Electrical computers and digital processing systems: multicomput – Computer-to-computer data routing
Reexamination Certificate
1999-09-28
2004-08-17
Wiley, David (Department: 2143)
Electrical computers and digital processing systems: multicomput
Computer-to-computer data routing
C709S223000, C702S002000, C702S179000, C703S002000, C370S902000, C370S230000
Reexamination Certificate
active
06779037
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to the field of the communications, preferably aggregate multimedia traffic, and more particularly the invention refers to a method for obtaining optimum utilization of a shared transmission medium for multimedia traffic, complying with customer quality standards for the traffic sessions. More particularly the invention refers to a technique for estimating the transmission rate necessary to satisfy the quality of service (QoS) requirements, in a plurality of real-time variable bit rate (rt-VBR) sessions, thus preventing the transmission medium capacity from being underestimated or overestimated this prevention resulting in an optimum cost rating of the transmission service.
2. Description of the Prior Art
A general multimedia transmission is established between a sender and a recipient and is made through a shared transmission medium, such as a communication network, more particularly in a ATM network having a predetermined transmission rate capacity or bandwidth for transmitting the ATM cells or data packets. It may be however that the bandwidth available for transmission is not enough to new calls or requests for connection, and such request exceeding the capacity of the network are rejected or queued until bandwidth availability is re-established. This availability, however, is estimated on the basis of real-time measurements or based on previously stored data transmission rate information. These measurements involves the calls in progress in the network, the new calls, requests for connection, frequency and duration of data bursts, data peaks, sojourn times, average usage, etc. In any event, the measurements of such parameters are complex, cumbersome and time consuming to make, even with the most recent and developed computer technology.
The data transmissions involving interactive voice conversations and videoconferences through modern networks must comply with the Quality of Service (QoS) requirements and the availability of the network must be constantly evaluated to comply with such requirements. Thus, a new rt-VBR session should be accepted provided that the network does not fail to comply with the QoS requirements. The QoS requirements can be defined by the Cell Lost Ratio (CLR), i.e. it is related to the probability of loosing a data unit or cell within a unit time or interval, defined to evaluate the traffic. The CLR is the ratio between the number of emitted cells according to contract rules that do not reach the receptor in time vs. the total number of cells emitted according to the contract rules; this is normally estimated in time periods that are larger than the duration of a call or connection. The QoS is also determined by the maximum time between the output of the signal from the transmitter and the reception thereof at the receptor, this being called Maximum Cell Transfer Delay or maxCTD. The services with CLR=0 are called deterministic services and the services with CLR>0 are called predictive services.
In agreement to ITU-T G.114, when the maximum end-to-end delay imposed by the application is less than 400 msec and the traffic type generated by the coder is variable bit rate (VBR), the application is called rt-VBR. Re-transmissions are not accepted by applications rt-VBR to correct transmissions errors or cell looses because the re-transmitted cell delay generally exceeds the maxCTD.
The total delay since a signal is captured by the transmitter until is reproduced at the receptor is the result of several delay components typically summarized as follows:
maxCTD>Codification+Propagation+Queuing
The maxCTD required by sessions imposes a maximum to total delay.
Queuing delay is the delay that a cell has in the queue of the system, is a random variable T depending of the service discipline, of A (i.e. quantity and type of traffic) and of r (transmission rate or speed reserved for the traffic A in the channel) . The quantity of cells in the queue is a random variable Q, and the above variables, in their form for discrete time, may be related with a recursive formula by the equation:
Q[n
]=(
Q[n
−1
]+A[n]−r)
+
(1)
where operator <X>
+
=max(X,O), and the queuing delay may be calculated by
T[k]=Q[n]/r
wherein:
n number of interval.
A[n]=amount of data units arrived during interval n.
Q[n]=amount of data units in queue during interval n.
r=transmission rate reserved for traffic A.
T[n]=queuing delay experimented by data units arrived in interval n.
k and n are the same if T is evaluated for A[n]>=0, and k<=n when T is evaluated only when A[n]>0.
The channel utilization is defined as
u
=
E
{
A
}
r
,
where E{A} is the traffic average, and E{A}<r must hold for system stability.
Generally, there are no analytical solutions for equation (1) because traffic A is an unknown stochastic process, and the conventional solutions are based in the construction of mathematical models of the traffic which do not reflect the exact behavior of the traffic.
There are several processes for modeling the traffic such as the Markov Modulated Poisson Process (MMPP) which is used for solving the queuing process. This process however requires of an extremely large quantity of samples that cause the processes to be no practical.
Another drawback is the estimation of the process parameters, wherein the mean bite rate (MBR) not only depends on the algorithm coding the session but also on the participants in a conversation and the type of conversation. There are predictive methods characterizing the traffic with a Probabilistic Burstiness Curve (PBC) using excessive quantity of actual samples.
U.S. Pat. No. 5,886,907 to Abu-Amara et al. discloses a method and system to model aggregate multimedia traffic in an ATM network with a purpose of constructing and utilizing a network. Different traffic services are specified and quality for each service is determined from available standards. Sojourn times are determined on the basis of measurements or published standards. The '907 patent is based in the determination of values for average and mean sojourn times, usage rate for each type of service for modeling the traffic.
U.S. Pat. No. 5,583,792 to San-Qi Li discloses a technique for the integration of traffic measurement and queue analysis comprising basically traffic measurement, statistical matching and queuing analysis. The traffic measurement is made by standard signal processing techniques, the statistical matching is made by constructing a Markov chain modulated rate process that can statistically match with each given traffic stream, and the queuing analysis is made by a folding-algorithm to determine a design of the network.
U.S. Pat. No. 5,604,731 to M. Grossglauser discloses a method of renegotiating transmission rates between a sender and a recipient in a renegotiated variable bit-rate (RVBR) network or a renegotiated constant bit-rate (RCBR) network, based on previously stored data transmission rate information. The method needs to count on previously stored information or new information about the real-time transmission must be obtained.
U.S. Pat. No. 5,757,771 to Kwok-Leung, Li discloses a method of processing data packets received from a data input port, temporarily stored in a buffer memory and transmitting the data packets to an output port, wherein the buffer memory is divided into a plurality of data sub-queues having assigned output rankings, an accumulation ratio threshold being calculated for each data sub-queue, and said transmission of data packets to the output port is made in accordance with said accumulation ratio threshold and said output ranking.
U.S. Pat. No. 5,774,455 to Fumiyoshi Kawase discloses a digital data transmission apparatus including a digital data buffer, a digital data transmitter and a controlle
Boutah Alina
Wiley David
LandOfFree
Method of obtaining optimum use of a shared transmission... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of obtaining optimum use of a shared transmission..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of obtaining optimum use of a shared transmission... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3327002