[Skip Header and Navigation] [Jump to Main Content]
Home

Secondary Links

  • Publications
  • Centers
  • Media
  • Important Forms
  • Contact Us

Languages

  • العربية
  • English

Primary Links

  • Home
  • About
  • Programs
  • Faculty Achivements
  • Photo Gallery
  • Quality Unit
  • Students
Home

Trained Neural Network on selecting DL speed for eNodeB

Academic Year: 
2013
Students: 
Dyala Fatayer
Oswa Ayoub
Tamarah Al-Joudeh
Department: 
Electrical Engineering

 

 

This project presents a new concept in the communication world, by linking LTE with Self-Organizing Networks (SON). The main idea is to use a Feed-Forward Back Propagation Network. In SON, we are concerned about training the network to select the required actual DL speed Mbps (BW) according to the code per service type. This can be accomplished by giving all possible service types a code number, which is used as an input matrix. On the other hand, we give each service type the bandwidth it requires in LTE system to serve with a good quality and use it as a target matrix, both matrices (input, target) are used as an input to the training process in the Feed-Forward Back Propagation Network. At the end of training the eNodeB will be able to choose the appropriate bandwidth for each service type automatically, leading to better utilization of system.

 

 

©2012 An-Najah National University|Faculty Of Engineering | P.O. Box: 7 | Nablus, Palestine | Phone: +970 (9) 2345113 Ext:2253 | Fax: +970 (9) 2345982 | email: [email protected]
[Jump to Top] [Jump to Main Content]