Multiple Description (MD) source coding has recently emerged as an attractive
framework for robust transmission over channels with transient shutdown characteristics. Examples of
such channels are found in best-effort
heterogeneous packet networks such as the Internet, where congestion and routing
delays can lead to ``holes'' in the transmitted data stream corresponding to lost packets, or wireless
systems during a deep fade. In the current
deployment of the Internet, switches in the network are typically oblivious to the content of the
packets that they process and switch. They provide only
a simple first-in-first-out (FIFO) queuing/scheduling policy, and indiscriminately discard incoming
packets when output queues are full. On the
other hand typical image/video bitstreams are highly structured, i.e., characterized
by a natural hierarchy of importance layers or resolutions. The reconstructed quality for such
hierarchical bitstreams depends on which layers are received.
This therefore underlines the need for an efficient mechanism that converts a scalable, prioritized
bitstream into a non-prioritized one so
that is is better ``matched'' to the existing infrastructure.
The basic idea in MD coding is to generate multiple independent descriptions of
the source such that each description independently describes the source with a certain fidelity, and
when more than one description is
available, they can be synergistically combined to enhance the quality. The reconstructed quality
depends only on the fraction of
descriptions delivered to the destination as opposed to "which ones" are actually delivered. We
have proposed a mechanism to transform a scalable source bit stream into a robust MD packet
stream by encoding source ``layers'' of decreasing importance with progressively weaker forward
error correction (FEC) channel codes and
spreading these codes across packets. The mechanism possesses features of graceful degradation in
quality as a function of the number of packet
losses.
Contact: Rohit Puri
The idea of this project is to model the wireless channel as a packet-loss channel by applying RCPC codes to the transmitted data and using CRC codes to determine if the packet is successfully decoded at the receiver, then applying Reed-Solomon codes to allow data to be reconstructed with packet losses. The novel part of this system is the use of an efficient optimizer to choose the protection level for each Reed-Solomon column to reduce the expected error at the receiver. (We actually bias the solution slightly toward reducing the peak error at the expense of the average MSE, so it's not optimal in the strict sense of the word.) The paper "Wireless Image Transmission Using Multiple Description Based Concatenated Codes" describes this system; it is available below.
Contact: Dan Sachs
We consider the problem of synchronous video transmission over noisy channels. Unlike data or image transmission, transmission of video is a delay constrained problem. In a situation where end-to-end delay is a bottleneck, an approach based on Automatic Repeat reQuest (ARQ) alone may not be the best solution, because of the retransmission delays involved. In situations where end-to-end delay requirements permit the use of ARQ, ARQ schemes have been found to work best when the probability of bit error in the transmission channel is low so that there are only occasional retransmissions. When error probabilities are higher, use of ARQ based schemes is not advisable because of the excessive number of retransmissions involved which eat away the available bandwidth and moreover result in large delays. Forward error correction (FEC) techniques, where error check information is appended to data for error correction are useful in such situations but they are wasteful for the low error case since they constantly burn the available bandwidth. We propose that for the case of synchronous video transmission over channels with small propagation delays (so that ARQ is not forbidden), hybrid ARQ/FEC schemes which have the flexibility to switch between the two error recovery modes in a ``soft'' fashion depending upon the error regime in which the channel is, have the potential to outperform the individual schemes.
We propose an Automatic Repeat Request (ARQ) / Forward Error
Correction
(FEC) scheme for synchronous transmission of video over a binary symmetric
constant rate channel. The approach consists of jointly allocating
source and channel rates to video blocks from a given admissible set subject
to the buffer or equivalently end-end delay constraints. The channel
codes used are the popular class of powerful FEC codes known as Rate-Compatible
Punctured Convolutional (RCPC) Codes. These codes have the rate
compatibility property in the sense that a weaker channel code is a prefix
of a stronger channel code. The method used involves independent
coding of the video units and the optimal partitioning of source rates and
channel rates so as to maximize the expected delivered end-to-end video
quality. The existence of a return channel is assumed through which the
decoder informs the encoder about the success/failure of the transmission. In
the event of a failure, the "incremental" parity information as opposed to
retransmission of the whole unit (modified Type III ARQ) is sent to the decoder
for correcting errors and a reallocation performed at the encoder.
Contact: Rohit Puri