Configuring dynamic heterogeneous wireless communications networks using a customised genetic algorithm

  • David Lynch
  • , Michael Fenton
  • , Stepan Kucera
  • , Holger Claussen
  • , Michael O’Neill

Research output: Chapter in Book/Report/Conference proceedingsChapterpeer-review

Abstract

Wireless traffic is surging due to the prevalence of smart devices, rising demand for multimedia content and the advent of the “Internet of Things”. Network operators are deploying Small Cells alongside existing Macro Cells in order to satisfy demand during this era of exponential growth. Such Heterogeneous Networks (HetNets) are highly spectrally efficient because both cell tiers transmit using the same scarce and expensive bandwidth. However, load balancing and cross-tier interference issues constrain cell-edge rates in co-channel operation. Capacity can be increased by intelligently configuring Small Cell powers and biases, and the muting cycles of Macro Cells. This paper presents a customised Genetic Algorithm (GA) for reconfiguring HetNets. The GA converges within minutes so tailored settings can be pushed to cells in real time. The proposed GA lifts cell-edge (2.5th percentile) rates by 32% over a non-adaptive baseline that is used in practice. HetNets are highly dynamic environments. However, customers tend to cluster in hotspots which arise at predictable locations over the course of a typical day. An explicit memory of previously evolved solutions is maintained and used to seed fresh runs. System level simulations show that the 2.5th percentile rates are boosted to 36% over baseline when prior knowledge is utilised.

Original languageEnglish
Title of host publicationApplications of Evolutionary Computation - 20th European Conference, EvoApplications 2017, Proceedings
EditorsJ.Ignacio Hidalgo, Carlos Cotta, Ting Hu, Alberto Tonda, Paolo Burrelli, Matt Coler, Giovanni Iacca, Michael Kampouridis, Antonio M. Mora Garcia, Giovanni Squillero, Anthony Brabazon, Evert Haasdijk, Jacqueline Heinerman, Fabio D Andreagiovanni, Jaume Bacardit, Trung Thanh Nguyen, Sara Silva, Ernesto Tarantino, Anna I. Esparcia-Alcazar, Gerd Ascheid, Kyrre Glette, Stefano Cagnoni, Paul Kaufmann, Francisco Fernandez de Vega, Michalis Mavrovouniotis, Mengjie Zhang, Federico Divina, Kevin Sim, Neil Urquhart, Robert Schaefer
PublisherSpringer Verlag
Pages205-220
Number of pages16
ISBN (Print)9783319558486
DOIs
Publication statusPublished - 2017
Externally publishedYes
Event20th European Conference on the Applications of Evolutionary Computation, EvoApplications 2017 - Amsterdam, Netherlands
Duration: 19 Apr 201721 Apr 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10199 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference20th European Conference on the Applications of Evolutionary Computation, EvoApplications 2017
Country/TerritoryNetherlands
City Amsterdam
Period19/04/1721/04/17

Fingerprint

Dive into the research topics of 'Configuring dynamic heterogeneous wireless communications networks using a customised genetic algorithm'. Together they form a unique fingerprint.

Cite this