Using Echo State Networks to Characterise Wireless Channels

We propose the use of echo state networks for the task of wireless channel characterisation to select the most similar channel to the current observed channel from a pre-defined set, based entirely on received signal information. This allows the system to select the optimal resource allocation scheme and transmission parameters from pre-computed solutions. Using suitable training data, the neural network was able to learn to characterise a signal correctly 68% of the time, which can be further improved to 72% by adding some simple location data to the signals being examined. Our system out- performs a comparable statistical method by a factor of two, demonstrating echo state networks' ability to infer information from their training data which other systems can not.

[1]  J. Salo,et al.  An interim channel model for beyond-3G systems: extending the 3GPP spatial channel model (SCM) , 2005, 2005 IEEE 61st Vehicular Technology Conference.

[2]  Xiongwen Zhao,et al.  WINNER II Channel Models Part I Channel Models , 2022 .

[3]  R. Ganesh,et al.  Statistical characterization of a partitioned indoor radio channel , 1992, [Conference Record] SUPERCOMM/ICC '92 Discovering a New World of Communications.

[4]  S. Sinanovic,et al.  Hardware implementation of a Kullback-Leibler Divergence based signal anomaly detector , 2009, 2009 2nd International Symposium on Applied Sciences in Biomedical and Communication Technologies.

[5]  S. Sinanovic,et al.  Anomaly detection using the Kullback-Leibler divergence metric , 2008, 2008 First International Symposium on Applied Sciences on Biomedical and Communication Technologies.

[6]  J. Salo,et al.  Practical Introduction to LTE Radio Planning , 2010 .

[7]  Reiner S. Thomä,et al.  Comparison of SCM, SCME, and WINNER Channel Models , 2007, 2007 IEEE 65th Vehicular Technology Conference - VTC2007-Spring.

[8]  Herbert Jaeger,et al.  The''echo state''approach to analysing and training recurrent neural networks , 2001 .

[9]  Harald Haas,et al.  Harnessing Nonlinearity: Predicting Chaotic Systems and Saving Energy in Wireless Communication , 2004, Science.

[10]  Ronald G. Harley,et al.  An Introduction to the Echo State Network and its Applications in Power System , 2009, 2009 15th International Conference on Intelligent System Applications to Power Systems.

[11]  K. Pahlavan,et al.  Statistical characterisation of a partitioned indoor radio channel , 1992 .

[12]  Moe Z. Win,et al.  Characterization of ultra-wide bandwidth wireless indoor channels: a communication-theoretic view , 2002, IEEE J. Sel. Areas Commun..