Artificial Intelligence for Scheduling Resource Blocks in
LTE/5G Networks
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis
Department of Computer Science
Universidade de Brasilia
July 26, 2018
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 1 / 7
Outline
1 Introduction
2 Motivation
3 Current Work State
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 2 / 7
Intro.
Brazil EU (Finland, Spain) Collaboration
Figure: 5G RANGE Project
5G Range Project
50 km range
100 mb/s end to end
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 3 / 7
Motivation
Problem
Such high demands require better resource allocation, meaning usual
solutions are unlikely to respond well enough.
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 4 / 7
Motivation
Problem
Such high demands require better resource allocation, meaning usual
solutions are unlikely to respond well enough.
Solution
Artificial Neural Networks are quite high, performance wise, in most areas
of technology at the moment due to it‘s great “learning” capabilities by
optimizing a function that tells how far from a good result the prediction
is.
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 4 / 7
Motivation
Problem
Such high demands require better resource allocation, meaning usual
solutions are unlikely to respond well enough.
Solution
Artificial Neural Networks are quite high, performance wise, in most areas
of technology at the moment due to it‘s great “learning” capabilities by
optimizing a function that tells how far from a good result the prediction
is.
New Problem
Rural Traffic data sets are not commonly found, meaning that a data set
will have to be generated.
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 4 / 7
Current Work State
A South African study has characterized Rural Traffic in some areas
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 5 / 7
Current Work State
A South African study has characterized Rural Traffic in some areas
68.45% web traffic
Second biggest area is videos, including Youtube, Livestreams and
VoIP Services (Skype, Hangouts etc)
A few percentages are for public alert, calls and SMS.
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 5 / 7
Current Work State
A South African study has characterized Rural Traffic in some areas
68.45% web traffic
Second biggest area is videos, including Youtube, Livestreams and
VoIP Services (Skype, Hangouts etc)
A few percentages are for public alert, calls and SMS.
Reinforcement Learning based Neural Network is the choice for the model.
Current reward system accounts for Quality of Service, Spectral Efficiency,
Justice and Guaranteed Bit Rate. All of the rewards are normalized
between 0 and 1, whereas 1 is a perfect value.
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 5 / 7
For Further Reading I
Okvist, Simonsson and Asplund. LTE Frequency Selective Scheduling
Performance and Improvements Assessed by Measurements
International Symposium on Personal, Indoor and Mobile Radio
Communications, 2011.
Yang, Xu, Han, Rehman and Tao. GA Based Optimal Resource
Allocation for Device to Device WCNC 2014 - Workshop on D2D and
Public Safety Communications, 2014.
3GPP, Policy and Charging Control Architecture TS 23.203, 2018.
Arulkumaran, Deisenroth, Brundage and Bharath. A Brief Survey of
Deep Reinforcement Learning. IEEE Signal Processing Magazine
Special Issue on Deep Learning for Image Understanding, 2017.
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 6 / 7
For Further Reading II
Johnson, Pejovic, Belding and Stam. Traffic Characterization and
Internet Usage in Rural Africa WWW ’11 Conference companion on
World Wide Web, 2011.
Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 7 / 7

Artificial Intelligence for Scheduling Resource Blocks in LTE/5G Networks

  • 1.
    Artificial Intelligence forScheduling Resource Blocks in LTE/5G Networks Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis Department of Computer Science Universidade de Brasilia July 26, 2018 Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 1 / 7
  • 2.
    Outline 1 Introduction 2 Motivation 3Current Work State Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 2 / 7
  • 3.
    Intro. Brazil EU (Finland,Spain) Collaboration Figure: 5G RANGE Project 5G Range Project 50 km range 100 mb/s end to end Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 3 / 7
  • 4.
    Motivation Problem Such high demandsrequire better resource allocation, meaning usual solutions are unlikely to respond well enough. Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 4 / 7
  • 5.
    Motivation Problem Such high demandsrequire better resource allocation, meaning usual solutions are unlikely to respond well enough. Solution Artificial Neural Networks are quite high, performance wise, in most areas of technology at the moment due to it‘s great “learning” capabilities by optimizing a function that tells how far from a good result the prediction is. Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 4 / 7
  • 6.
    Motivation Problem Such high demandsrequire better resource allocation, meaning usual solutions are unlikely to respond well enough. Solution Artificial Neural Networks are quite high, performance wise, in most areas of technology at the moment due to it‘s great “learning” capabilities by optimizing a function that tells how far from a good result the prediction is. New Problem Rural Traffic data sets are not commonly found, meaning that a data set will have to be generated. Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 4 / 7
  • 7.
    Current Work State ASouth African study has characterized Rural Traffic in some areas Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 5 / 7
  • 8.
    Current Work State ASouth African study has characterized Rural Traffic in some areas 68.45% web traffic Second biggest area is videos, including Youtube, Livestreams and VoIP Services (Skype, Hangouts etc) A few percentages are for public alert, calls and SMS. Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 5 / 7
  • 9.
    Current Work State ASouth African study has characterized Rural Traffic in some areas 68.45% web traffic Second biggest area is videos, including Youtube, Livestreams and VoIP Services (Skype, Hangouts etc) A few percentages are for public alert, calls and SMS. Reinforcement Learning based Neural Network is the choice for the model. Current reward system accounts for Quality of Service, Spectral Efficiency, Justice and Guaranteed Bit Rate. All of the rewards are normalized between 0 and 1, whereas 1 is a perfect value. Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 5 / 7
  • 10.
    For Further ReadingI Okvist, Simonsson and Asplund. LTE Frequency Selective Scheduling Performance and Improvements Assessed by Measurements International Symposium on Personal, Indoor and Mobile Radio Communications, 2011. Yang, Xu, Han, Rehman and Tao. GA Based Optimal Resource Allocation for Device to Device WCNC 2014 - Workshop on D2D and Public Safety Communications, 2014. 3GPP, Policy and Charging Control Architecture TS 23.203, 2018. Arulkumaran, Deisenroth, Brundage and Bharath. A Brief Survey of Deep Reinforcement Learning. IEEE Signal Processing Magazine Special Issue on Deep Learning for Image Understanding, 2017. Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 6 / 7
  • 11.
    For Further ReadingII Johnson, Pejovic, Belding and Stam. Traffic Characterization and Internet Usage in Rural Africa WWW ’11 Conference companion on World Wide Web, 2011. Gabriel Ferreira, Guilherme Branco, Marcos Caetano, Priscila Solis (Universidade de Brasilia)Artificial Intelligence for Scheduling Resource Blocks in LTE/5G NetworksJuly 26, 2018 7 / 7