JPH02299343A - Packet transmission characteristic measurement system - Google Patents

Abstract

PURPOSE:To measure the variance or deviation of a transmission delay time distribution based on the interval of the arrival time of a received measuring packet by writing a serial number to a packet at a sender side of a transmitter. CONSTITUTION:A transmitter 1 uses a trigger signal from an equal time interval transmission trigger generating section 1-3 measuring the time interval to send the measuring packets with a serial number written therein in the order of transmission by a serial number writing section 1-2 to a receiver 2 via a measured system 3. The receiver 2 measures the arrived time with a counter 2-4 and reads the serial number of the received measurement packets with a serial number readout section 2-2 and stores the both in a storage section 2-5. The variance of the transmission delay time distribution is calculated by equation I, where ai and bi (i=1-n) are the serial number and the arrival time stored in the storage section 2-5 and tau is a trigger interval of an equal time interval transmission trigger. Moreover, the deviation is calculated as a root of the variance.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a method for measuring packet transmission characteristics of a system that handles packets. In particular, the present invention relates to a method of writing serial numbers in some of the packets used for measurement in the order of input, and measuring transfer characteristics using the serial numbers at the output.

(Prior Art) A first example of a conventional method for measuring the dispersion or deviation of transmission delay time distribution, which is one of the packet transmission characteristics in a system that handles packets, is shown in FIG.

Reference numeral 1 denotes a measurement packet transmitter (hereinafter referred to as a transmitter), 1-1 a transmitter, 1-4 a transmission time measurement clock or counter (hereinafter referred to as a transmitter clock), 1-5 a transmission time storage unit, Reference numeral 2 denotes a measuring packet receiving device (hereinafter referred to as a receiving device); 2-1 denotes a receiving unit; 2-7 denotes an arrival time measurement clock or counter (hereinafter referred to as a receiving clock) indicating the same absolute time as 1-4; Reference numeral 3 indicates the system to be measured.

The transmitter transmits only one measurement packet to the system under test, reads the transmission time from the transmission clock 1-4, and stores it in the transmission time storage section 1-5. The transmission delay time is calculated by reading the time at which the measurement packet output from the system under test is received in the receiving device from the receiving clock 2-7, and subtracting the time stored in the transmission time storage unit 1-5 from that time. is measured. The variance or deviation of the propagation delay time distribution can be determined by measuring the propagation delay time many times.
It is calculated from that data. That is, when the transmission delay time data is T, (i=1 to n), the variance of the transmission delay time distribution is calculated by equation (1). Also, the deviation is calculated as the square root of the variance.

Second method for measuring dispersion or deviation of propagation delay time distribution
An example of this is shown in FIG. Reference numeral 1-6 indicates a transmission time writing unit that writes the transmission time read from the transmission clock 1-4 into a part of the measurement packet, reference numeral 2-3 indicates a received packet counter, and reference numeral 2-8 indicates a received measurement packet. A transmission time reading section is shown for reading the transmission time written by the transmission time writing section 1-6.

The transmitter reads the packet transmission time from the transmission clock 1-4, writes the time into a measurement packet in the transmission time writing section 1-6, and transmits the packet to the system under test. The dispersion or deviation of the transmission delay time distribution is determined by reading the time at which the measurement packet output from the system under test is received in the receiving device from the reception clock 2-7, and using the transmission time reading section 2-8 to calculate the measurement packet from the reception clock 2-7. It is calculated using the transmission delay time measured by subtracting the packet transmission time read from . That is, if the count value of the received packet number counter 2-3 is n, and the measured transmission delay time data is T+ (i = 1 to n), the variance of the transmission delay time distribution is calculated by equation (1). Ru. Also, the deviation is calculated as the square root of the variance.

FIG. 5 shows an example of a conventional method for measuring the number of discarded packets, which is one of the packet transmission characteristics in a system that handles packets. Reference numerals 1-7 indicate transmission packet number counters.

