JPH0260362A - How to avoid traffic concentration - Google Patents

Abstract

PURPOSE:To avoid concentration of traffic on a communication line by forecasting the period, when traffic concentration will occur, based on past data and transferring data of prescribed subscribers to a data base to which forecasted terminals belong. CONSTITUTION:Terminals 31 to 3M or a host computer 1 connected by a communication line 2 forecasts concentration of traffic of the day beforehand based on past traffic data or information of traffic in days of a month, traffic in days of a week, traffic in days preceding and following holidays, etc. Forecasted conditions in this year obtained in this manner are compared, and the normal processing is performed when traffic concentration is lower than a limit value, but terminals used by subscribers are forecasted when it is not lower than the limit value. A using subscriber forecasting means 30 obtains the terminal, which each subscriber most frequently used within a past fixed period, and preliminarily transfers data of the subscriber to this terminal by a data communicating means 40 and receives data of the subscriber from the other terminals.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to an online system via a telephone or dedicated communication line, and in particular, to a traffic concentration avoidance method and method suitable for automatic cash dispensing systems such as banks. Regarding equipment.

[Conventional technology]

In conventional systems, especially online systems such as banks, when input from a terminal occurs, the data is transmitted to the place where the accessing person's data is stored (host computer or account opening branch database) and predetermined processing is performed. After that,
Perform payment processing.

In this case, if the subscriber accesses all terminals at once, traffic will concentrate on the communication line (telephone or dedicated line, etc.), and if this continues, the host computer will be overloaded and the system may go down. .

That is, the number of requests that the host computer can process is set at the time of system design, and the number of accesses from terminals exceeds the limit.

However, such a system is designed to have a sufficient margin H1 so that overload does not occur.

On the other hand, as networks have recently become more widespread and the number of terminals has increased, the probability of access exceeding the design settings has increased, and there are cases in which such a phenomenon actually occurs and develops into a social problem.

[Problem to be solved by the invention]

In the above-mentioned conventional technology, the overload of the system is often set based on statistics and probability distributions, so there is always a possibility that the overload is exceeded.

It is a well-known fact that traffic concentration tends to occur particularly around holidays and holidays.

The object of the present invention is to avoid and prevent system overload and system down due to such concentration of traffic.

[Means to solve the problem]

The above object is achieved by accurately predicting the time when the concentration of 1~1000 subscribers will occur and the terminals used by subscribers, and by transferring predetermined subscriber data to the predicted terminals in advance.

[Effect]

The traffic concentration prediction means uses past data, month, day, and day of the week to predict when traffic concentration will occur.

Prediction is made using parameters such as before and after holidays.

Further, the usage subscriber prediction means predicts the current access terminal from the subscriber's past access terminals and transfers data of a predetermined subscriber in advance to a database to which the predicted terminal belongs. As a result, traffic concentration due to communication does not occur, and each branch office can be processed within the processing range of the host computer, thereby preventing overload or system down.

〔Example〕

An embodiment of the present invention will be described below with reference to FIG. 1st
In the figure, 1 is a host computer, 2 is a communication line, 31 to
3M is a terminal (or branch), 311-31. M is an automatic payment machine, and 4 and 41 to 4M are storage devices.

one The operation of the present invention will be explained below with reference to FIGS. 1 and 2.

The computers at each terminal or branch shown in Figure 2 collect past traffic data, month, day, and day of the week.

Based on information such as before and after a holiday, the concentration of rahitsuku in 1 on the day is predicted before the day using, for example, equation (1).

Tp=ax ・Tr α2Ma+a8Wd+α4 Hd
+...(1) α1 to α4: Weighting coefficient Tr: Past traffic situation on the same month and day M,: Traffic situation due to different days Wd: Traffic situation due to different days of the week Hd: Traffic situation before and after the combination (1) Equation Each parameter in is stored in advance as shown in FIG. 3, for example. Also, the weighting coefficient α is (2)
It is determined in advance so that the conditions of the expression are satisfied.

α engineering + α2 + α3 + α4 = 1 ・(
2) Here, for Tr, data from the previous year is used, or traffic conditions for different months are determined from past data using equation (3).

Tr(M)=(Tro(M)+T,-+(M))/2
-(3) Tr (M): Past 1-2 hour traffic situation in month M Tro (M): Traffic situation from 1 year ago in month M +-1 (M): Last year 1-2 hour traffic situation in month M This year's predicted situation obtained in this way is compared using equation (4), and Tp is the traffic concentration limit value (T p L
M T ) In the case of the terminal, normal processing is performed, and if the limit value or more is exceeded, the subscriber's terminal is predicted.

Tp<TpLMT...(4)Tp
LMT: Traffic concentration limit value This usage subscriber prediction means determines the terminal that each subscriber has used the most within a certain period of time in the past, and transfers the subscriber's data to that terminal in advance using the data transfer means. , receives subscriber data from other terminals.

By performing the above two processes, it is possible to avoid traffic concentration on the day. In addition, if the subscriber uses a terminal other than the one predicted, a communication line will be used.
Most of the traffic can be reduced and there will be no overload.

Meanwhile, the transferred subscriber data is returned to the original terminal at night on the same day to predict traffic concentration the next day.

According to the above-described embodiment, it is possible to avoid concentration of traffic due to the use of terminals by subscribers and overload of the host computer or the terminals at each branch office, thereby not causing any inconvenience to the subscribers.

The present invention does not prescribe when to perform processing. In other words, even if you predict the next day in your free time every day, -
You may also make predictions and execute them all at once on holidays.

Alternatively, prediction may be made for several days before and after a holiday without making predictions on weekdays.

The present invention is effective as long as it does not go against the gist of transferring subscriber data in advance to a storage device (data storage) to which the terminal belongs based on predictions of traffic concentration and the prediction results of the terminals used by the subscriber.

Further, the subscriber data mentioned above refers to a series of information regarding subscriber number, deposit amount, deposit/withdrawal history, etc.

〔Effect of the invention〕

According to the present invention, concentration of traffic on communication lines can be avoided, so that the host computer or terminal will not be overloaded and the subscribers will not be inconvenienced.

[Brief explanation of the drawing]

FIG. 1 is a system configuration diagram of an embodiment of the present invention, FIG. 2 is a software configuration diagram of terminals at each branch, and FIG. 3 is a diagram showing an example of the traffic situation of each parameter. 1...Host computer, 2...Communication line, 31-3M terminal or computer at each branch, 4 and 41-4M storage device, 10...Traffic concentration prediction means, 20 comparison means, 3
o...Using subscriber prediction means, 40...Data exchange means.

Claims (1)

[Scope of Claims] 1. A method for avoiding traffic concentration in an online system using a communication line of a bank or the like, comprising a traffic concentration prediction means, a user subscriber prediction means, and a data exchange means. 2. A method for avoiding traffic concentration, wherein the traffic concentration prediction means according to claim 1 predicts the traffic situation based on at least information about the month and day of the prediction date, day of the week, and holidays. 3. A method for avoiding traffic concentration, characterized in that the usage subscriber prediction means according to claim 1 predicts the terminal that the subscriber will use this time based on the terminals that the subscriber has used within a certain period in the past. 4. A traffic concentration avoidance method, characterized in that the data exchange means according to claim 1 exchanges subscriber data (subscriber number, deposit amount, deposit/withdrawal history, etc.) based on paragraph 3.