Comparative Analysis of Channel Estimation Techniques in SISO, MISO and MIMO Systems

Abstract

The ever-growing need for high data rate, bandwidth
efficiency, reliability, less complexity and less power consumption
in our communication systems is on the increase.
Modern techniques have to be developed and put in place to
meet these requirements. Research has shown, that compared to
conventional Single input Single output (SISO) systems, Multipleinput
Single output (MISO), and Multiple-input multiple-output
(MIMO) can actually increase the data rate of a communication
system, without actually requiring more transmit power or
bandwidth. This paper aims at the investigation of the existing
channel estimation techniques. Based on the pilot arrangement,
the block type and comb type are compared, employing the
Least Square estimation (L.S) and Minimum Mean Squared
Error (MMSE) estimators. Pilots occupy bandwidth, minimizing
the number of pilots used to estimate the channel, in order
to allow for more bandwidth utilization for data transmission,
without compromising the accuracy of the estimates is taken
into consideration. Various channel interpolation techniques and
pilot-data insertion ratio are investigated, simulated and compared,
to determine the best performance technique with less
complexity and minimum power consumption. As performance
measures, the Mean squared error (MSE) and Bit error rate
(BER) as a function of Signal to noise power ratio (SNR) of
the different channel estimation techniques are plotted, in order
to identify the technique with the most optimal performance.
The complexity and energy efficiency of the techniques are also
investigated. The system modelling and simulations are carried
out using Matlab simulation package. The MIMO gives the
optimum performance, followed by the MISO and SISO. This
is as a result of the diversity and multiplexing gain experienced
in the multiple antenna techniques using the STBC.