Multiplex communications – Diagnostic testing – Determination of communication parameters
Reexamination Certificate
2000-03-08
2003-06-24
Chin, Wellington (Department: 2664)
Multiplex communications
Diagnostic testing
Determination of communication parameters
Reexamination Certificate
active
06584072
ABSTRACT:
1.0 FIELD OF THE INVENTION
This method determines the drop rate, the transit delay and the break state of communications objects using the topology (connectivity) of these objects.
1.1 BACKGROUND OF THE INVENTION
Existing methods for determining whether or not a communications device is broken depend on periodically sending frames to it which require the device to respond (e.g. SNMP requests and responses (RFC 1157)). The absence of any response to a sequence of requests indicates the device is either broken or that the communications path to the device is broken. The best method for exploiting this information using knowledge of the network topology is reported by Dawes et al (Network Diagnosis by Reasoning in Uncertain Nested Evidence Spaces: N. W. Dawes, J. Altoft, B. Pagurek: IEEE Transactions on Communications, #2, 43, pp 466-476, 1995). This earlier method does not exploit measurements of the traffic rates on lines connected to devices and so is far more complex and far later to detect break faults than the method described below. It also is marginally less accurate. Commercially deployed break fault methods are very significantly inferior to even this previous method.
Existing methods for determining the transit delay across a device rely on requesting this information from the device itself, in the case where the device measures this delay and records it so it can be read externally. However, many devices do not have these facilities. Many of those that do, do so in a manner which is particular to that version of that manufacturer's device, placing the information in certain variables somewhere in the MIB (RFC 1213). This makes the process of determining the transit delay across a device cumbersome and complex, as variation need to be made for the particular device type.
Existing methods for determining the drop rate of a device depend on what percentage of responses it makes to management requests. They do not use knowledge of the local topology of objects and so are far less accurate than the present invention.
1.2 SUMMARY OF THE INVENTION
A method of determining the topology of a network of objects has been filed for patent, Dawes et al, U.S. Ser. No. 08/558,729 filed Nov. 16, 1995, Ser. No. 08/599,310 filed Feb. 9, 1996 and (unknown) filed Nov. 15, 1996 incorporated herein by reference. A manual method or some alternative automatic method, allows the connectivity of communications objects to be determined.
A new method described below also works on unmanaged objects and sets of unmanaged objects, which is novel.
The invention exploits knowledge of the detailed local topology of communicating objects.
1.21 DEFINITIONS
Communications objects such as routers have multiple communications lines. They accept frames from these lines and determine from information in each frame which line each frame should be sent out on.
Transit delay:
The time between the receipt of a frame and its dispatch out again is called the transit delay.
Drop rate:
Sometimes routing or switching communications devices cannot dispatch frames as fast as they receive them and run out of memory to store the ones they receive, so they discard some. In addition, internal queues may fill up and for other reasons, frames get lost between acceptance and onward dispatch. The overall discard rate is usually called the drop rate.
Break:
Communications devices, routing or otherwise, can break. The break state for a device is true when it can neither send nor receive on any communications line, yet all the lines are ok. For example, when a device is powered down its break state is true. The break state is true for a line when the devices at each end are not broken and yet cannot send or receive traffic across it. For example, a line is broken when it is cut through.
NMC:
The network management center is the computer which is operating the software that performs this method. It also either performs interrogation of devices to provide data for the method below or receives such data to use in the method.
The NMC periodically requests from each device in a communications network the amount of traffic flowing in and out of each interface and the line status (OK or OFF) on the line for each interface on that device. This request should result in a set of replies from each device returned to the NMC. Not all devices need report the OK or OFF line status values or do so correctly.
If a device breaks then the NMC may detect four changes. First that it now receives no replies to its requests of this device. Second that it receives no replies from devices lying beyond this device and which are only reachable through this device. Third no traffic will now be detected flowing in any lines to or from this device. Four the line status bits on lines connected to this broken device will change (e.g. from ok to off). Any subset of two or more of these four changes will be adequate to determine that the device is broken.
If a line between two devices is broken, the status bits on the interfaces at each end may change and no traffic will flow. Should neither device be broken then and yet should either of these conditions be met, then the line itself is broken. This diagnosis depends on the device break diagnosis above.
The drop rate in a device is the difference between the mean drop rate measured to devices just beyond it (and connected to it) and the mean drop rate measured to devices just before it (and connected to it), where closeness is measured in terms of the number of hops to the NMC. Devices diagnosed as broken should not be included in any part of this calculation.
The mean frame transit-delay in a device is the difference between the mean round trip time measured to devices just beyond it (and connected to it) and the mean round trip time measured to devices just before it (and connected to it), where closeness is measured in terms of the number of hops to the NMC. Devices diagnosed as broken should not be included in any part of this calculation.
The result is far simpler and far more generally applicable method which gives similar or better results. This means that all the devices in communications networks can now be analyzed, without any undue burden on the network bandwidth or in machine facilities.
In accordance with an embodiment of the invention, a method for determining the mean transit delay of frames through one or more communications devices which receive and forward frames. p In accordance with another embodiment, a method for determining the mean drop rate of frames through one or more communications devices which receive and forward frames.
In accordance with another embodiment, a method for determining the break state of one or more communications devices and interfaces or lines to and from communications devices.
In accordance with another embodiment, a method of analyzing a communication network comprising determining a mean drop rate in a device x by polling each device from a network management computer (NMC) which is in communication with the network, and processing signals in the NMC to determine a drop rate D(x), in accordance with:
D
(
x
)=((
L+
(
x
)−
L−
(
x
))/2,
and
L
(
x
)=1−
A
(
x
)
where
A(x): the fraction of poll requests from the NMC to device x for which the NMC receives replies (measured over the last M sampling periods), (wherein device x must not be broken),
D(x): the mean frame drop rate in device x,
L(c): NMC's perception of the loss rate to device x and back,
L−(x): the NMC's perception of the mean value of L(z) for all devices z connected to device x, closer to the NMC than device x and which are not broken, and
L+(x): the NMC's perception of the mean value of L(z) for all devices z connected to device x, further away from the NMC than device x and which are not broken.
In accordance with another embodiment, a method of analyzing a communication network comprising determining a mean frame transit delay in a device x by polling each device from a network management computer (NMC) which is in communication wi
Baker Harold C.
Chin Wellington
Hendry Robert G.
Jones Prenell
Loran Network Management Ltd.
LandOfFree
Method for determining the drop rate, the transit delay, and... 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 for determining the drop rate, the transit delay, and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for determining the drop rate, the transit delay, and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3093481