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
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.