| 1. |
- Imtiaz, Shamoona, et al.
(författare)
-
AI-based automatic detection of IP network performance in telecommunication
- 2024
-
Ingår i: Industry 4.0 Key Technological Advances and Design Principles in Engineering, Education, Business, and Social Applications. - : CRC Press. - 9781040048986 - 9781032376875 ; , s. 189-220
-
Bokkapitel (övrigt vetenskapligt/konstnärligt)abstract
- The convergence of complex networks, IoT, various services, the enormous amount of data over the network, and 5G have brought challenges to the telecom industry for enhanced service delivery and network maintenance and monitoring. With numerous Netrounds probes and test agents generating massive volumes of data, the knowledge produced is underutilized due to the manual search for insight within the data. With the help of Netrounds APIs and machine learning, the automation of Netrounds metrics data aimed to predict the network performance degradation and anomaly detection ahead of time. The automation was expected to provide meaningful data insight and minimize the violation of SLAs in terms of delay and packet loss. These are one of the primary objectives of business continuity management to handle threats and risks for various network and information systems. Automating Netrounds’ open and programable APIs through Python fed data to an automatic machine learning model (supervised learning for prediction and unsupervised learning for anomaly detection). The analytics were used to predict network behavior, anomaly detection, and maintenance of the SLA threshold.
|
|
| 2. |
- Rius i Riu, Jaume, 1973-
(författare)
-
Gas phase molecular relaxation probed by synchrotron radiation experiments
- 2002
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- This thesis presents experimental studies of gas phasemolecular relaxation after excitation with synchrotron photonsin the 15-35 eV and in the 70-350 eV regions.In the 15-35 eV region, molecular relaxation by neutraldissociation processes and nonFranck-Condon effects in N2 and O2 molecules have beenstudied by means of dispersed fluorescence and photoelectronspectroscopy experimental techniques, respectively. From thedispersed fluorescence data, excitation functions for themeasured atomic fluorescence spectra have been obtained. Fromthe recorded photoelectron spectra vibrational branching ratioshave been produced. The results obtained reveal that Rydbergseries and singly and doubly excited valence states of theappropriate symmetry energetically accessible in the studiedregion and interactions between themaccount for most of theobserved effects in these two type of experiments.In the 70-350 eV range, molecular relaxation processesresulting in fragmentation of CD4 and SF6 after absorption ofsynchrotron light have been studied by energy resolved electronion coincidence technique using a multicoincidence experimentalstation developed by our group during the last five years forsuch type of experiments. The coincidence measurements yieldedmass spectra from which information about the kinematics of thedetected fragments has been deduced by means of Monte Carlosimulations of the experimental peak shapes. The obtainedresults show completely different dissociation patternsdepending on the molecular electronic states studied. Thesepatterns reflect the bonding properties of the excited orbitalsand they permit the description and in some cases theidentification of the different molecular relaxation pathwaysobserved. The achievements presented in this thesis exemplifythe potential of the multicoincidence station used in thereported experiments.
|
|
