Scalable H.264 Wireless Video Transmission over MIMO-OFDM Channels - Advanced Concepts for Intelligent Vision Systems - page 34

On the other hand, OFDM mitigates the undesirable effects of a frequency-
selective channel by converting it into a parallel collection of frequency-flat sub-
channels. OFDM is basically a block modulation scheme where a block of N
information symbols is transmitted in parallel on N subcarriers. The subcarriers
have the minimum frequency separation required to maintain orthogonality of
their corresponding time domain waveforms, yet the signal spectra corresponding
to the different subcarriers overlap in frequency. Hence, the available bandwidth
is used very eciently. An OFDM modulator can be implemented as an in-
verse discrete Fourier transform (IDFT) on a block of N information symbols.
To mitigate the effects of intersymbol interference (ISI) caused by channel time
spread, each block of N IDFT coecients is typically preceded by a cyclic prefix
(CP) or a guard interval consisting of G samples, such that the length of the
CP is at least equal to the channel length. As a result, the effects of the ISI
are easily and completely eliminated. Recent developments in MIMO techniques
promise a significant boost in performance for OFDM systems. A parallel trans-
mission framework for multimedia data over spectrally shaped channels using
multicarrier modulation was studied in [4]. A space-time coded OFDM system
to transmit layered video signals over wireless channels was presented in [5].
Video transmission with OFDM and the integration of STC with OFDM have
been studied recently [6,7,8]. In [9] an optimal resource allocation method was
proposed for multilayer wireless video transmission by using the large-system
performance analysis results for various multiuser receivers in multipath fad-
ing channels. However, the above approaches have not exploited wireless video
transmission over MIMO-OFDM systems with bandwidth optimization.
In this paper, we consider the bandwidth constrained transmission of temporal
and quality scalable layers of coded video over MIMO-OFDM wireless networks,
with optimization of source coding, channel coding and physical layer parameters
on a per group of pictures (per-GOP) basis. The bandwidth allocation problem
is addressed by minimizing the expected end-to-end distortion (for one GOP
at a time) and optimally selecting the quantization parameter (QP), channel
coding rate and the constellation for the STBC symbols used in this MIMO-
OFDM system. At the source coding side, we use the scalable video coding
(SVC) extension of the H.264/AVC standard which has an error-resilient network
abstraction layer (NAL) structure and provides superior compression eciency
[10]. The combined scalability provided by the codec is exploited to improve the
video transmission over error-prone wireless networks by protecting the different
layers with unequal error protection (UEP). In [11,12], we proposed bandwidth
optimization algorithms for SVC video transmission over MIMO (non-OFDM)
channels using O-STBC.
A good knowledge of the total end-to-end decoder distortion at the encoder
is necessary for such optimal allocation. Accordingly, we use the low-delay, low-
complexity method for accurate distortion estimation for SVC video as discussed
in [11] and also propose a new modified version of this distortion estimation
method. The two distortion estimation methods differ in the priority order in
which different types of scalability inherent in the SVC codec, namely temporal