Data processing: artificial intelligence – Knowledge processing system
Reexamination Certificate
1997-10-16
2001-05-08
Hafiz, Tariq R. (Department: 2762)
Data processing: artificial intelligence
Knowledge processing system
C370S230000, C370S232000, C706S004000, C706S006000
Reexamination Certificate
active
06230152
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to queue management systems and particularly, to a controller implementing fuzzy logic for controlling dequeuing operations of customer screening queues in a loop management operating system used by regional telephone companies.
BACKGROUND OF THE INVENTION
Currently, in the telephone industry, front end systems are provided that contain information about telephone customers, and particularly, telephone numbers, telephone service histories, and other relevant comments regarding, e.g., the customer's dwelling and service record. As shown in
FIG. 1
, such front end systems
50
a,b
, . . .
50
n
are commonly known as Loop Management Operating Systems (hereinafter “LMOS”) and interface with a remote host mainframe computer
75
containing the primary service and operational databases. The front end systems
50
a,b
, . . .
50
n
maintain a facility for receiving telephone calls from customers requesting particular telephone service, e.g., when a telephone line or cable breaks, or when service is interrupted, etc., and service personnel enter details regarding the customer's complaint and service requested. Once the phone company investigates the particular problem(s) and determines the causality of the problem, it will dispatch maintenance personnel to rectify the problem.
As further shown in
FIG. 1
, each LMOS front end system deploys a plurality of screening maintenance centers (“SMC”), indicated as
80
a
1
,
2
,
3
, . . . ,
80
b
1
,
2
, . . . and
80
n
1
,
2
,
3
, . . . that respond to particular calls, e.g., associated with a particular area or maintenance center, etc., and typically may have up to several hundred SMCs. Furthermore, as shown in
FIG. 1
, embedded in the LMOS front end systems are rule-based expert systems
90
a,b
, . . .
90
n
that can automatically determine the nature of a trouble before dispatching a service rig, or, e.g., determine if the source of the problem is to be correctly assigned to that front end center. One expert system in particular is the screener decision unit (“SDU”) which takes a trouble description, e.g., as typically called-in by a customer, looks at the various properties of the trouble, e.g., telephone number, exchange, etc., and looks to see if there is any way to automatically handle the call. For example, the SDU may automatically determine if another front end should be handling the call, or, automatically send the call to a dispatch maintenance center. All trouble calls are placed in a queue to await disposition via the expert systems just described. The queue contains, for each trouble, a pointer to relevant information such as customer name, telephone number, trouble code, etc. There is one queue for each SMC on the front end.
As shown in
FIG. 2
, the front end will generate a queue
85
, corresponding to a particular SMC associated with that front end, with each queue containing pointers
86
, as described above, that point to the particular record of the customer calling in the trouble situation. Depending upon circumstances, e.g., system load, etc., the queue for any given SMC at any given instant may have few or many pointers attached to it. These queues may increase or decrease in size, and, in extreme circumstances, may increase in size dramatically in very short period of times, for instance, when there is a storm or other natural disaster that interrupts telephone service. For example, as shown in
FIG. 2
, associated with front end
50
n
, one SMC queue, e.g.,
80
n
m−1
, may have ten customer complaints, i.e., pointers
86
′ associated with it, while another SMC queue, e.g.,
80
n
m
, being associated with an area having just incurred a natural disaster, may have tens of thousands of customer complaints, i.e., pointers
86
″ associated with that SMC. There is, in principle, no limit to the size of any queue, i.e., to the number of pointers (troubles) that may be enqueued upon it.
Currently, as shown in
FIG. 2
, each LMOS SDU employs a “Drainer” mechanism, indicated by arrow
99
, that iteratively loops through all of the queues (one for each SMC) and selects the queue for screening in a fixed, orderly fashion, with the processing time allocated to a single SMC determined on the basis of three variable parameters: Queue Size (Q), the single-pass drain limit controlling how many pointers will be drained from the queue at a maximum before the Drainer passes on to examine the queue for the next SMC; Throttle Time (T), the predetermined amount of time that the process idles or sleeps after each item in the queue for that SMC is screened; and, Sleep Time (S), the time delay before the SDU begins to drain the next SMC queue. Thus, as an example, if there are greater or equal to“Q” troubles on a queue “x”, and the throttle time is “t” seconds, then exactly Q troubles will be drained from the queue “x” in Qt seconds (ignoring processing overhead) when queue “x” is allocated its “turn” by the Drainer. The sleep time “s” is additional time the Drainer “rests” after processing a queue (and before it begins processing the next queue). By varying the Q,T and S parameters, it is possible to increase or reduce the time slice used by the Drainer relative to LMOS as a whole.
Currently, there is no mechanism for dynamically adjusting these parameters in response to changes in load distribution, i.e., unbalanced queue sizes. Thus, it is possible that some of the queues may increase in size to the point where a number of hours may elapse between the time a trouble is called in and the time that it is screened. Traditional queue management/queue selection techniques are not particularly useful in this connection because it is not possible to determine how many elements (pointers) reside on each queue.
Thus, it would be highly desirable to provide a mechanism that dynamically adjusts the S,T and/or Q parameters to properly balance the drain queue load and optimize Drainer performance in the face of unpredictable real-time demands.
Furthermore, it would be highly desirable to find a queue management system that manages queues in the absence of knowledge of the number of queues and size of each queue, since as mentioned herein, such knowledge is not readily available to the Drainer.
SUMMARY OF THE INVENTION
Generally, the invention is a queue management system implementing a fuzzy controller that manages queues in the absence of knowledge of the size of each queue and which can be applied to a system which contains an arbitrary number of queues.
Particularly, the queue management system is a watchdog expert system employing a fuzzy logic controller implemented by the LMOS Drainer to monitor the queuing environment and to effect increases or decreases in the Q, S and T parameter values as needed. As opposed to most queue management implementations, the controller does not dictate which queue to select from, but rather controls the environment in which the queue manager operates. Thus, it determines a first stable state which the Drainer attempts to maintain over time and then adjusts the parameters to keep the controller output as close to this first stable state as possible.
The various features of novelty which characterize the invention are pointed out with particularity in the claims annexed to and forming a part of the disclosure. For a better understanding of the invention, its operating advantages, and specific objects attained by its use, reference should be had to the drawings and descriptive matter in which there are illustrated and described preferred embodiments of the invention.
REFERENCES:
patent: 5515428 (1996-05-01), Sestak et al.
patent: 5812526 (1998-09-01), Chang et al.
patent: 5841084 (1998-11-01), Thangavelu
“Effective control of traffic flow in ATM networks using fuzzy, explicit rate marking (FERM)”, Pitsillides, A.; Sekercioglu, Y.A.; Ramamurthy, G., Selected Areas in Communications, IEEE Journal on, vol.: 15, Feb. 1997, pp. 209-225.*
“Fuzzy service rate control of queueing systems”, Zhang, R.; Phillis, Y.A., Systems, Man an
Brown Edward G.
Hafiz Tariq R.
Lucent Technologies Inc
LandOfFree
Fuzzy controller for loop management operating system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Fuzzy controller for loop management operating system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Fuzzy controller for loop management operating system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2564946