Multiplex communications – Data flow congestion prevention or control – Control of data admission to the network
Reexamination Certificate
1999-10-15
2003-05-27
Hsu, Alpus H. (Department: 2665)
Multiplex communications
Data flow congestion prevention or control
Control of data admission to the network
C370S352000, C370S395410, C370S412000, C370S429000, C370S471000, C370S477000
Reexamination Certificate
active
06570849
ABSTRACT:
FIELD OF INVENTION
This invention deals with the interleaving of data streams of differing priorities in such a manner that real time data can be delivered on a managed packet-based network in real time, while non-real time data is delivered in the bandwidth remaining after the real time data has been transmitted.
BACKGROUND
Historically, entire separate communications systems have been employed to transmit audio data, traditionally referred to as “voice”, and computer data, conveniently referred to as “data”, although at present audio data and computer data both are considered to fall within the broad definition of “data”. Over a century ago analogue telephone networks were developed to carry analogue audio signals. Telephone networks allow communication of audio signals between two or more users by establishing, with central switching equipment, a temporary dedicated communication circuit or channel between the users. Because the channel, once established, is dedicated exclusively to the transmission of signals between the users, the signals are not required to compete for the channel's bandwidth with other unrelated signals. The advantage of having a dedicated channel for a voice conversation is that any transmission delay from speaker to listener is purely a function of the unfettered speed of the audio signal through the telephone network. Since this speed does not significantly vary over time, such dedicated channels are capable of providing isochronous transmission. Unfortunately, one significant disadvantage of dedicated channels is that they require significant bandwidth; that is, the complete bandwidth of the channel remains available and dedicated to carrying the conversation even when no information is being transmitted, such as during either conversational or inter-syllable pauses. To allow better utilisation of a given communication link, time-division multiplexing (TDM) systems, where voice signals from a number of users are digitised and then the resulting bits are time compressed for transmission over the same link. This scheme also provides isochronous transmission while sharing the communication link.
In contrast, more recent packet transmission systems are asynchronous, allowing the allocation of bandwidth on an as-demanded basis. For instance, if a channel is highly active, it may receive more than its pro-rata share of overall bandwidth. When a channel's activities decline, its allocated bandwidth likewise declines. Thus, packet transmission is adept at handling “bursty” transmission of data, wherein the activity of each individual channel is subject to relatively wide variation.
All networks experience certain delays in end-to-end data transmission therethrough, this delay (termed “latency”) affects the overall efficiency and effective bandwidth of the network. Packet-based computer networks, because they are asynchronous, are further subjected to “jitter”, defined as a change in the latency of the network as a function of time. Jitter is largely unpredictable; however, the overall quantity of traffic on a network tends to increase both latency and jitter. Jitter can only be corrected by ensuring constant latency, or by developing methods to compensate for its effects. The most common of these methods is the use of a “jitter-buffer” to store incoming data, thus hiding jittery transmissions from the listener.
Given the ongoing rapid growth of data traffic compared with the slow growth of voice traffic, it is foreseeable that the volume of data traffic will outstrip that of voice on public carrier networks. At that point, the existing Public Switched Telephone Network (PSTN), whose technology is primarily based on circuit switching, will be used predominantly to carry data for which a packet switching technology is a more appropriate alternative. Packet-switched networks can carry voice calls using as little as 8 kilobits per second (Kbps) of bandwidth to provide TDM quality voice compared to the 64 Kbps that is reserved for each call in the conventional PSTN. The potential increase in capacity is attractive to telecommunications service providers, as are the reduced infrastructure costs associated with building packet-switched networks. As a result, providers of new real time telephony services are increasingly using packet- switched network architectures, a trend that will continue with the arrival of competition in the local phone market. The predominant packet-based network protocol in use today is the ubiquitous Internet Protocol (IP). The first attempts to use this technology were implemented by hobbyists on the Internet. In these early methods, an Internet telephony software program would convert the user's analogue voice signal to digital data, compress it, and transmit it as packets to an Internet server. For example, a full duplex real time telephone conversation may be initiated over the Internet from one computer to another. Both computers generally require the same Internet telephony software, a microphone, a sound card, a minimum processor speed, a modem, specific software, and a connection to an Internet Service Provider. Both computers must be on-line simultaneously; which requires either a pre-arrangement between callers or a separate telephone toll call. Software for that type of system is commercially available.
Data packets traverse packet based networks by being routed from one node to the next. Each of these hops takes the packet closer to its destination. Each node along the route is designated by a globally unique address. Each node in the route looks at the destination address contained in the header of a packet and sends the packet in the direction of its destination. At any time, a node along a particular route can stop accepting, or block one or more packets. This may be due to any number of reasons: congestion, maintenance, node crashes, etc. Each routing node constantly monitors its adjacent nodes and adjusts its routing table when problems occur. As a result, sequential packets often take different routes as they traverse the Internet. The audio quality of duplex phone conversation over the Internet is often poor because of delays of transmission of packets, lost packets, and lost connections. The delays are unpredictable and are usually caused by the dynamically changing conditions of the network and the changing and often long routes through which packets must pass to arrive at their destination. One consequence of these delays and differing routes the packets follow, other than message lag, is that individual voice packets may arrive in a non-sequential order.
Existing methods for reducing delay-related problems include assorted error correction schemes, primarily repeat codes, which transmit data packets multiple times, and parity codes, which use modulo arithmetic to achieve a fixed value over a series of cells in a check accumulator. These techniques are well known in the art, as are their drawbacks. The use of repeat codes is undesirable because of the network resources wasted by sending multiple copies of a single packet. Parity codes are hampered by implementation problems that make both source encoding and destination decoding complicated. A simple “row-parity” scheme can be employed easily by placing parity packets after a set number of packets, but is flawed because it can only be used to correct an error if the receiver knows that all but one packet has been received correctly. More complex schemes, such as the one outlined in U.S. Pat. No. 5,883,891 use a “matrix-parity” concept which seeks to improve the audio quality of voice communication over the Internet by reconstituting delayed and/or missing packets based upon the packets that arrive in time. The system is “robust”, in comparison to simpler row-parity schemes, because packets constituting the matrix are deliberately transmitted over multiple routes. If one route is subject to delays, or packet loss, the lost or delayed packets can be fully reconstituted in many cases. The system receiving the voiceband signal from a caller (host system) u
Aly Sami
Brown Robert W.
Skemer Terry
Donnelly Victoria
Hsu Alpus H.
Kinsman Anne
Tropic Networks Inc.
LandOfFree
TDM-quality voice over packet does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with TDM-quality voice over packet, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and TDM-quality voice over packet will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3079591