RT Dissertation/Thesis
T1 Performance Evaluation And Anomaly detection in Mobile BroadBand Across Europe
A1 Moulay Brahim, Mohamed Lamine Touhami
A2 IMDEA Networks Institute, 
AB With the rapidly growing market for smartphones and user’s confidence for immediateaccess to high-quality multimedia content, the delivery of video over wireless networks hasbecome a big challenge. It makes it challenging to accommodate end-users with flawlessquality of service. The growth of the smartphone market goes hand in hand with thedevelopment of the Internet, in which current transport protocols are being re-evaluated todeal with traffic growth. QUIC and WebRTC are new and evolving standards. The latteris a unique and evolving standard explicitly developed to meet this demand and enablea high-quality experience for mobile users of real-time communication services. QUIChas been designed to reduce Web latency, integrate security features, and allow a highqualityexperience for mobile users. Thus, the need to evaluate the performance of theserising protocols in a non-systematic environment is essential to understand the behaviorof the network and provide the end user with a better multimedia delivery service. Sincemost of the work in the research community is conducted in a controlled environment, weleverage the MONROE platform to investigate the performance of QUIC and WebRTCin real cellular networks using static and mobile nodes. During this Thesis, we conductmeasurements ofWebRTC and QUIC while making their data-sets public to the interestedexperimenter. Building such data-sets is very welcomed with the research community,opening doors to applying data science to network data-sets. The development part of theexperiments involves building Docker containers that act as QUIC and WebRTC clients.These containers are publicly available to be used candidly or within the MONROEplatform. These key contributions span from Chapter 4 to Chapter 5 presented in PartII of the Thesis.We exploit data collection from MONROE to apply data science over networkdata-sets, which will help identify networking problems shifting the Thesis focus fromperformance evaluation to a data science problem.Indeed, the second part of the Thesis focuses on interpretable data science. Identifyingnetwork problems leveraging Machine Learning (ML) has gained much visibility in thepast few years, resulting in dramatically improved cellular network services. However,critical tasks like troubleshooting cellular networks are still performed manually by expertswho monitor the network around the clock. In this context, this Thesis contributes by proposing the use of simple interpretableML algorithms, moving away from the current trend of high-accuracy ML algorithms(e.g., deep learning) that do not allow interpretation (and hence understanding) of theiroutcome. We prefer having lower accuracy since we consider it interesting (anomalous)the scenarios misclassified by the ML algorithms, and we do not want to miss them byoverfitting. To this aim, we present CIAN (from Causality Inference of Anomalies inNetworks), a practical and interpretable ML methodology, which we implement in theform of a software tool named TTrees (from Troubleshooting Trees) and compare it toa supervised counterpart, named STress (from Supervised Trees). Both methodologiesrequire small volumes of data and are quick at training. Our experiments using realdata from operational commercial mobile networks e.g., sampled with MONROE probes,show that STrees and CIAN can automatically identify and accurately classify networkanomalies—e.g., cases for which a low network performance is not justified by operationalconditions—training with just a few hundreds of data samples, hence enabling precisetroubleshooting actions. Most importantly, our experiments show that a fully automatedunsupervised approach is viable and efficient. In Part III of the Thesis which includesChapter 6 and 7.In conclusion, in this Thesis, we go through a data-driven networking roller coaster,from performance evaluating upcoming network protocols in real mobile networks tobuilding methodologies that help identify and classify the root cause of networkingproblems, emphasizing the fact that these methodologies are easy to implement and canbe deployed in production environments.
YR 2022
FD 2022-06
LK https://hdl.handle.net/10016/35952
UL https://hdl.handle.net/10016/35952
LA eng
NO This work has been supported by IMDEA Networks Institute
DS e-Archivo
RD 27 jul. 2024