TY - GEN

T1 - Stadium Antennas Deployment Optimization

AU - Yuskov, Alexander

AU - Kulachenko, Igor

AU - Melnikov, Andrey

AU - Kochetov, Yury

N1 - Conference code: 23

PY - 2024

Y1 - 2024

N2 - The stadium is divided into sectors. Each sector is split into cells. Users in the cells must be provided with a certain quality of signal from antennas assigned to their sector. Our goal is to select antenna types, their location, assignment to sectors, and orientation to optimize the signal distribution, measured by three different metrics under some technical constraints. The quality metrics are signal quality, average signal-to-interference ratio (SIR), and consistency. Each variant of antenna deployment is evaluated by a simulator. Thus, we deal with a constrained black-box optimization problem with three objectives. To tackle the problem, we design a three-stage algorithmic approach. In the first stage, we apply a fast constructive heuristic. Later on, a local improvement procedure is called. Finally, a VNS metaheuristic is used to get high-quality solutions. The approach demonstrates strong performance and ability to improve signal quality by 7% and SINR by at least 14% without worsening the given consistency threshold for test instances with up to 7 antenna types, 19 sectors, and 4426 cells.

AB - The stadium is divided into sectors. Each sector is split into cells. Users in the cells must be provided with a certain quality of signal from antennas assigned to their sector. Our goal is to select antenna types, their location, assignment to sectors, and orientation to optimize the signal distribution, measured by three different metrics under some technical constraints. The quality metrics are signal quality, average signal-to-interference ratio (SIR), and consistency. Each variant of antenna deployment is evaluated by a simulator. Thus, we deal with a constrained black-box optimization problem with three objectives. To tackle the problem, we design a three-stage algorithmic approach. In the first stage, we apply a fast constructive heuristic. Later on, a local improvement procedure is called. Finally, a VNS metaheuristic is used to get high-quality solutions. The approach demonstrates strong performance and ability to improve signal quality by 7% and SINR by at least 14% without worsening the given consistency threshold for test instances with up to 7 antenna types, 19 sectors, and 4426 cells.

KW - black box optimization

KW - local search

KW - quality of signal

KW - simulation

KW - wireless network

UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-85198405288&origin=inward&txGid=7c09af7b834c4041cc374e453e77a904

UR - https://www.mendeley.com/catalogue/1ef8f34d-6d37-3977-a126-fbdf63252d73/

U2 - 10.1007/978-3-031-62792-7_30

DO - 10.1007/978-3-031-62792-7_30

M3 - Conference contribution

SN - 9783031627910

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 449

EP - 461

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

PB - Springer Science and Business Media Deutschland GmbH

T2 - 23rd International Conference on Mathematical Optimization Theory and Operations Research

Y2 - 30 June 2024 through 6 July 2024

ER -