On Placement of LAA / LTE-U Base Stations in the Presence of Wi-Fi

Baswade, Anand M and M R, Shashi Kiran and Tamma, Bheemarjuna Reddy and Franklin, Antony (2018) On Placement of LAA / LTE-U Base Stations in the Presence of Wi-Fi. In: 19th International Conference on Distributed Computing and Networking, 4 - 7 January 2018, Varanasi.

Preview

Text
ICDCN_2018.pdf
Download (618kB) | Preview

Abstract

Recently, the use of LTE in unlicensed has gained a lot of attention to improve the data rate and satisfy the increasing user demand. Use of unlicensed spectrum is allowed with restricted transmission power hence mostly LAA/LTE-U nodes are Femto cells and mostly will be used to improve the indoor data rates. In future, the LAA nodes will get deployed inside residential apartment complexes and office buildings to provide high data rates for indoor User Equipment (UEs). The UEs with high Signal-to-Interference plus Noise Ratio (SINR) experience good throughput, but the SINR decreases significantly if building walls and other obstacles are present in the communication path. The coverage difference of licensed and unlicensed spectrum provides different SINR so, efficient placement of LAA Femtos in buildings with consideration of interference from other LAA Femto nodes, Wi-Fi Access Points (APs) and Macro Base Stations (BSs) is very crucial for attaining desirable SINR for the indoor UEs. In this paper, we have considered obstructions (walls, floors) and interference from Macro, Femto BSs, and Wi-Fi APs. Our goal is to find the optimal number of Femtos with minimum SINR requirements in the unlicensed band, and next thing is to achieve good SINR in all sub-regions in licensed band as well. To do this, we formulated an optimization model (MinLF) to find the minimum number of LAA/LTE-U Femtos along with its positions in indoor scenarios which resulted in Mixed Integer Linear Programming (MILP) optimization problem. Further, to maximize minimum SINR in the licensed spectrum, we formulated MinLPow model to find the optimal power of each located Femto inside the building which is Mixed Integer Non-Linear Programming (MINLP). MinLF formulation is solved using GAMS CPLEX optimization solver. After getting positions of Femtos, we solved MinLPow formulation using Genetic Algorithm. Finally, the efficiency of proposed formulation model is shown with sufficient simulation study.

[error in script]

IITH Creators: