Magnet: Real-Time Trace Stream Analytics Framework for 5G Operations Support Systems

The era of petabyte data has arrived as the digital big data universe continues its expansion toward exascale with massive volumes of data generated by diverse distributed sources. The size of big data makes it very difficult to gain insight into the meaning of data. In industrial applications, in order to explore both the meaning of data and the complex relationship between data components, big data needs to be processed and reduced enabling further deeper analysis in a timely manner. In this article an integrated data analytics framework is presented designed to extract the set of instances exhibiting statistical dependency from the massive volume of data in a pre-defined quasi real-time manner. The parallel computing model of MapReduce is enhanced to realize Magnet. The solution presented in this article is applicable to the telecommunications market where it optimizes next-generation network management systems for heterogeneous radio access technologies.

[1]  Albert Mo Kim Cheng,et al.  An auto-scaling mechanism for virtual resources to support mobile, pervasive, real-time healthcare applications in cloud computing , 2013, IEEE Network.

[2]  Jennifer Widom,et al.  Models and issues in data stream systems , 2002, PODS.

[3]  W. B. Johnson,et al.  Extensions of Lipschitz mappings into Hilbert space , 1984 .

[4]  Liam Fallon,et al.  Load balanced telecommunication event consumption using pools , 2015, 2015 IFIP/IEEE International Symposium on Integrated Network Management (IM).

[5]  Gabriel-Miro Muntean,et al.  A recommender system architecture for predictive telecom network management , 2015, IEEE Communications Magazine.

[6]  Andrey Brito,et al.  Scalable and Low-Latency Data Processing with Stream MapReduce , 2011, 2011 IEEE Third International Conference on Cloud Computing Technology and Science.

[7]  Jimmy O'Meara,et al.  A system for monitoring mobile networks using performance management events , 2013, 2013 IFIP/IEEE International Symposium on Integrated Network Management (IM 2013).

[8]  Sanjay Ghemawat,et al.  MapReduce: Simplified Data Processing on Large Clusters , 2004, OSDI.

[9]  Gabriel-Miro Muntean,et al.  A heuristic correlation algorithm for data reduction through noise detection in stream-based communication management systems , 2014, 2014 IEEE Network Operations and Management Symposium (NOMS).

[10]  Declan O'Sullivan,et al.  SECCO: A test framework for controlling and monitoring end user service sessions , 2014, 2014 IEEE Network Operations and Management Symposium (NOMS).

[11]  Gabriel-Miro Muntean,et al.  E-stream: Towards pattern centric network incident discovery and corrective action recommendation in telecommunication networks , 2015, 2015 IFIP/IEEE International Symposium on Integrated Network Management (IM).