Managing Tropospheric Ducting Effect in Mobile Networks Using Unsupervised Machine Learning

Self-optimized networks continue to gain more immense importance by mobile network manufacturers and operators. Obtaining optimized operation with minimum human intervention contributes to better customer satisfaction and enhanced cost efficiency. Deep learning is commonly applied to achieve self-optimization by intelligently interpret patterns of network-statistics and develop action plans that achieve optimum performance. In this paper, we apply deep neural networks to develop a novel machine learning system that performs automatic coverage-tuning to optimize performance under tropospheric ducting condition. In a live-trial, our algorithm was set to monitor 180 site that include clusters of different configuration (Sub-urban, rural and roads). Our system improves the spectrum efficiency under the external effects like weather and our trial show enhancement of 40\% in call drops as per a detected cell's data.

[1]  Erhardt Barth,et al.  A Hybrid Convolutional Variational Autoencoder for Text Generation , 2017, EMNLP.

[2]  Heng Tao Shen,et al.  Principal Component Analysis , 2009, Encyclopedia of Biometrics.

[3]  I. Sirkova,et al.  Influence of Tropospheric Ducting on Microwave Propagation in Short Distances , 2003 .

[4]  Ayman El-Baz,et al.  Athlete-Customized Injury Prediction using Training Load Statistical Records and Machine Learning , 2018, 2018 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT).

[5]  Olav Tirkkonen,et al.  Channel Charting: Locating Users Within the Radio Environment Using Channel State Information , 2018, IEEE Access.

[6]  Irina Sirkova,et al.  Parabolic‐equation‐based study of ducting effects on microwave propagation , 2004 .

[7]  I. Sirkova,et al.  Parabolic Wave Equation Method Applied to the Tropospheric Ducting Propagation Problem: A Survey , 2006 .

[8]  Michael E. Tipping,et al.  Probabilistic Principal Component Analysis , 1999 .

[9]  Vincent K. N. Lau,et al.  The Mobile Radio Propagation Channel , 2007 .

[10]  Abraham U. Usman,et al.  Instantaneous GSM Signal Strength Variation with Weather and Environmental Factors , 2015 .

[11]  Ahmed Naglah,et al.  Real-time scale-adaptive compressive tracking using two classification stages , 2015, 2015 IEEE International Conference on Robotics and Biomimetics (ROBIO).

[12]  Kurt Hornik,et al.  Neural networks and principal component analysis: Learning from examples without local minima , 1989, Neural Networks.

[13]  Carl Doersch,et al.  Tutorial on Variational Autoencoders , 2016, ArXiv.

[14]  Regina Barzilay,et al.  Junction Tree Variational Autoencoder for Molecular Graph Generation , 2018, ICML.

[15]  Max Welling,et al.  Auto-Encoding Variational Bayes , 2013, ICLR.