Patent Application
20070171942
(Not applicable)
The invention relates to telecommunications systems, and more particularly, to cable television systems using data stream compression.
It is known in the art to reduce channel change delay due to MPEG compression by reducing the duration between random access points such as I frames. Specifically, according to the well-known MPEG standard, the duration between random access points, or I-frames, which are referred to as anchor frames because they are “free-standing” and can be used as the basis to begin image construction, can be as long as 2 seconds. This means that a decoder that first begins to construct an MPEG-encoded image may introduce a delay of up to 2 seconds or more before the image can be fully reconstructed. A user switching channels will find such a delay frustrating.
To alleviate this problem, upon a channel change request, subscribers can be sent two versions of a programming channel. The first version is the fully-compressed data stream in accordance with the MPEG standard. The second version is one in which I-frames, or random access points, appear more frequently, for example every other frame, reducing the delay before an I-frame is received and reconstruction can begin. Once the image of the second data stream is reconstructed, the subscriber system monitors the fully-compressed data stream, and at the next random access point available in this fully-compressed data stream, the subscriber system seamlessly switches from the second, less-compressed data stream to the fully-compressed data stream. The subscriber system then signals that the switch to the fully-compressed data stream has occurred, indicating that the second, less-compressed version of the data stream no longer needs to be transmitted, in order to conserve transmission resources. This approach works well to reduce channel change delay. However, it requires transmission to the subscriber, at least initially, of both the fully-compressed and second version of the data streams. This is problematic because bandwidth to the subscriber in particular is at a premium and there is a long-felt need to conserve as much of it as possible.
In accordance with one aspect of the invention, there is provided a method for reducing channel switch delay on a unicast channel through which programming is transmitted from a switch to a subscriber. The method includes, responsive to a channel change signal, transmitting over the unicast channel a partially-compressed data stream containing anchor frames that are undergoing transmission at a first rate; monitoring a fully-compressed data stream corresponding to the partially-compressed data stream and containing anchor frames that are undergoing transmission at a second rate lower than the first rate; and replacing transmission of the partially-compressed data stream with transmission of the fully-compressed data stream over the unicast channel when an anchor frame is detected in the monitored fully-compressed data stream.
Further in accordance with an aspect of the invention, there is provided a method for enabling fast channel change. The method includes generating for each of a plurality of programs a partially-compressed data stream and a fully-compressed data stream, the partially-compressed data stream containing more anchor frames per time duration than the fully-compressed data stream; transmitting the partially-compressed and fully-compressed data streams to a switch; and, at any given time, transmitting from the switch to a subscriber a single, either fully-compressed or partially compressed data stream.
Further in accordance with an aspect of the invention, there is provided a fast channel change system. The system includes a plurality of encoder pairs, each pair being associated with a program and including first and second encoders respectively generating a fully-compressed data stream and a partially-compressed data stream that are associated with the program, the partially-compressed data stream containing more anchor frames per time duration than the fully-compressed data stream. The system also includes a switch adapted to receive fully-compressed and partially-compressed data streams from a plurality of encoder pairs and to transmit a single data stream in response to a channel change signal.
Further in accordance with an aspect of the invention, there is provided, in a fast channel change system in which, for each of a plurality of programs, a partially- and a fully-compressed data streams are generated, the partially-compressed data stream containing more anchor frames per time duration than the fully-compressed data stream, a switch that includes a switching module and a frame monitor in communication with the switching module. The frame monitor communicates an anchor frame detection signal to the switching module to thereby cause the switch to switch transmission from the switch from a partially-compressed data stream associated with a first program to a fully compressed data stream associated with the first program.
Many advantages of the present invention will be apparent to those skilled in the art with a reading of this specification in conjunction with the attached drawings, wherein like reference numerals are applied to like elements, and wherein:
Switch 24 is responsive to channel change requests received over a return, upstream channel from the subscribers 26 as these subscribers make their programming selections. When a selection is made, switch 24 directs the selected program to the subscriber and informs the subscriber of the tuning particulars (channel, frequency, encoding information, IP address, and so forth) required to receive the selected program.
To avoid a delay associated with compression during channel program switches, additional encoders 12′ and 14′ each dedicated to the same programming or data streams as encoders 12 and 14 are provided. It will be recalled that in the MPEG2 compression standard, three types of frames are used. The I-frames are intra-coded—that is, they can be reconstructed without any reference to other frames. The price for this is that they are typically the most information-intensive of the frames and accordingly consume the highest transmission resources. The P-frames are forward-predicted from the last I-frame or P-frame—that is, it is impossible to reconstruct them without the data of another frame (I or P). The B-frames are both forward-predicted and backward-predicted from the last/next I-frame or P-frame—that is, there are two other frames necessary to reconstruct them. P- and B-frames are less dense than I-frames and consume less transmission resources. In this encoding scheme, a typical interval between I frames during playback or decoding is from about 0.5 to 2 seconds or more. I-frames are also referred to as anchor frames and are used as the point of reference here because it is from the I-frames that a newly-tuned channel is initially constructed for viewing, and a subscriber who tunes to the channel must await an I-frame to begin viewing fully reconstructed pictures. Accordingly, in the worst case scenario in which the viewer has tuned to a channel just after the last frame was sent from the DSLAM, the viewer must wait as long as 2 seconds or more before the requested channel makes a fully reconstructed visual appearance, for example on a television set. This 2-second delay can be reduced by employing a less efficient compression structure. For instance, rather than transmitting an I frame every fifteen or so frames, the I-frame transmission rate can be increased so that every other transmitted frame can be made an I-frame. This either increases the amount of information that must be transmitted, or reduces the quality of the compressed information, but it also reduces the delay between I-frames and the consequent reconstruction delay when a channel is newly-tuned to. Programming or data streams subjected to this less efficient compression structure will for convenience be herein referred to as being partially-compressed, to distinguish them from fully-compressed data streams in accordance with standard MPEG2 compression. Of course the invention is not limited to MPEG2 compression, as it is contemplated that it is applicable to other types of compression, such as standard MPEG or MPEG4/AVC. The equivalent of an MPEG2 I-frame in MPEG4/AVC is an IDR (instantaneous decoder refresh) frame. There is a guaranty that the frames following an IDR will not be predicted from frames transmitted before the IDR. So for fast channel change the secondary—that is, partially compressed—stream will be encoded mostly with IDRs and transmitted until the switch can happen to the primary—that is, fully compressed—stream when an IDR is detected.
Returning to
Based on the above description, fast-channel change operation in accordance with the invention can be described with reference to the following procedure. As mentioned above, from the switch 24 to the subscribers 26, there is a unique, or unicast connection or stream per subscriber, which in
The above are exemplary modes of carrying out the invention and are not intended to be limiting. It will be apparent to those of ordinary skill in the art that modifications thereto can be made without departure from the spirit and scope of the invention as set forth in the following claims.
