Performance Enhancement of Smart Antennas Algorithms for Mobile Communications System

Abstract

—In order to improve the performance 
of mobile communications systems, this research 
suggests two novel smart antenna algorithms based 
on a combined approach. The first suggested 
combination approach combines the pure 
Normalized Least Mean Square (NLMS) and 
Conjugate Gradient Method (CGM) algorithms to 
create a new algorithm known as CGM-NLMS. 
The second suggested technique, known as the 
CGM-MLLMS algorithm, will combine the pure 
CGM and Modified NLMS algorithms. One 
variable regularization parameter that is fixed in 
the traditional NLMS technique is the MNLMS 
algorithm. Rather than using a fixed regularization 
parameter, the regularization parameter makes use 
of a reciprocal of the estimate error square of the 
update step size of NLMS. The estimated weight 
coefficients obtained from the first stage of the 
measurements of the relevant correlation 
functions are needed [1].
LMS algorithm that uses 
CGM algorithm are kept and then used as starting 
weight coefficients for processing the NLMS (or 
MNLMS) algorithm using the newly suggested 
CGM-NLMS and CGM-MNLMS algorithms. 
According to simulation results of an adaptive 
beamforming system using a fading channel and a 
Jakes power spectral density channel model, the 
two newly suggested algorithms outperform the 
pure CGM and pure NLMS algorithms in terms of 
speedy convergence, increased interference 
suppression, and low levels of mean square 
deviation (MSD) and minimum mean square error 

(MSE) at steady state.