Multiplex communications – Data flow congestion prevention or control – Control of data admission to the network
Reexamination Certificate
1997-04-11
2003-06-24
Hsu, Alpus H. (Department: 2665)
Multiplex communications
Data flow congestion prevention or control
Control of data admission to the network
C370S253000, C370S395100, C370S474000
Reexamination Certificate
active
06584070
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to an asynchronous transfer mode (ATM) transmission apparatus for multiplexing coded audio signals into a cell for transmission over a transmission network in an ATM mode.
Research is progressing on the so-called ISDN (integrated services digital network). This is a branch of technologies for concurrently transmitting over a single network multiple pieces of information that have different characteristics, such as audio information and dynamic image information. Drawing attention in this area presently is asynchronous transfer mode (ATM), a switching technique indispensable for implementing a broadened ISDN (B-ISDN). This technique involves dividing communication information into packets called cells of a fixed length for transmission.
The most commonly utilized method today for coding telephone voice signals in digital format is pulse code modulation (PCM) at a transmission rate of 64 kilobits per second. Where it is desired to lower the transmission rate (also known as the bit rate) without degrading the quality of voice transmitted, one known method employed is ADPCM (adaptive differential pulse code modulation) at a transmission rate of 32 kilobits per second.
About to be put into practice is what is known as low delay code excited linear prediction (LD-CELP:CCITT G728). This is a method for converting every five values sampled at 8 kHz into a predetermined code of 10 bits, whereby a transmission rate of 16 kilobits per second is provided.
Where voice signals are transmitted as communication information, the quality of voice sound deteriorates if the transmission delay time involved is prolonged. Thus there are strict limits as to how long the transmission delay time is allowed to be.
Described below is a typical setup of the above-mentioned ATM transmission using voice signals.
FIG. 2
is a block diagram of a typical prior art ATM transmission apparatus, and
FIG. 3
is a view showing the constitution of an ATM cell used by the conventional apparatus of FIG.
2
. In
FIG. 2
, an exchange
1
accommodates subscriber lines from a plurality of subscriber terminals
2
and is connected to the ATM transmission apparatus
4
via a plurality of channels
5
.
Suppose that one of the subscriber terminals
2
(i.e., calling subscriber) makes a call to communicate with another subscriber terminal (i.e., called subscriber) via an ATM transmission line
3
. In that case, the exchange
1
first connects the terminal
2
of the calling subscriber to the ATM transmission apparatus
4
over a given channel
5
.
In turn, the ATM transmission apparatus
4
converts into a predetermined digital code (called coded information) the voice signal transmitted from the subscriber terminal
2
(calling subscriber) through the exchange
1
and channel
5
. The ATM transmission apparatus
4
then generates an ATM cell
10
, multiplexes it with another cell made of the voice signal from the subscriber terminal
2
, and transmits the multiplexed result over the ATM transmission line
3
.
As depicted in
FIG. 3
, cells generated by the ATM transmission apparatus
4
are each composed of 53 octets. The first five octets constitute an ATM header
7
. The ATM header
7
includes a virtual path identifier (VPI) and a virtual channel identifier (VCI). The remaining 48 octets make up a payload
8
comprising coded information.
Of the 48 octets constituting the payload, the first octet contains a sequence number identifier (SN) and a data type identifier (IT);, the last two octets make up an effective data length identifier (LI) and a cyclic redundancy check identifier (CRC). The remaining 45 octets (i.e., 360 bits) constitute a payload user information part
11
for transmitting the coded information.
The ATM transmission apparatus
4
of
FIG. 2
is equipped for each channel
5
with a coder-decoder
41
, a code buffer
42
, a payload assembler
43
and an ATM multiplexer
13
. The channels
5
are provided commonly with a cross connection multiplexer
45
. These components work as follows:
The coder-decoder
41
digitizes a voice signal illustratively according to the LD-CELP method. The voice signal has been transmitted from a subscriber terminal
2
(calling subscriber) over a channel
5
and through the exchange
1
. The signal in digital format is stored in the code buffer
42
downstream.
The payload assembler
43
monitors the amount of coded information in the code buffer
42
. On detecting an accumulation of 36 items of coded information (i.e., 360 bits. or 45 octets) in the code buffer
42
, the payload assembler
43
gets the accumulated
36
items of coded information from the code buffer
42
and assembles them into a payload
8
. The payload
8
is then transferred to the ATM multiplexer
13
downstream.
Upon receipt of the payload
8
from the payload assembler
43
, the ATM multiplexer
13
composes a cell by adding an ATM header
7
to the payload coming from the payload assembler
43
. The cell when composed is transferred to the cross connection multiplexer
45
.
The cross connection multiplexer
45
stores temporarily in a queue (i.e., buffer) the cells transferred from the ATM multiplexers
13
upstream. The cells are then output onto the ATM transmission line
3
in the order in which they were stored into the buffer.
As described, in the prior art ATM transmission apparatus
4
, the coded information made of the voice signals coming from subscriber terminals
2
is transmitted over the ATM transmission line
3
after 36 items of the coded information are accumulated in the code buffer
42
and are assembled into a cell for transmission.
It takes 625 microseconds (&mgr;s) for the coder-decoder
41
to generate one item of coded information (i.e., 125 &mgr;d×5). That is, a delay time of 22,500 &mgr;s occurs by the time 36 items of coded information are accumulated in the code buffer
42
(i.e., 625 &mgr;s×36). This often makes it difficult to comply with the time constraints on transmission delay under the LD-CELP method. As a result, a serious adverse effect on the quality of the transmitted voice may occur.
Presently there is a possibility that in-house LAN's (local area networks), based on the DQDB (distributed queue dual bus) system proposed under IEEE (Institute of Electrical and Electronics Engineers) 802.6, will gain widespread acceptance. If that happens, the congestion of different types of communication information, which will affect transmission, can be a severe disadvantage to the system.
Packet data transmitted over the LAN's have variable lengths while ATM cells
10
have a fixed length. When communication information is divided, the divided items are multiplexed into an ATM cell
10
. If a fraction of the cell
10
constitutes the information, the remaining vacant parts are filled with dummy patterns so that the finished cell will be a complete cell. The smaller the fraction and the higher the frequency at which a fractionally complete cell occurs, the more dummy patterns are needed to fill the gap. As a result, the transmission efficiency decreases.
More and more terminals connected to in-house LAN's including those in compliance with the DQDB system will likely be multi-media terminals such as TV telephone sets and audio/visual output devices. There is little doubt that the number of available channels will not keep up with the growing number of multi-media terminals. Furthermore, if equipped with a transmitter-receiver for each different medium, the multi-media terminal will bloat in size and cost and will run counter to today's trend toward downsized terminals with compact functions.
Certain kinds of communication information such as motion pictures require synchronism between dynamic image information and audio information when transmitted. Ensuring synchronism between the different kinds of information is necessary so as to keep the received information meaningful. It may be arranged technically that each cell comprises either audio or image information alone. In that case, a
Kajiwara Masanori
Mase Hideki
Nobuyasu Kosuke
Norizuki Reiko
Takashima Tomonobu
No associations
LandOfFree
Cell multiplexing apparatus handling multiple items of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Cell multiplexing apparatus handling multiple items of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Cell multiplexing apparatus handling multiple items of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3096418