Information Technology Reference
In-Depth Information
architecture for transmitting the video streams from a sensor to the UMTS
user. This architecture integrates mobile network architecture with sensor
ad hoc networks, which considerably reduces the implementation cost and
the power consumption. It also suits the deployment of large-scale sensor
networks. Video sensor can provide important visual information in a num-
ber of applications such as health care, environmental monitoring, emergency
response, and secure surveillance. However, when transmitted to a mobile
user, it is difficult to predict and estimate the quality of a video, since the
encoding and decoding and transmission lose will dramatically decrease the
original video quality. This chapter describes an algorithm for approximating
the received video quality by combining the information of the original video
type with channel status. The video type information consists of encoded
video quality, which is decided by the original video motion frequency and
frame lost rate, which is decided by the original video bit rate. The channel
status information indicates the packet lost rate. Predicting the video quality
has many uses, such as choosing the best channel to transmit the video.
3.1 Introduction
With the development growth of third generation telephony (video telephony),
there is a tremendous demand on the third generation Circuit Switched (3G-CS)
video service. The 3G wireless network, foreseen to be the enabling technology
for multimedia services with up to 64 kbps for CS network and 2 Mbps for the
Packet Switched (PS) Network, makes it feasible for visual communication over
t h e w i r e l e is is l i in k . T h e t r a in is m i is is i o in r at e i is u is u a l ly ve r y u in is t a bl e i in w i r e l e is is c h a in -
nel, which is caused by multipath fading, intersymbol interference, and noise
disturbances. The transmission rate varies with the changing external environ-
ment, resulting in devastating effect on multimedia transmission. To cope with
this on a wireless channel, accurate network-condition estimation and effective
flow control are essential for robust video transmission. It is known that video
transmission is delay-sensitive but may be tolerable to some kinds of errors.
Moreover, different portions of video bitstream have different importance to the
reconstructed video quality; thereby giving rise to different network quality of
service (QOS) requirements, such as transmission latency, bit error rate, and so
on. For instance, it is intuitive that lower layers of a layered scalable video codec
have higher network QOS requirements than those of higher layers. Therefore,
adopting different multimedia transmission rate schemes for each portion is
more appropriate for such a compressed video bit stream. However, channel
coding introduced by dynamic multimedia transmission rate would generate
increased computational complexity. Considering the limited bandwidth in the
wireless network of the mobile device, the resource, such as bandwidth, should
be allocated appropriately for source and channel coding.
