AI Anomaly Detection for Cloudified Mobile Core Architectures

Michelinakis, Foivos; Pujol-Roig, Joan S.; Malacarne, Sara; Xie, Min; Dreibholz, Thomas; Majumdar, Sayantini; Poe, Wint Yi; Patounas, Georgios; Guerrero, Carmen; Elmokashfi, Ahmed; Theodorou, Vasileios

doi:10.1109/tnsm.2022.3203246

Cited by 1 publication

(10 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In Table 8, we compare our proposed framework against two recent works from state-of-the-art [7], [4]. We choose these methods because they show some similarities to our work.…”

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

“…We choose these methods because they show some similarities to our work. The work presented in [7] leverages DAE and Convolutional Autoencoder (CAE) for bottleneck detection in a cloudified mobile core testbed. The authors did only active and passive monitoring and did not collect measurements from UEs.…”

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

“…In our work, with UE-based measurements and VAE, we can detect bottlenecks both at RAN and at core. The accuracy is slightly better than that of DAE (see Table 4), that is a preferred detection model of [7]. Other than using DAE, [7] examines only four types of high severity bottlenecks at the core network and does not consider bottlenecks occurring at RAN.…”

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

“…At this stage (i.e., at stage-1), we use Variational Autoencoder and achieve an impressive performance of 85% F1-score in detecting bottlenecks. This accuracy is not only 2% higher but also includes detection of bottlenecks at RAN that are missed by a recent related work [7]. 2) Presenting a viable solution to the problem of using supervised ML for classifying problems with potentially, yet, unlearned classes in the context of mobile networks.…”

Section: Introductionmentioning

confidence: 95%

“…Thirdly, to overload the network resources, we stress out the CPU and memory resources and increase input/output disk operations with the stress-ng tool [45] 6 . Lastly, to create radio interference we deploy a GNU RADIO 7 noise source on a separate system using a dedicated SDR ETTUS USRP B210. The noise source generates an additive white Gaussian noise (AWGN) signal which has central frequency similar to that of the eNodeB radio carrier.…”

Section: Types Of Bottlenecksmentioning

confidence: 99%

See 4 more Smart Citations

Bottleneck Identification in Cloudified Mobile Networks Based on Distributed Telemetry

Fida,

Ahmed,

Dreibholz

et al. 2024

IEEE Trans. on Mobile Comput.

Self Cite

View full text Add to dashboard Cite

Cloudified mobile networks are expected to deliver a multitude of services with reduced capital and operating expenses. A characteristic example is 5G networks serving several slices in parallel. Such mobile networks, therefore, need to ensure that the SLAs of customised end-to-end sliced services are met. This requires monitoring the resource usage and characteristics of data flows at the virtualised network core, as well as tracking the performance of the radio interfaces and UEs. A centralised monitoring architecture can not scale to support millions of UEs though. This paper, proposes a 2-stage distributed telemetry framework in which UEs act as early warning sensors. After UEs flag an anomaly, a ML model is activated, at network controller, to attribute the cause of the anomaly. The framework achieves 85% F1-score in detecting anomalies caused by different bottlenecks, and an overall 89% F1-score in attributing these bottlenecks. This accuracy of our distributed framework is similar to that of a centralised monitoring system, but with no overhead of transmitting UE-based telemetry data to the centralised controller. The study also finds that passive in-band network telemetry has the potential to replace active monitoring and can further reduce the overhead of a network monitoring system.

show abstract

“…In Table 8, we compare our proposed framework against two recent works from state-of-the-art [7], [4]. We choose these methods because they show some similarities to our work.…”

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

Section: Comparison With State-of-the-art Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 95%

Section: Types Of Bottlenecksmentioning

confidence: 99%

See 3 more Smart Citations

Bottleneck Identification in Cloudified Mobile Networks Based on Distributed Telemetry

Fida,

Ahmed,

Dreibholz

et al. 2024

IEEE Trans. on Mobile Comput.

Self Cite

View full text Add to dashboard Cite

show abstract

AI Anomaly Detection for Cloudified Mobile Core Architectures

Cited by 1 publication

References 46 publications

Bottleneck Identification in Cloudified Mobile Networks Based on Distributed Telemetry

Bottleneck Identification in Cloudified Mobile Networks Based on Distributed Telemetry

Contact Info

Product

Resources

About