A Study on Modeling of MIMO Channel by Using Different Neural Network Structures
Hadi Alipour1, Mohammad Reza Noorbakhsh2, Zahra Mansourian3

1Hadi Alipour, Payame Noor University, Tehran, Iran.
2Mohammad Reza Noorbakhsh, Department of Education, Shiraz, Iran.
3Zahra Mansoorian, free university, Yazd Meybod, Iran.

Manuscript received on 18 November 2012 | Revised Manuscript received on 25 November 2012 | Manuscript published on 30 November 2012 | PP: 47-50 | Volume-1 Issue-5, November 2012 | Retrieval Number: E0347101512/2012©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Recognition of Radio Channel (channel Parameters) is one of Main Challenges in Signal Transformation, and has important role in cognitive radio approach. Goal of this paper is “Channel modeling” to estimate coefficients of transmission functions affected on data being transformed in the channel. We use Multilayer perceptron(MLP) Neural Network with Back-propagation learning algorithm, block-structured Neural Network with Least Squares(LS) method(cost function) and a multilayer neural network with multiple back-propagation(MBP) learning algorithm for error estimation. These networks will be trained with received signals to be compatible with channel, then give us an estimation of these coefficients. Simulation will show that this MBP method is better than the other two method in error estimation. It has good performance and also consume less execution time. Then, we will use this network for estimating coefficients of non-linear transmission functions of actual radio channel.
Keywords: Cognitive Radio, Channel Recognition, Channel Modeling, Least Squares, Multiple Back-propagation(MBP), Neural Network, Transmission function.

Scope of the Article: Neural Information Processing