Multiplex communications – Data flow congestion prevention or control
Reexamination Certificate
1998-06-05
2001-07-24
Chin, Wellington (Department: 2664)
Multiplex communications
Data flow congestion prevention or control
C370S252000, C379S111000
Reexamination Certificate
active
06266322
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method for dimensioning link bandwidth for elastic-data traffic. More particularly, the invention relates to the dimensioning of bandwidth for elastic-data traffic in a high-speed communications network in which the number of connections to be supported on a particular link is determined. Moreover, given that the bandwidth of the link is determined, the invention relates to a method for determining the maximum number of connections that can be supported on the link.
2. Description of the Related Art
The use of data networks for the reliable high speed transport of information, such as text, voice, video, etc., over wide areas in digital format via a variety of media including optical fibers is bringing major changes to network services and network architecture/infrastructure design. Many new services having diverse characteristics and bandwidth requirements are now feasible.
FIG. 1
illustrates a small portion of a typical high speed network
10
, comprising switching nodes
11
which are connected by bandwidth data links
12
. Inputs to the network
10
are received by a plurality of user end-systems
13
which are connected to network
10
by corresponding links
14
. The inputs may be of any form, but are conveniently formatted at the user end-systems
13
into packets for transmission to other user end-systems
13
. Data is transmitted between user end-systems
13
as elastic-data traffic. A link
12
,
14
may be the entire bandwidth available to a transmission path between network nodes or between a user end-system
13
and a network node
11
, or may be a portion of the bandwidth of the transmission path. Inherent in the transfer of elastic-data from one user end-system
13
to another is that the rate of transfer is directly dependent on the constraints imposed by the network
10
as well as the constraints imposed by equipment employed at the user end-system
13
. For example, some factors which may constrain the transport of data include the protocols used to transmit the data, the amount of congestion in a given link
12
,
14
, the bandwidth of the link
12
,
14
and how many connections are simultaneously being transported over the link
12
,
14
.
In designing a high speed network, an important step is the dimensioning or capacity assignment of each link within the network. The dimensioning step typically occurs once the location of the nodes (e.g., the city in which a particular node will reside) and the connectivity between nodes (e.g., which nodes will be interconnected) is determined.
Elastic-data traffic adapts to time-varying available bandwidth over a communications link using a feedback control scheme. Examples of such schemes are the Transmission Control Protocol (TCP) over the Internet or an Available Bit Rate (ABR) transfer capability in Asynchronous Transfer Mode (ATM) networks. Typical elastic-data applications are the transfer of files supporting e-mail or the transfer of files over the world wide web. A file is a digital object that encodes information which, for example, may be a document, a computer program, data-base records, or an audio or video recording. A file resides in some memory or storage device associated with the user's end-system. The problem of dimensioning bandwidth for elastic-data applications can be viewed as a variation of the well-known “capacity assignment problem” described in the literature on design of computer networks. For example, see D. Bertsekas et al., Data Networks 2nd Edition, Prentice Hall, Englewood Cliffs, N.J., 1992.
Prior art approaches to the capacity assignment problem calculate load based on a flow of packets whose characteristics are assumed to be exogenous (independent) of the state of the network including the presence of other flows. The problem also considers a performance criterion in the form of delay of packets. The outcome of the capacity assignment problem is flawed in that the calculations do not take into consideration the dependent character of the packet flows of elastic data traffic, or that the performance criterion of most interest to the user is the transfer delay of the entire file.
Dimensioning bandwidth for elastic-data traffic can also be compared to conventional traffic dimensioning for telephone networks. Conventional dimensioning of telephone circuits uses the well-known Erlang blocking formula, while a generalized Erlang blocking model is used for multi-rate circuits. Recursive solutions associated with conventional dimensioning of telephone circuits are disclosed by J. S. Kaufman, Blocking in a shared resource environment, IEEE Transactions on Communications, Vol. 29, 1981, pp. 1474-1481; and by J. W. Roberts, G. Pujolle (ed.), A service system with heterogenous user requirements, Performance of data communications systems and their applications, North Holland, 1981, pp. 432-431. Asymptotic approximations for conventionally dimensioning telephone circuits are disclosed by Y. Kogan and Michael Shenfild, “Asymptotic solutions of generalized multiclass Engset model,” in
The Fundamental Role Teletraffic in the Evolution of Telecommunications Networks, Proc. of
14
th
Inter. Teletraffic Congress,
J. Labetoule, J. W. Roberts Eds., Elsevier, 1994, pp. 1239-1249, and by D. Mitra et al., Erlang capacity and uniform approximations for shared unbuffered resources, IEEE/ACM Transactions on Networking, Vol. 2, 1994, pp. 581-587. Both recursive and asymptotic approaches for dimensioning telephone networks assume constant-rate connections, which is natural for traditional circuits.
Recently, the concept of effective bandwidth has extended the applicability of the conventional techniques used for dimensioning telephone circuits to variable-rate connections by using the concept of a constant “effective” rate. For recent summaries of conventional techniques using constant effective rate concepts, see, for example, C.-S. Chang et al., Effective bandwidth in high-speed digital networks, IEEE Journal on Selected Areas of Communications, Vol. 13, August 1995, pp. 1091-1100; G. de Veciana et al., Resource management in wide-area ATM networks using effective bandwidths, IEEE Journal on Selected Areas of Communications, Vol. 13, August 1995, pp. 1081-1090; and, F. P. Kelly, Notes on effective bandwidths, Stochastic Networks, Claredon Press, Oxford, 1996, pp. 141-168.
The concept of assigning an effective bandwidth is reasonable for specific parameter regions of classes of traffic, such as packet voice, packet video, frame relay, and Statistical Bit Rate (SBR) service in ATM networks. Nevertheless, elastic-data traffic does not have an inherent transmission rate because the traffic adapts to available bandwidth. Consequently, the concept of an effective bandwidth for elastic-data traffic seems dubious. Further, the conventional performance criterion of a blocking probability of a new connection request is irrelevant for best-effort type services if, generally speaking, all connection requests are granted, or if a request is not even made prior to the transmission of the user data.
SUMMARY OF THE INVENTION
The present invention provides a technique for dimensioning link bandwidth for elastic-data traffic in a communications network, such as an ATM network or a network that uses the Transmission Control Protocol for sending the elastic-data traffic through the network. The advantages of the present invention are provided by a method for dimensioning link bandwidth for elastic-data traffic for a link in a communications network, in which a number of connections N having elastic-data traffic flowing over a link in a communications network is determined. The term “connection”, as used herein, applies to either an ATM connection, or more generally a virtual circuit in a connection-orientated packet network or, with some blurring of meaning, to an IP flow of packets. A transfer rate u for each connection is determined for the condition that the communications network is not a constraining resource for the connection
Berger Arthur W.
Kogan Yaakov
AT&T Corp.
Chin Wellington
Jones Premell
LandOfFree
Dimensioning bandwidth and connection admission control for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Dimensioning bandwidth and connection admission control for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Dimensioning bandwidth and connection admission control for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2524496