All practical communications systems introduce errors during the transmission of information. Errors alter or delete parts of the information and usually have a deep impact on the quality of the received information. Their effect depends on the extent and the significance of the error-struck information segment, and in many cases they deem the transmitted message unusable.
For packet networks particularly, errors are usually modelled as packet loss. This could refer to the actual loss of the packet in networks that discard packets when congested, and to the reception of a packet altered by the network infrastructure above a certain degree. In real-time video communication, the packet loss concept is further extended to packets arriving too late to be displayed at the decoder. For example, in streaming video applications over the Internet, delay is the biggest source of packet loss.
The outline of the chapter follows. The following two introductory paragraphs are dealing with the effects of packet loss on coded video and the model applied for error control, respectively. Next, some of the techniques which have been developed in order to protect and recover from packet loss, are presented. They are categorised according to the place where error control or concealment is performed.