WO1998032284A1

WO1998032284A1 - Improvements in or relating to switching between compressed bitstreams

Info

Publication number: WO1998032284A1
Application number: PCT/GB1997/003547
Authority: WO
Inventors: Alois Martin Bock; Nigel Stephen Dallard
Original assignee: Nds Limited
Priority date: 1997-01-17
Filing date: 1997-12-24
Publication date: 1998-07-23
Also published as: GB9700956D0; EP0951779A1; AU5333198A

Abstract

According to one aspect of the present invention, there is provided a method of switching from a first bitstream to a second bitstream, in a system having a decoder with a buffer through which the bitstream is transmitted, the method comprising: stopping the decoder from carrying out the decoding process in a predetermined manner at a predetermined point on the first bitstream, such that the buffer is substantially emptied; and restarting the decoder in a controlled manner at a predetermined position on the second bitstream, thereby switching from the first bitstream to the second bitstream. This invention allows insertion of additional programs, advertisements and the like into a standard MPEG stream without any of the disadvantages experienced using the hooks provided for in MPEG-2. One advantage of the proposed method is its simplicity. In terms of the MPEG-2 standard it requires intervention essentially only at the transport layer. The transport packet headers in MPEG-2 indicate the start of packets of elementary streams (PES packets). This flag is used for the control of all events in the switching process. The invention also allows for simple insertion of additional information in other types of bitstream.

Description

Improvements in or relating to Switching between compressed bitstreams

This invention relates to improvements relating to switching between

compressed bitstreams. Particularly but not exclusively to insertion of

regional programs or advertisements to the compressed bitstream.

In the field of digital transmission of information, including

broadcasting, data for transmission may undergo compression to reduce

the amount of bandwidth required to transmit the information. The method

by which the information is compressed is standardised. This means that

information can be compressed and decompressed by all users in a known

manner. The description below is based on the requirements for switching

a video bitstream which has been compressed according to the ISO/IEC

international standard 13818, also known as "MPEG-2". Although the same

principle can be used for audio signals and indeed for any bitstream with

timing information.

An MPEG transmission system allows several video, audio and

associated services to be multiplexed and sent over a single digital

transmission channel. The information to be transmitted is compressed into a

single continuous transport stream, with the different programs, data,

advertisements etc. at different points along the stream. Compression of

video signals makes use of the spatial and temporal redundancy, i.e.

predictability of the source signal to reduce the data rate of the compressed

signal. This means that the decoded signal is, to a large extent, dependent

on the coding history of the signal. A switch from one compressed signal to another at an arbitrary point will, therefore, generally result in a total break-up

of the displayed picture, recovering only after synchronisation is regained in

the decoding algorithm. Furthermore, even if the switch between the

compressed bitstreams is carefully timed to coincide with the start of a new

sequence, the occupancy of the decoding buffer after the switch is generally

not defined resulting in possible buffer overflow or underflow some time after

the switch. This scenario is further complicated when the bitstreams are

encrypted as would be common for a Pay-TV channel.

Accordingly, one object of the present invention is to provide a

method and apparatus which will allow seamless or near seamless

insertion of additional information in a bitstream, for example, a transport

stream.

According to one aspect of the present invention, there is provided a

method of switching from a first bitstream to a second bitstream, in a system

having a decoder with a buffer through which the bitstream is transmitted, the

method comprising: stopping the decoder from carrying out the decoding

process in a predetermined manner at a predetermined point on the first

bitstream, such that the buffer is substantially emptied; and restarting the

decoder in a controlled manner at a predetermined position on the second

bitstream, thereby switching from the first bitstream to the second bitstream.

This invention allows insertion of additional programs, advertisements

and the like into a standard MPEG stream without any of the disadvantages

experienced using the hooks provided for in MPEG-2. One advantage of the

proposed method is its simplicity. In terms of the MPEG-2 standard it requires intervention essentially only at the transport layer. The transport

packet headers in MPEG-2 indicate the start of packets of elementary

streams (PES packets). This flag is used for the control of all events in the

switching process. The invention also allows for simple insertion of additional

information in other types of bitstream.

According to a second aspect of the present invention there is provided

apparatus for switching from a first bitstream to a second bitstream, in a

system having a decoder with a buffer through which the bitstream is

transmitted, comprising: means for causing the decoder to stop the decoding

process in a predetermined manner at a predetermined point on the first

bitstream, such that the buffer is substantially emptied; and means for

causing the decoder to restart in a controlled manner at a predetermined

position on the second bitstream, thereby switching from the first bitstream to

the second bitstream.

Reference will now be made, by way of example, to the accompanying

drawings, in which:

Figure 1 is a timing diagram of the switching process according to one

aspect of the present invention; and

Figure 2 is a block diagram of the proposed switching circuit for

accomplishing the invention.

