Multiplex communications – Diagnostic testing – Determination of communication parameters
Reexamination Certificate
1999-03-15
2003-05-13
Ton, Dang (Department: 2666)
Multiplex communications
Diagnostic testing
Determination of communication parameters
C370S389000
Reexamination Certificate
active
06563796
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method of traffic quality evaluation and traffic measurements in packet network and asynchronous transfer mode network. Traffic quality evaluation is utilized primarily for traffic design and confirmation of quality of service evaluation after the system has been constructed. Traffic quality evaluation apparatus observes packets that have been duplicated from the packets passing an attention point between switching systems, and determines arrival time of packet itself or a specific cell contained in the packet, thereby estimating the arrival time of packet or cells contained in the packet as well as loss ratio of cells.
This application is based on patent application No. Hei 10-68772, No. Hei 10-264769 and No. Hei 10-267076 filed in Japan, the contents of which are incorporated herein by reference.
2. Description of the Related Art
In information communication systems, variations in the quality of service such as delays and data loss are produced depending on the traffic intensity. Therefore, to construct an economic and high performance information communication system, quantitative understanding of the relationship between the traffic intensity and the quality performance under a given set of conditions of operating environment, such as buffer size and circuit speed, CPU power, memory and the like, become essential so that an appropriate size of buffer memories, for example, can be provided for optimum performance of the network.
To achieve this end objective, two basic methodologies have been used. The first method is based on simulation. A logic to simulate a target system is assembled into a computer. The computer is then input with statistically processed results obtained from traffic measurement in the actual network system, and traffic based on the results of processing is simulated by generating random numbers, and the resulting quality for the generated traffic is evaluated.
The second method is based on evaluating a mathematical relationship between quality and operating environment such as traffic buffer size obtained by traffic theory. Traffic conditions are derived from statistical processing of traffic measurement in the information network system, as in the first method, or the person evaluating the system selects suitable traffic conditions under certain assumptions. In either method, the evaluation approach is based summarizing the measured results once into a small number of statistics, and analyzing the summarized statistics.
However, these conventional methods present the following problems in providing a quality of service (QoS) evaluation to perform accurate traffic design for a given network.
(1) Traffic measurements in information communication systems lose a lot of information related to the traffic, because of many restrictions in measurement items pre-provided in the system, such as measurement duration.
(2) Much information is lost by statistically summarizing the measured traffic results into a small number of prearranged statistical parameters to fit certain probability/statistical model. Actual traffic activities are far more complex than those that can be described by a few parameters of probability/statistical models. Consequently, errors introduced by modeling are such that they cannot be ignored.
Furthermore, traffic measurement techniques are important in contributing significantly to design an economical highspeed network. Methodology for measuring network traffic can be divided into two broad categories: one utilizes devices within the network such as end systems, switches and routers; while the other utilizes external measuring devices. In either case, items to be measured are determined by the devices or softwares provided in the switches and externally attached traffic measuring devices. It is true that some flexibility in choice can be provided when measuring the traffic by application programs in end systems, because they allow modifications to be made to the programs, In such cases, however, it is important to remember that what is being measured is not the traffic within the network, but the traffic that the relevant end system is transmitting and receiving.
Internet and ATM networks are realizing not only monolithic tasks such as voice transmission but a variety of information transmission tasks. Thus, to achieve an effective use of the network capability, it is necessary that the network operators come to clear understanding of the variety of traffic conditions that may exist in the network. However, because the existing measurement devices can provide only fixed traffic items for measurements, it is difficult to perform traffic measurements within a network in a variety of conditions, without exhaustively seeking out all combinations of a variety of traffic measurement conditions. However, it is impossible in practice to respond to all such combinations, considering the limited nature of the processor capability and memory capacity in a given measuring system in existing communication networks.
SUMMARY OF THE INVENTION
The present invention related to the quality of service evaluation has been developed in consideration of the background described above, and an object is to provide an evaluation apparatus to enable traffic measurements without loss of traffic information. Also, another object of the present invention is to provide an evaluation apparatus that enables to evaluate the quality of service without being affected by the modeling.
The object related to the quality of service evaluation has been achieved by the following features disclosed in the present invention.
(1) Raw traffic data are obtained by inserting a splitting device or by using accessing history. Traffic measurement items provided in a communication system are not used.
(2) Raw data obtained are directly input into a simulation model without summarizing into pre-arranged statistical parameters. As a result, near real-time traffic is reproduced. To perform such a simulation continually, the measuring point and the execution section for the simulation logic are connected on-line.
(3) QoS evaluation is performed by simulation and numerical equations are not used.
Specifically, the present QoS evaluation apparatus is provided with means for splitting and accepting the packets passing through an attention point in a switching system, and evaluates the quality of service dependent on by observing the packets splitted off from the network.
The features provided for the present quality of service evaluation apparatus are:
means for accepting a packet passing through an attention point in a switching system and duplicated and splitted at a splitter, and assigning a timestamp to a specific processing unit included in said packet;
means for estimating arrival times of processing units included in said packet at said attention point on a basis of said timestamp for said specific processing unit given by said arrival detection section; and
means for simulating arrival time of said packet and processing discipline at said attention point which arrival time to correspond to an estimated arrival time of said packet at said attention point obtained by said estimating section, according to a pre-determined simulation logic.
The switching system is an ATM switching system, and said attention point may be an output-line buffer in said ATM switching system, then said processing unit is a cell and said specific processing unit can be the first cell in an object packet. Or, said processing unit may be a cell, and said specific processing can be an individual cell in an object packet.
In this case, the estimating section may include means for estimating an arrival time of an (i)th cell in said packet according to a relation:
[a timestamp of said first cell+(i−1)×(cell length/input-line speed)], or
an estimated arrival time of an (i)th cell comprising said packet may be calculated from a relation:
[an estimated arrival time of said first cell+(i−1)&t
Burns Doane Swecker & Mathis L.L.P.
Nippon Telegraph and Telephone Corporation
Ton Dang
LandOfFree
Apparatus for quality of service evaluation and traffic... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus for quality of service evaluation and traffic..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus for quality of service evaluation and traffic... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3055073