TY - CHAP
T1 - Configuring dynamic heterogeneous wireless communications networks using a customised genetic algorithm
AU - Lynch, David
AU - Fenton, Michael
AU - Kucera, Stepan
AU - Claussen, Holger
AU - O’Neill, Michael
N1 - Publisher Copyright:
© Springer International Publishing AG 2017.
PY - 2017
Y1 - 2017
N2 - 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.
AB - 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.
UR - https://www.scopus.com/pages/publications/85017569580
U2 - 10.1007/978-3-319-55849-3_14
DO - 10.1007/978-3-319-55849-3_14
M3 - Chapter
AN - SCOPUS:85017569580
SN - 9783319558486
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 205
EP - 220
BT - Applications of Evolutionary Computation - 20th European Conference, EvoApplications 2017, Proceedings
A2 - Hidalgo, J.Ignacio
A2 - Cotta, Carlos
A2 - Hu, Ting
A2 - Tonda, Alberto
A2 - Burrelli, Paolo
A2 - Coler, Matt
A2 - Iacca, Giovanni
A2 - Kampouridis, Michael
A2 - Mora Garcia, Antonio M.
A2 - Squillero, Giovanni
A2 - Brabazon, Anthony
A2 - Haasdijk, Evert
A2 - Heinerman, Jacqueline
A2 - D Andreagiovanni, Fabio
A2 - Bacardit, Jaume
A2 - Nguyen, Trung Thanh
A2 - Silva, Sara
A2 - Tarantino, Ernesto
A2 - Esparcia-Alcazar, Anna I.
A2 - Ascheid, Gerd
A2 - Glette, Kyrre
A2 - Cagnoni, Stefano
A2 - Kaufmann, Paul
A2 - de Vega, Francisco Fernandez
A2 - Mavrovouniotis, Michalis
A2 - Zhang, Mengjie
A2 - Divina, Federico
A2 - Sim, Kevin
A2 - Urquhart, Neil
A2 - Schaefer, Robert
PB - Springer Verlag
T2 - 20th European Conference on the Applications of Evolutionary Computation, EvoApplications 2017
Y2 - 19 April 2017 through 21 April 2017
ER -