KR100791628B1 - Method for active controlling cache in mobile network system, Recording medium and System thereof - Google Patents

Abstract

An active cache control method of a mobile network system, a recording medium thereof, and a system thereof are disclosed.

The present invention comprises the steps of: receiving an update list including an update history that is a weighted average of the times when data on a server is updated; retrieving the data from the cache when a query for any data occurs at the client; If the data exists in the cache, if the elapsed time from the time when the previous update list is received from the server exceeds the update history, the next update list is received from the server, and if the elapsed time is equal to or less than the update history, Reading data present in the cache and requesting the data from the server if the data exists in the next update list.

According to the present invention, when a query occurs in a mobile environment, clients can use their caches in a short time, thereby reducing unnecessary response delay time and immediately obtaining a response to the query when there is data requested in the cache. Therefore, by increasing the cache utilization rate, the cache availability can be increased, and the false alarm problem can be eliminated.

Description

Active cache control method of mobile network system, recording medium and system therein {Method for active controlling cache in mobile network system, Recording medium and System}

1 illustrates a process of requesting a query and receiving a response in a conventional IR-based method.

2 illustrates a process of requesting a query and receiving a response in a conventional UIR based method.

3 illustrates an example of a situation in which unnecessary response delay occurs in a conventional UIR based method.

4 is a simplified diagram of a broadcast environment to which the present invention is applied.

5 shows a process of requesting a query and receiving a response in the present invention.

6 is a block diagram of a mobile network system according to the invention.

7 is a flowchart of an active cache control method of a mobile network system according to the present invention.

8 is a graph showing a conventional cache control method and the average response time according to the present invention.

The present invention relates to a mobile computing environment, to an active cache control method of a mobile network system, a recording medium and a system thereof.

Cellular phones. As the use of mobile devices such as PDAs increases, more services are expected to be serviced through the wireless environment. However, these services have limitations due to the mobile environment such as narrow bandwidth, limited battery capacity and frequent disconnection of clients. Storing data frequently used in a mobile environment in a user's storage space, or a user's cache, is an efficient way to efficiently use limited battery and reduce competition for narrow bandwidth usage.

However, the data stored in the cache may be invalidated because the original data may have been updated, but the copy may not be updated, resulting in the user not being able to use his cache immediately. Many recent studies have demonstrated that IR-based cache invalidation methods are very effective because of their scalability.

1 illustrates a process of requesting a query and receiving a response in a conventional IR-based method.

The Invalidation Report (IR) -based method has long response times and low cache availability, and in many cases users have to get data in a certain amount of time.

Even if the data is acquired within a certain time, a false alarm problem may occur that causes unnecessary response time delays. Many studies have been done to reduce this long response time. In these studies, an updated invalidation report (UIR) was used.

2 illustrates a process of requesting a query and receiving a response in a conventional UIR based method.

For example, when a client needs data called D1, if D1 is in its cache, it checks to see if its data is the same as the data on the server. First, the server periodically broadcasts the ID of the updated data. The ID of the updated data is included in the IR and UIR. The client checks the next IR or UIR if the data D1 is in its cache, and uses D1 in the cache if it is not updated. If it is updated, it requests D1 data from the server and receives the requested data after the next IR.

3 illustrates an example of a situation in which unnecessary response delay occurs in a conventional UIR based method.

When data d _x is in the cache and updated at time A and updated at time B, both UIRs broadcast at Ti _{, 2} contain information that d _x has been updated.

If d _x in the cache is used between A and B, when the update occurs in B, the conventional UIR-based method checks the UIR and if the update is successful, requests data from the server and sends the requested data. Receive behind the next IR.

However, no update occurred at the time the user confirmed that the actual data was in the cache. In other words, the data in the cache is the latest data at that time. Therefore, according to the UIR-based method, a false alarm is given to the client.

Therefore, in the conventional cache control method, since the client only passively waits for an IR or UIR, there is a problem that unnecessary response time delay, low cache availability, and false alarm occur.

