TY - GEN
T1 - Optimisation of CDMA-based mobile telephone networks
T2 - 9th International Conference on Parallel Problem Solving from Nature
AU - Weal, Paul
AU - Corne, David
AU - Murphy, Chris
PY - 2006
Y1 - 2006
N2 - CDMA and WCDMA mobile phone networks depend on a network of antennae, each defining a geographic 'cell' that handles the transmissions to and from users' handsets within that cell. These antennae have adjustable settings whose values have a large effect on both quality of service (and consequent subscriptions) and resource consumption. We consider the optimisation of these parameters, and describe experiments that compare a range of optimisation algorithms with the methods currently used in the field for this purpose. The aim of the current project was to achieve faster (necessary) and better (if possible) results than the existing methods used by field engineers. We find that certain evolutionary algorithm configurations achieve both of these requirements on test problems arising from real data from a high-traffic urban environment. To some extent the ideal algorithm depends on the size and load in the network being optimised, and this is the main topic of ongoing research. © Springer-Verlag Berlin Heidelberg 2006.
AB - CDMA and WCDMA mobile phone networks depend on a network of antennae, each defining a geographic 'cell' that handles the transmissions to and from users' handsets within that cell. These antennae have adjustable settings whose values have a large effect on both quality of service (and consequent subscriptions) and resource consumption. We consider the optimisation of these parameters, and describe experiments that compare a range of optimisation algorithms with the methods currently used in the field for this purpose. The aim of the current project was to achieve faster (necessary) and better (if possible) results than the existing methods used by field engineers. We find that certain evolutionary algorithm configurations achieve both of these requirements on test problems arising from real data from a high-traffic urban environment. To some extent the ideal algorithm depends on the size and load in the network being optimised, and this is the main topic of ongoing research. © Springer-Verlag Berlin Heidelberg 2006.
UR - http://www.scopus.com/inward/record.url?scp=33750230546&partnerID=8YFLogxK
M3 - Conference contribution
SN - 3540389903
SN - 9783540389903
VL - 4193
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 402
EP - 412
BT - Parallel Problem Solving from Nature, PPSN IX - 9th International Conference, Procedings
Y2 - 9 September 2006 through 13 September 2006
ER -