (ISBN 978-1-78646-108-7)
  • (en) Meklenburg, Specter, Wentz, Balakrishnan et al., « SonicPACT: An Ultrasonic Ranging Method for the Private Automated Contact Tracing (PACT) Protocol », arXiv preprint arXiv:2012.04770,‎
  • (en) Vaudenay, « Centralized or Decentralized? The Contact Tracing Dilemma », IACR Cryptology ePrint Archive,‎ (DOI white paper)
  • (en) Pandl, Thiebes, Schmidt-Kraepelin et Sunyaev, « How detection ranges and usage stops impact digital contact tracing effectiveness for COVID-19 », Scientific Reports, vol. 11,‎ (DOI 10.1038/s41598-021-88768-6)
  • (en) Shubina, Ometov et Simona Lohan, « Technical Perspectives of Contact-Tracing Applications on Wearables for COVID-19 Control », 12th International Congress on Ultra Modern Telecommunications and Control Systems and Workshops (ICUMT),‎ (DOI 10.1109/ICUMT51630.2020.922224)
  • (en) Zhang, Xu, Sharif et Zhu, « Privacy-preserving contact tracing in 5G-integrated and blockchain-based medical applications », Computer Standards & Interfaces,‎ (DOI 10.1016/j.csi.2021.103520)
  • (en) Xu, Zhang, Onireti, Fang et al., « Beeptrace: Blockchain-enabled privacy-preserving contact tracing for covid-19 pandemic and beyond », IEEE Internet of Things Journal, vol. 8,‎ (DOI 10.1109/JIOT.2020.3025953)
  • (en) Castelluccia, Bielova, Boutet, Cunche et al., « DESIRE: A Third Way for a European Exposure Notification System Leveraging the best of centralized and decentralized systems », arXiv preprint arXiv:2008.01621,‎ (DOI white paper)
  • (en) Hasan, Salah, Jayaraman, Yaqoob et al., « COVID-19 Contact Tracing Using Blockchain », IEEE Access, vol. 9,‎ (DOI 10.1109/ACCESS.2021.3074753)
Ce document provient de « https://fr.wikipedia.org/w/index.php?title=Utilisateur:Karolinasea/Brouillon&oldid=189770898 ».