The first technical problem to be achieved by the present invention is to provide an active cache control method of a mobile network system that can reduce unnecessary response delay time, increase cache availability, and eliminate false alarm problems.

A second technical object of the present invention is to provide a computer-readable recording medium having recorded thereon a program for executing the active cache control method of the mobile network system in a computer.

A third technical problem to be achieved by the present invention is to provide a mobile network system for actively controlling a cache to which the active cache control method of the mobile network system is applied.

In order to achieve the first technical problem, the present invention provides a mobile network system including a client that includes a cache and a server that broadcasts an update list, the update history being a weighted average of the times when the data of the server is updated. Receiving an update list including, If a query for any data in the client occurs, Searching the data in the cache, If the data exists in the cache, If the previous update list from the server If the elapsed time from the time exceeds the update history, receiving a next update list from the server; if the elapsed time is less than or equal to the update history, reading data existing in the cache and updating the next update list. When data is present, and provides an active cache of the mobile network control system comprises the step of requesting the data to the server.

In addition, in order to achieve the first technical problem, the present invention provides a mobile network system comprising a client with a cache and a server that broadcasts an update list, when the query occurs for any data in the client, the data Retrieving from the cache; if the data exists in the cache, receiving a next update list from the server if a time elapsed from the time of receiving a previous update list from the server exceeds a threshold previously stored in the client. And reading out data existing in the cache when the elapsed time is less than or equal to the pre-stored threshold value, and if the data exists in the next update list, requesting the data from the server. Active Cache of the System It provides a control method.

In order to achieve the second technical problem, the present invention provides a computer-readable recording medium having recorded thereon a program for executing the active cache control method of the mobile network system.

In order to achieve the third technical problem, the present invention provides a mobile network system including a client having a cache and a server that broadcasts an update list, wherein the data broadcast by the server and the time when the data of the server are updated. A client receiving unit which receives an update list including an update history which is a weighted average of about, a cache access unit which retrieves the data from the cache when a query for arbitrary data occurs in the client, and the data exists in the cache If the time elapsed from the time when the previous update list is received from the server exceeds the update history, the client receiver receives the next update list, and the elapsed time is the update history. Hayimyeon reading the data present in the cache, if said data present in the active cache controller and the next update list, and provides a mobile network system to actively control the cache that includes a request for requesting the data to the server.

One feature of the present invention is to eliminate unnecessary response delays. Another feature of the present invention is the use of data in the cache directly, thereby eliminating unnecessary delays that result in false alarms.

Due to this false alarm problem, the last value consistency model, which is a conventionally used consistency model, is not suitable for obtaining fast access time. Because the consistency model is application dependent, the present invention uses a weaker consistency model than before to increase cache availability.

The present invention uses the latest IR consistency. In this consistency model, the client is guaranteed consistency by the most recent IR / UIR.

In conventional IR / UIR-based algorithms, the client passively waits for IR / UIR for cache use, but actively uses the cache in the protocol proposed by the present invention.

The present invention provides two protocols, a Direct Cache Access Protocol (DCAP) and a Threshold Cache Access Protocol (TCAP).

In DCAP, if a client has data in its cache, it uses that data immediately without waiting for the next IR or UIR.

In order for the data to be not up-to-date when the client immediately uses its own cache, first, the data must be up-to-date until the latest IR / UIR, and second, the client's cache is generated from the latest IR / UIR. Updates must occur between points in time, and third, the data must be used by the client. Since the probability of all three conditions occurring at the same time is very small, in DCAP, even if the cache is used immediately, the loss of current is not significant.

The TCAP sets a threshold at a certain time and uses the cache immediately before that point and waits for the next IR / UIR after the threshold, just like a conventional IR / UIR. In fact, if the cache is read directly before the threshold, which results in a long response delay, the response time is close to that of the DCAP. However, since the second of the above conditions is reduced in proportion to the threshold, the loss of current is proportional to this. Can be reduced.

