Creation of prototype for automatic fault detection in ISP network (Q3815732): Difference between revisions
Jump to navigation
Jump to search
(Changed an Item: import item from Lithuania) |
(Changed an Item: Adding English translations) |
||||||||||||||
| Property / summary | |||||||||||||||
The aim of the project is to create an artificial intelligence-based solution that can identify problems (attacks, hardware failures, software failures, lack of resources) and learn to predict their occurrence even before negative consequences occur by receiving Internet network surveillance data from all devices. A solution is being developed for network monitoring data collected from switches, routers, which are somewhat simpler, but their total volume is significantly higher. On the basis of these data, trained machine-learning algorithms would be able to quickly identify whether there is any abnormality in the network. Also identify in which particular device the anomaly is monitored and, if possible, specify a specific port or path. Large-scale monitoring in this way would make it possible to resolve faults and solve network problems more quickly and to predict them. (English) | |||||||||||||||
| Property / summary: The aim of the project is to create an artificial intelligence-based solution that can identify problems (attacks, hardware failures, software failures, lack of resources) and learn to predict their occurrence even before negative consequences occur by receiving Internet network surveillance data from all devices. A solution is being developed for network monitoring data collected from switches, routers, which are somewhat simpler, but their total volume is significantly higher. On the basis of these data, trained machine-learning algorithms would be able to quickly identify whether there is any abnormality in the network. Also identify in which particular device the anomaly is monitored and, if possible, specify a specific port or path. Large-scale monitoring in this way would make it possible to resolve faults and solve network problems more quickly and to predict them. (English) / rank | |||||||||||||||
Normal rank | |||||||||||||||
| Property / summary: The aim of the project is to create an artificial intelligence-based solution that can identify problems (attacks, hardware failures, software failures, lack of resources) and learn to predict their occurrence even before negative consequences occur by receiving Internet network surveillance data from all devices. A solution is being developed for network monitoring data collected from switches, routers, which are somewhat simpler, but their total volume is significantly higher. On the basis of these data, trained machine-learning algorithms would be able to quickly identify whether there is any abnormality in the network. Also identify in which particular device the anomaly is monitored and, if possible, specify a specific port or path. Large-scale monitoring in this way would make it possible to resolve faults and solve network problems more quickly and to predict them. (English) / qualifier | |||||||||||||||
point in time: 1 February 2022
| |||||||||||||||
Revision as of 17:24, 1 February 2022
Project Q3815732 in Lithuania
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | Creation of prototype for automatic fault detection in ISP network |
Project Q3815732 in Lithuania |
Statements
21,997.56 Euro
0 references
24,441.73 Euro
0 references
90.0 percent
0 references
1 January 2022
0 references
30 December 2022
0 references
"Tinklainė", MB
0 references
85160
0 references
Projekto tikslas yra sukurti dirbtiniu intelektu pagrįstą sprendimą, kuris gaudamas interneto tinklo stebėjimo duomenis iš visų įrenginių galėtų laiku identifikuoti problemas (atakas, techninės įrangos gedimus, programinės įrangos gedimus, resursų trūkumą) ir išmokti prognozuoti jų atsiradimą dar iki pasireiškiant neigiamoms pasekmėms. Kuriamas sprendimas tinklo stebėjimui naudotų duomenis surinktus iš komutatorių, maršrutizatorių, kurie yra kiek paprastesni, tačiau jų bendras kiekis yra ženkliai didesnis. Remiantis šiais duomenimis apmokyti mašininio mokymo algoritmai gebėtų greitai nustatyti ar tinkle vyksta kokia nors anomalija. Taip pat identifikuotų, kuriame konkrečiai įrenginyje ta anomalija yra stebima ir pagal galimybę nurodytų konkretų prievadą ar kelią. Tokiu būdu atliekama plataus masto stebėsena leistų greičiau pašalinti gedimus ir spręsti tinklo problemas, bei jas prognozuoti. (Lithuanian)
0 references
The aim of the project is to create an artificial intelligence-based solution that can identify problems (attacks, hardware failures, software failures, lack of resources) and learn to predict their occurrence even before negative consequences occur by receiving Internet network surveillance data from all devices. A solution is being developed for network monitoring data collected from switches, routers, which are somewhat simpler, but their total volume is significantly higher. On the basis of these data, trained machine-learning algorithms would be able to quickly identify whether there is any abnormality in the network. Also identify in which particular device the anomaly is monitored and, if possible, specify a specific port or path. Large-scale monitoring in this way would make it possible to resolve faults and solve network problems more quickly and to predict them. (English)
1 February 2022
0 references
Ramučių g. 38, Naujoji Akmenė
0 references
Identifiers
01.2.1-MITA-T-852-04-0015
0 references