Multiplex communications – Pathfinding or routing – Switching a message which includes an address header
Reexamination Certificate
1997-10-23
2002-06-04
Yao, Kwang Bin (Department: 2664)
Multiplex communications
Pathfinding or routing
Switching a message which includes an address header
C370S529000, C370S537000
Reexamination Certificate
active
06400723
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of Invention
The invention relates to communication technology, and particularly to efficient transmission of information over networks that service both voice and data traffic.
2. Description of Related Art
Integrated multiplexing schemes attempt to optimize the use of instantaneous transmission bandwidth available in networks for different kinds of data streams. In the case of communication over telephone networks, one prevailing policy is to allocate the transmission capacity of the Public Switched Telephone Network (PSTN) to various traffic types, according to instantaneous needs of originating sources and data destinations.
Integrated transmission on PSTN networks is expected to handle increasing traffic of various types, including voice, facsimile, voiceband data, digital data and video services and others. In particular, when the PSTN interfaces with the Internet a typical network node in this type of communications environment receives packets of different traffic types from various terminals. The traffic is stored according to the order of arrival, and then transmitted to desired destinations using some transmission policy. Because the switching and transmission resources of the networks are shared among the various types of traffic, efficient use of the channel bandwidth becomes important to meet performance requirements (e.g., packet loss, delay, etc.) of differing traffic types. For example, voice is more tolerant to the loss of information than data, but is intolerant of substantial delay or jitter.
There have been many approaches in the art to multiplexing heterogeneous traffic types, some of which are currently or have been used in commercial systems. Existing methods can be classified according to several types.
These types can be divided according to the type of traffic to be multiplexed (voice and voiceband data, voice and facsimile, etc.), the access method (i.e., synchronous or asynchronous multiplexing), and the bandwidth over which the multiplexing is taking place. For example, digital types include narrowband ISDN at 64 kbit/s channel, wideband ISDN using several 64 kbit/s channels, primary rate (1544 kbit/s or 2048 kbit/s) channels, broadband channels (at ATM bit rates of 100 Mbit/s or above) or sub-rate channels, i.e., <64 kbit/s channel.
In terms of emerging digital technology, conventional multiplexing schemes for narrowband and wideband ISDN fall into one of the three following classes.
1. Fixed Boundary Multiplexing Scheme
This ISDN scheme is based on synchronous time-division multiplexing (TDM), where the TDM frame consists of N slots each b bits wide. N
1
of the N channels are allocated to voice, and the remaining (N
2
=N−N
1
) to data. The value of N
1
is chosen according to the voice bit rate and the duration of the TDM frame. However, this scheme suffers from large time delay and blocking probabilities. The problem is compounded when one type of traffic is temporarily absent; the corresponding allocated slots will not be used, even if the other type of traffic is delayed or blocked. For example, excess voice packets are blocked (or dropped) even if slots allocated for data packets are available (e.g., when there are no data packets), degrading voice quality.
2. Movable Boundary Multiplexing Scheme
In this scheme, voice and data traffic still share channel capacity on the basis of a synchronous TDM scheme. Here, voice traffic receives priority over data traffic, but when there is no voice traffic the transmission bandwidth is used exclusively for data traffic. Thus, data packets may occupy any of the N
1
slots temporarily not used, while voice traffic pre-empts data traffic and occupies one of its allocated slots if necessary to receive service. This scheme reduces the average queuing delay for data, but does not improve blocking performance for voice. This scheme also cannot be relied upon to increase data throughput because extraordinarily long data queues may result. Flow control of data traffic is thus required to ensure that the data traffic load is kept within reasonable bounds. The scheme also does not take into account the fact that silent periods constitute a significant amount (60%) of the time that a person speaks. These factors have been taken into account in the modified movable boundary schemes of the following digital circuit multiplication equipment (DCME).
3. DCME Schemes
Several proprietary DCME multiplexing schemes are in commercial use. To facilitate networking, a standard scheme has been developed in the various versions of industry standard ITU-T (formerly CCITT) Recommendation G.763. The idea is to combine talk-burst with various voice-encoding schemes using lower bit rates than the traditional 64 kbit/s PCM for voice. In addition, DCME demodulates facsimile traffic to transport the baseband signal instead of a 64 kbit/s stream of the digitized modulated baseband signal. These features have been introduced to balance high channel multiplication ratios with high quality voice and data transmission, as known in the art.
DCME accomplishes this by generating bearer channels consisting of full-time four-bits/sample for 32 kbit/s ADPCM derived from the 64 kbit/s time slots. 24 kbit/s and 16 kbit/s overload channels are created from the bearer channels whenever the demand for voice service exceeds the number of available channels. This presumes use of variable bit rate coding for voice, so that voice channels can be coded with less bits per sample. The reduction in bits per sample is spread among all voice channels on a pseudo-random basis. For voiceband transmission, data transmission up to 9600 bit/s can be supported by five-bits/sample transmission operation. Higher bit rates and digital data transmission use 8-bits/sample transmission operations for “clear” or “transparent” 64 kbit/s operation. Thus, the bearer channels are divided into pre-assigned bearer channels for data and voiceband data operation, voice bearer channels, facsimile channels and overload channels. The pre-assigned channels are fixed. The boundaries between the other channels vary according to the traffic mix and the desired quality. For example, to minimize “freeze-out” of speech (clipping that occurs when the number of talk-bursts exceed the transmission capacity), the requests for assignment to servers are placed in an assignment queue. Whenever the load increases beyond a given threshold, the controller increases the number of available overload channels to serve the additional load, and improves utilization of the available bandwidth. Finally, when the dynamic load reaches a given threshold, DCME signals to the switch that no more calls should be accepted. The configuration data of a DCME frame includes all the information necessary to define the structure of the transmit and receive bearers.
The main differences between the various schemes just described reside in how transmission bandwidth is divided, and how incoming channels are mapped to the bearer channels. Because DCME schemes are based on time-division multiplexing, they can be efficient in their use of transmission bandwidth if the traffic is predominantly voice, voiceband data at rates of 14,400 bits/s, 9600 bits/s or less, or facsimile. However, because voice traffic is not allowed to use the idle time slots of the pre-assigned channels, it could result in bandwidth inefficiency when data traffic is low. Although DCME can be adapted to multi-point applications, this requires extensive coordination among the various destinations.
4. (T
1
, T
2
) Techniques
A further implementation of an integrated multiplexing scheme known in the art is (T
1
, T
2
) multiplexing. The basic idea of the (T
1
, T
2
) scheme is to share the available transmission bandwidth between voice and data on a statistical basis. The objective is to make efficient use of transmission bandwidth while meeting the performance requirements of three types of traffic: signaling, voiceband and digital data traffic. This makes the scheme well adapted to packet networks. Sig
Burg Frederick Murray
Sherif Mostafa Hashem
Tewani Kamlesh T.
AT&T Corp.
Yao Kwang Bin
LandOfFree
Multiplexing system and method for integrated voice and data... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Multiplexing system and method for integrated voice and data..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Multiplexing system and method for integrated voice and data... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2981362