A method is described below which avoids the problems of the prior art

by going through a sequence of events such that the decoder will re-gain

synchronisation with the new bitstream in a defined and consistent way, and

with only a freeze-frame effect on the displayed picture or a short mute in audio. Optional insertion of a marker signal indicating that a switch is

imminent can make the decoding more robust and help to achieve a faster

switch. Although video switching is discussed below, the same principle can

be used for audio signals and indeed for any bitstream with timing

information.

The principle behind the proposed system is the emulation of a

receiver or decoder channel change in the switching circuit, i.e. the bitstream

being fed to the decoders in the receiver is essentially the same as if the user

had locally changed to a different channel. When the user changes channel

the decoder buffers are flushed and decoding re-starts on the new channel.

To achieve the same effect in a switch upstream of the decoder the bitstream

is interrupted for a short period of time. The duration of the interruption is just

long enough to make sure that the decoding buffer in the receiver is

completely emptied. This forces the decoder to re-start the decoding process

as soon as the signal transmission is resumed. Insertion of an end-of-

sequence marker just prior to the interruption makes sure that the bitstream

remains compliant to the standard throughout the switching process and

helps the decoder to re-gain synchronisation more quickly by giving advanced

warning of the signal interruption. This ensures that a clean transition from

the first bitstream to the second bitstream occurs, such that the transition

appears to be seamless to the viewer.

Figure 1 shows a timing diagram of the switching process. Initially,

input bitstream 1 is selected. The switching process is initiated by an external

request. PES packet 1a remains switched to the output until the start of the next PES packet 1 b. At this point the switching sequence starts by inserting

an optional end-of-sequence header into the bitstream and starting the guard

interval. The guard interval is calculated as the time it takes to empty the

receiver buffer, i.e.

t_guard = buffer size / bitrate

After the guard interval the bitstream remains switched off until the arrival of

the next PES packet of the new bitstream. At that stage input bitstream 2 is

switched to the output and the switching process is complete.

Figure 2 shows a block diagram of an example implementation of the

proposed system. Part of a decoder or receiver 20 is shown, in which a

decoder buffer 22 is represented. It is assumed that switching should occur

between two independent transport streams 24, 26. The transport packets of

the two streams are aligned using a FIFO 28 in one of the two input paths 30,

32. The transport headers of both streams are decoded at filters 34, 36 and

the information, i.e. transport-packet-identification (PID), payload-unit-start-

indicator, etc. is passed on to the control circuit 38. The transport packet

headers are unencrypted, even if the payload is encrypted, so this information

is always available. The control circuit also receives the switching request, for

example, from an external control computer. After a switching request is

received the control circuit goes through the sequence of events as shown in

Figure 1 by selecting one of the four signal sources. The transport packet

identification (PID) in the transport packet headers of the video PES packets

in input transport stream 2 and the sequence-end-code inserter are adjusted such that they match those in the transport packet headers of the video PES

packets in input transport stream 1.

The Program Specific Information (PSI) carried in the transport stream

to enable a decoder to locate the various components (e.g. video, audio) of a

program in the transport stream is not changed during the switching process.

If the PSI is changed the decoder will detect this change at some

indeterminate time, and commence to reconfigure itself, flush its buffers and

attempt to resynchronise with the incoming bitstream. By changing the

transport-packet-identification (PID) fields in the headers of the transport

packets to those used by the original program, changing the PSI is not

required, and this ensures that the decoder is in a known state throughout the

switching process.

If transport stream 26 is encrypted, it is necessary to deliver the

Entitlement Control Messages (ECMs) associated with that stream to the

decoder ahead of the packets containing the program components, in order

that they can be decrypted by the decoder to obtain the key used to encrypt

the program. These messages replace those from input transport stream 24

immediately the request to switch is received, and the PID field in the header

is adjusted to match that of the equivalent messages in transport stream 24.

The switch to PES packets from transport stream 26 should be delayed for

long enough to ensure that the encryption key carried in the ECMs is

available at the receiver before the PES packets from transport stream 26 are

received. Irrespective of the state of the transport-scrambling-control field in the

packet headers of transport streams 24 and 26, the packets from the

sequence-end-code inserter are unencrypted, and the transport-scrambling-

control field should indicate such. As this data contains nothing of any value

there is no need for it to be encrypted.

In MPEG bitstreams, audio frames may not be aligned with transport

stream packets. If such non-aligned bitstreams are spliced at the transport

stream layer, e.g. at PES packet boundaries as is the case in video, there is a

danger that fragments of audio frames are presented to the decoder at the

splice points. This can lead to severe audio distortion and noise levels.

