FULL PAPER
International Journal of Recent Trends in Engineering, Vol 2, No. 5, November 2009
80
DHT Based 4 QAM OFDM Baseband System and
Channel Estimation
Vijay Kumar Ganapati Panda Prasant Kumar Sahu
Department of E. E. School of Electrical Sciences School of Electrical Sciences .
I.I. T. Madras, Chennai I. I. T. Bhubaneswar, Orissa I. I. T. Bhubaneswar, Orissa
Email: nitrkl.vijay@gmail.com Email: ganapati.panda@gmail.com Email: prof.prasant@gmail.com
Abstract - This paper presents the simulation of 4 Quadrature
Amplitude Modulation (QAM) orthogonal frequency division
multiplexing (OFDM) baseband system and channel
estimation which uses inverse discrete Hartley transform
(IDHT) and discrete Hartley transform (DHT). As the
calculation of DHT and IDHT involves real operations hence
the computational complexities are less as compared to DFT
and IDFT. Moreover as IDHT is same as DHT hence we can
use same hardware for both, while the DFT and IDFT require
separate hardware to implement. As compared to DFT based
OFDM system, the simulated DHT based OFDM system
achieves approximately the same transmission performance
with less computational complexity and hardware
requirements.
Index Terms - Channel estimation, DHT, DFT, QAM, OFDM
I INTRODUCTION
Frequency division multiplexing (FDM) is used to
transmit multiple signals simultaneously over a wired or a
wireless system. Each signal is limited by a specific
frequency band, and is modulated by a data stream.
Orthogonal Frequency Division Multiplexing (OFDM) [7,
8] is a special case of this, where the data is distributed over
a large number of carriers that are ‘orthogonal’ to each
other. OFDM is spectrally efficient compared to the
conventional FDM system, since it does not need any guard
bands between adjacent channels. This orthogonality
property is the heart of OFDM, since the interference due to
other carriers is prevented, when the receiver demodulates a
particular carrier.
The OFDM system avoids the tracking of a time varying
channel by the use of differential phase shift keying
(DPSK) in. However this will limit the number of bits per
symbol and results in 3 db loss in signal-to-noise ratio
(SNR) [2]. If the receiver contains a channel estimator,
multipath signaling scheme can be used. OFDM system
offers high bit rate transmission over a frequency fading
channel due to use of spectrally efficient quadrature
amplitude modulation (QAM). Because of high efficiency
with no inter-symbol-interference (ISI), OFDM has been
standardized or extensively investigated for various
applications like digital radio applications and wireless
communications [1].
Implementation of the DFT based OFDM system requires
the complex calculation of a long length IDFT and DFT on
the transmitter and receiver side of the OFDM system.
Such a long length IDFT and DFT calculation requires a
huge number of complex multiplications and additions.
In this paper we simulated a discrete Hartley transform
(DHT) [10] based OFDM system which uses IDHT and
DHT on the transmitter and receiver side. DHT is purely
real transform. The calculation of DHT involves only real
multiplications and additions and it is having identical
inverse. Here, we present a comparative analysis of DFT
and DHT based OFDM system. The objective is to examine
the computational complexities and bit error rate (BER)
performance after channel estimation of DHT and DFT
based OFDM system [12]. The rest of the work is
organized as follows. In Section II DHT based OFDM
system model is described. Section III includes the channel
estimation algorithms. Section IV contains performance
evaluation results obtained by means of simulations. This is
followed by conclusions in Section V.
The similar type of work might have been done by
someone else too but the methodology adapted in this paper
is totally new up to best of our knowledge and data
available to us.
II DHT BASED OFDM SYSTEM MODEL
The OFDM system is modeled employing the
following assumptions.
1) Accurate time & frequency synchronization at the
receiver.
2) The channel impulse response length (L
t
) is
smaller than the cyclic prefix (CP) length of the
OFDM symbol in order to avoid inter-block
interference & preserve orthogonality of the
OFDM symbol.
We will consider the system shown in Fig. 1, where
k
x are
the transmitted symbols, ) (t g is the channel impulse
© 2009 ACADEMY PUBLISHER