The transmitting device transmits measurement packets to the system under test, and counts the number of transmitted packets using a transmitted packet number counter 1-7. The receiving device counts the number of received measurement packets using a received packet number counter 2-3. The number of discarded packets is measured by subtracting the count value of the received packet number counter 2-3 from the count value of the transmitted packet number counter 1-7.

(Problems to be Solved by the Invention) In the conventional packet transfer characteristic measurement method described above, (1)
When measuring the dispersion or deviation of the transmission delay time distribution, it is necessary to match the absolute time of the transmitting clock and receiving clock.
This has the disadvantage that it is difficult to implement, especially when the transmitting device and receiving device are located far apart. In addition, (2) there is a drawback that in measuring the number of discarded packets, it is necessary to know the values of the packet number counters of both the transmitting device and the receiving device. Furthermore, even if the input order of packets to the system under test and the order of output of packets from the system under test are the same, there is a possibility that the packets remain in the system under test, and the remaining packets may be lost after transmission is stopped. It is impossible to determine whether a packet has been discarded or remains in the system under test until after waiting enough time for all packets to reach the receiving device.
The drawback was that it was not possible to measure the exact number of packets discarded. Furthermore, there is a drawback that only the total number of packets discarded over the entire measurement time can be measured, and it is not possible to specify at what point within the measurement time the packet discard occurred.

An object of the present invention is to solve the above-mentioned problems by writing a serial number of the transmission order in a part of the measurement packet and using that number in the receiving device.

(Means for Solving the Problems) One feature of the present invention for achieving the above object is that in a method that treats information in packet units and measures the packet transfer characteristics from the input to the output of the system under test, A packet transfer characteristic that writes serial numbers in the order of input to packets input to the system, and measures at least one of the packet transmission delay time and its dispersion in the system under test according to the serial numbers and arrival times of the packets on the output side of the system under test. It's in the measurement method.

Another feature of the present invention is that in the method of measuring the packet transfer characteristics from the input to the output of the system under test, where information is handled in packet units, serial numbers are written in the packets input to the system under test in the order of input. This transfer characteristic measurement method measures the number of packets lost within the system under test by checking the serial numbers of received packets at the output side of the system.

(Function) In the present invention, serial numbers are written in part of the measurement packets in the order of transmission, and the measurement packets are transmitted to the system under test.
The relative relationship of packets received by the receiving device can be known, and the dispersion or deviation of the transmission delay time distribution and the number of discarded packets can be measured by the receiving device alone.

(Example) The present invention will be described in detail below with reference to the drawings.

FIG. 1 is a diagram illustrating a method of measuring dispersion or deviation of packet transmission delay time distribution according to a first embodiment of the present invention.

Reference numeral 1-2 is a serial number writing section, reference numeral 1-3 is an equal time interval transmission trigger generation section, reference numeral 2-2 is a serial number reading section, reference numeral 2-4 is a clock or counter for measuring arrival time, and reference numeral 2-5 is a serial number. and an arrival time storage section.

Unlike the conventional measurement of the dispersion or deviation of the transmission delay time distribution using absolute time, this embodiment uses a transmitting device to write serial numbers to packets in the order of transmission, so that the measurement packets received only at the receiving device can be compared to each other. The feature is that it is possible to measure the variance or deviation of the transmission delay time distribution from the relationship (arrival time interval).

In the present embodiment, the transmitting device receives a trigger from an equal time interval transmission trigger generation unit 1-3 that determines the time interval using a clock for measuring arrival time or a clock of the same frequency that is synchronized with the operating clock of the counter 2-4. As a result, measurement packets in which serial numbers have been written in the order of transmission by the serial number writing unit 1-2 are transmitted to the system under test. The receiving device is
A measurement packet output from the system under test is received, the arrival time is measured by the arrival time measuring clock or counter 2-4, and the serial number of the received measurement packet is read by the serial number reading unit 2-2. is stored in the storage unit 2-5. Further, the number of received packets is counted by a received packet number counter 2-3. The variance of the transmission delay time distribution is expressed by the following formula, where the serial number and arrival time stored in the storage unit 2-5 are respectively a and (i=1 to n), and the trigger interval of the equal time interval transmission trigger is (2). Also, the deviation is calculated as the square root of the variance.