This problem can be overcome by processing the bitstreams at the

PES layer. At the splice-out position, after the last complete audio frame, the

audio bitstream is replaced by a series of zeros to the end of the current PES

packet. Similarly, the new PES packet at the splice-in position is replaced

with zeros until the first full valid audio frame is found in the target bitstream.

Since audio frame headers, in particular, are removed from the bitstream

during the splice period, the decoder is forced to stop the decoding process

and mute its output until audio frame headers are once again found.

Alternatively, MPEG-2 adaptation fields may be used. In this case the

audio bitstream is interrupted exactly at the end of an audio frame. The rest of

the corresponding transport stream packet is replaced with an empty

adaptation field. Similarly, the new bitstream is started up exactly with an

audio frame header with the first part of the corresponding transport stream

packet replaced by an adaptation field. This invention is of use where additional information is to be included

into a transport stream. For example, the insertion of advertisements,

regional programs, announcements and the like into a broadcast transport

stream. As the present invention takes place at the transport packet level, the

insertion of information can occur at regional broadcast stations and main

head end broadcast stations. In addition it would be possible to apply the

invention not only to the broadcasting of television, but also the broadcast of

any other type of signals.

The invention provides a clean manner by which information or

material can be inserted into a digital transport media.

Claims

1. A method of switching from a first bitstream to a second bitstream, in a

system having a decoder with a buffer through which the bitstream is

transmitted, the method comprising:

stopping the decoder from carrying out the decoding process in a

predetermined manner at a predetermined point on the first bitstream, such

that the buffer is substantially emptied;

restarting the decoder in a controlled manner at a predetermined

position on the second bitstream, thereby switching from the first bitstream to

the second bitstream.

2. The method of claim 1 , further comprising providing the first and

second bitstreams with timing information.

3. The method of claim 1 or claim 2, further comprising providing the first

and second bitstream as first and second transport streams each having one

or more packets of information.

4. The method of claim 3, further comprising stopping the decoder at the

end of a packet of information in the first bitstream.

5. The method of claim 3 or claim 4, further comprising restarting the

decoder at the start of a packet of information of the second bitstream.

6. The method of any preceding claim, further comprising emptying the

buffer by allowing the information therein to pass out of the buffer over a time

period.

7. The method of claim 6, further comprising calculating the time period

as a function of the size of the buffer and the bitrate of the first bitstream.

8. The method of any of claims 1 to 5, further comprising emptying the

buffer by flushing the buffer.

9. The method of any preceding claim, further comprising stopping the

decoder in response to a user input.

10. The method of any of claims 1 to 8, further comprising stopping the

decoder in response to a signal embedded in the first bitstream.

11. The method of any preceding claim, further comprising providing the or

each bitstream with encryption information.

12. The method of claim 11 , further comprising providing the encryption

information at a known location in the second bitstream.

13. The method of any preceding claim, further comprising adjusting any

control information relating to the second bitstream to be equivalent to that of

the first bitstream.

14. Apparatus for switching from a first bitstream to a second bitstream, in

a system having a decoder with a buffer through which the bitstream is

transmitted, comprising:

means for causing the decoder to stop the decoding process in a

predetermined manner at a predetermined point on the first bitstream, such

that the buffer is substantially emptied; and

means for causing the decoder to restart in a controlled manner at a

predetermined position on the second bitstream, thereby switching from the

first bitstream to the second bitstream.

15. The apparatus of claim 14, wherein the first and second bitstreams are

provided with timing information.

16. The apparatus of claim 14 or claim 15, wherein the first and second

bitstream comprise first and second transport streams each having one or

more packets of information.

17. The apparatus of claim 16, wherein the decoder is stopped at the end

of a packet of information in the first bitstream.

18. The apparatus of claim 16 or claim 17, wherein the decoder is

restarted at the start of a packet of information of the second bitstream.

19. The apparatus of any preceding claim, wherein the buffer is emptied by

allowing the information therein to pass out of the buffer over a time period.

20. The apparatus of claim 19, wherein the time period is calculated as a

function of the size of the buffer and the bitrate of the first bitstream.

21. The apparatus of any of claims 14 to 18, wherein the buffer is flushed.

22. The apparatus of any preceding claim, wherein the decoder is stopped

in response to a user input.

23. The apparatus of any of claims 14 to 21 , wherein the decoder is

stopped in response to a signal embedded in the first bitstream.

24. The apparatus of any preceding claim, wherein the or each bitstream

includes encryption information.

25. The apparatus of claim 24, wherein the encryption information is at a

known location in the second bitstream.

26. The apparatus of any of claims 14 to 25, wherein any control

information relating to the second bitstream is adjusted to be equivalent to

that of the first bitstream.

27. The apparatus of any of claims 14 to 26, wherein the buffer is

completely emptied.