Scalable H.264 Wireless Video Transmission over MIMO-OFDM Channels - Advanced Concepts for Intelligent Vision Systems - page 35

and Signal to Noise Ratio (SNR), are considered for estimation purposes. We also
propose two different priority mappings of the scalable layers produced by SVC.
Comparison results for the two priority mappings are presented for bandwidth
constrained and distortion optimized video transmission over a MIMO-OFDM
system. The results exemplify the advantages of the use of each priority mapping
for different video sequences.
The rest of the paper is organized as follows. In section 2, the proposed system
is introduced. In section 3, the scalable extension of H.264 is described. In section
4, the cross-layer optimization problem is formulated and solved. In section 5, the
two video distortion estimation methods are discussed. In section 6, the priority
mapping of the temporal and FGS layers of SVC is discussed. In section 7,
experimental results are presented. Finally, in section 8, conclusions are drawn.

2 System Description

In our packet-based video transmission system, we utilize channel coding fol-
lowed by orthogonal space-time block coding for MIMO-OFDM systems. After
video encoding, the scalable layers of each frame are divided into packets of con-
stant size γ , which are then channel encoded using a 16-bit cyclic redundancy
check (CRC) for error detection and rate-compatible punctured convolutional
(RCPC) codes for UEP. These channel-encoded packets are modulated with a
particular constellation size and further encoded using O-STBC for each sub-
carrier for transmission over the MIMO wireless system. A 6-ray typical urban
(TU) channel model with AWGN is considered (details shown in Table 1) and
ML decoding is used to detect the transmitted symbols at each subcarrier, which
are then demodulated and channel decoded for error correction and detection.
All the error-free packets for each frame are buffered and then fed to the source
decoder with error concealment for video reconstruction. For the MIMO-OFDM

Table 1. Six-ray typical urban (TU) channel model

Delay ( µ s) 0.0 0.2 0.5 1.6 2.3 5.0
Power (mean) 0.189 0.379 0.239 0.095 0.061 0.037

system used here, we consider M t = 2 transmit and M r = 2 receive antennas.
We used the O-STBC design for MIMO-OFDM systems in which two codewords
(corresponding to two time instances) are transmitted. The channel is assumed
to be quasi-static for these two codeword time periods. The codeword structure
is as follows:

⎡
⎣

⎤
⎦

x 1 x 2
x 3 x 4
| |
| |
| |
x 2 N− 1 x 2 N

C OF DM 1 = , (1)