FIG. 2 is a diagram illustrating a method for measuring the number of discarded packets, which is a second embodiment of the present invention.

Reference numeral 2-6 is a serial number storage section of the received measurement packet.

This second embodiment is different from the conventional method of measuring the number of discarded packets.
By writing serial numbers into packets in the transmission order in the transmitting device, it is possible to measure the number of discarded packets based on missing serial numbers of packets received only in the receiving device.

In the second embodiment, the transmitting device includes the serial number writing unit 1-
2, the measurement packets in which serial numbers are written in the transmission order are transmitted to the system under test. The receiving device receives the measurement packet output from the system under test, reads its serial number in the serial number reading section 2-2, and reads the serial number in the serial number reading section 2-2.
6. The number of packets discarded is determined by (1) If the input order of packets to the system under test and the order of output of packets from the system under test are the same, calculate the serial number of the previously received measurement packet from the serial number of the received measurement packet. By subtracting it, we can measure the number of packets discarded while receiving these two packets. Therefore, it is possible to specify the total number of packets discarded over the entire measurement time as well as the time at which the discard occurred with the resolution of the arrival time interval.

(2) If the order of input of packets to the system under test is different from the order of output of packets from the system under test, all the packets remaining in the system under test will arrive at the receiving device after transmission is stopped. After continuing reception for a sufficient period of time, the number of discarded packets can be measured without knowing the number of transmitted packets by checking the serial numbers stored in the storage section 2-6 and counting the number of missing serial numbers.

(Effects of the Invention) As explained above, the present invention writes serial numbers in the measurement packets in the order of transmission in the transmitting device, thereby eliminating the need for difficult absolute time matching between the transmitting and receiving devices and improving the distribution of packet transmission delay time. It has the effect of measuring variance or deviation. Furthermore, in a system under test where the input order and output order of packets are the same, it is possible to specify the time at which discard occurs with the resolution of the arrival time interval. Furthermore, in a system under test where the input order and output order of packets are different, the number of discarded packets can be measured only on the receiving device side without knowing the number of transmitted packets.

[Brief explanation of drawings]

FIG. 1 is a diagram showing a first embodiment of the present invention, FIG. 2 is a diagram showing a second embodiment of the present invention, and FIG. 3 is a diagram showing a conventional method for measuring dispersion or deviation of packet transmission delay time distribution. FIG. 4 is a diagram showing a second example of a conventional method for measuring dispersion or deviation of packet transmission delay time distribution, and FIG. 5 is a diagram showing a conventional method for measuring the number of discarded packets. 1: Measuring packet transmitting device, 1-1 = transmitting unit, 1-2 two serial number writing units, 1-3 equal time interval transmission trigger generating unit, 1-4 = transmission time measurement clock or counter, 1-5 : Transmission time storage section, 1 to 6: Transmission time writing section, 1-7: Transmission packet number counter, 2: Measurement packet receiving device, 2-1: Receiving section, 2-2; Serial number reading section, 2- 3: Received packet number counter, 2 to b 2-5: Serial number and arrival time storage unit, 2-6 = Serial number storage unit, 2-7: Arrival time measurement clock that indicates the same absolute time as 1-4 or Counter, 2-8 = Transmission time reading section, 3: System under test.

Claims (2)

[Claims] (1) In a method that treats information in packets and measures the packet transfer characteristics from the input to the output of the system under test, a serial number is written in the packets input to the system under test in the order of input, and a serial number is written on the output side of the system under test. A method for measuring packet transfer characteristics, characterized in that at least one of a packet transfer delay time and its dispersion in a system under test is measured according to a packet serial number and arrival time. (2) In a method that treats information in packets and measures the packet transfer characteristics from the input to the output of the system under test, a serial number is written in the packets input to the system under test in the order of input, and a serial number is written on the output side of the system under test. A packet transfer characteristic measurement method characterized by measuring the number of packets lost within a system under test by checking the serial numbers of received packets.