In the case of TCAP, the threshold can be adjusted by the client, which is the same as the DCAP if the threshold is one IR / UIR cycle. In other words, DCAP is a special case of TCAP.

Hereinafter, with reference to the drawings will be described a preferred embodiment of the present invention. However, embodiments of the present invention illustrated below may be modified in many different forms, and the scope of the present invention is not limited to the embodiments described below.

4 is a simplified diagram of a broadcast environment to which the present invention is applied.

4 is a diagram illustrating a system environment assumed in an IR-based cache invalidation method. In the present invention, an environment similar to that of FIG. 4 is assumed.

The server 400 includes a database DB and may include communication means for communicating with the clients 410-490. Clients 410-490 each include a cache. When information is transmitted from server 400 to clients 410-490, it is over a broadcast channel, and when information is transmitted from clients 410-490 to server 400, an on-demand channel (On-Demand). Channel).

5 shows a process of requesting a query and receiving a response in the present invention.

Preferably, the threshold may be an update history broadcast by the server at predetermined intervals or a constant value previously stored in the client.

When using update history as a threshold value, the update history is calculated by the weighted average of the recent update time and the past update history of the specific data as shown in Equation 1 below.

d _h = α * d _h-1 + (1-α) * (t _x -T _i )

Where d _x is the updated data ID, t _x is the time d _x was updated, d _h is the update history of d _x , d _h-1 is the past update history of d _x before d _h , and T _i represents the current time.

α is a system parameter for determining how much the past update history is reflected in the current update history, and is a real value between 0 and 1 that can be arbitrarily determined by those skilled in the art.

When using update history as the threshold, the client is updated by the server with the appropriate threshold.

If the threshold value is stored in the client as a constant in advance, those skilled in the art can determine an appropriate value.

After receiving the previous update list (IR or UIR) 510, 520, the threshold value Vt is set for a predetermined time, and before that time point, the cache is directly accessed to read data, and the time point after the threshold value Vt. Waits for the next update list (IR or UIR) 530, 540.

By reading the cache directly in the case before the threshold Vt, the response time delay is eliminated, resulting in a response time close to the DCAP.

At this time, between the time when the latest update list (IR or UIR) 510, 520 is received and the time when the cache check (cache search) of the client occurs, the update occurs is reduced in proportion to the threshold value, Damage may be reduced in proportion.

The TCAP protocol, which is one of the features of the present invention, uses the cache immediately after a reference point around a specific reference point, and after the next IR, the same as a conventional IR-based protocol.

An embodiment of the TCAP protocol is described as follows.

Where Q _i = {d _x | d _x is the data queried after T _i } and Q _{i, k} = {d _x | d _x is the data queried at interval [T _{i, k-1} , T _{i, k} ]}, t ^c _x is the timestamp of the cached data item d _x , and T _t is the timestamp of the last IR received. , T _c is the current time.

Also, the T _threshold is a time reference point of the _threshold and has a range of 0 ≦ T _threshold ≦ L. T _threshold of FIG. 5 is the same concept as the threshold value Vt.

When using update history as a threshold, the update history is calculated by the weighted average of the recent update time and the past update history of the specific data, as shown in equation (1).

L _data is an ID list of data items requested by the user to the server and corresponds to a client request. t _x is the update time of the data item included in the IR, T _l is the timestamp (received time) of the last IR received, and T _i is the timestamp (received time) of the received IR _i , L _bcast is the list of data broadcast to the user by the server, d _x is the data item ID, UIR _{i, k} is the _kth UIR between IR _i and IR _{i + 1} , and L is the broadcast period of the IR. Where w is the size of the window (IR contains update information as much as w periods).

Table 1 below shows a protocol when user C _j receives IR _i and L _bcast .

if T _l <(T _i -L * w) then delete the entire cache entry; invalidate each data item <d _x , t ^c _x > do if ((d _x , t _x ) ∈ IR _i ) ∧ (t ^c _j <t _x ) then d _x in the for cache; for each data d _x ∈ L _bcast do if (d _x ∈ L _data ) then download d _x to the client local cache; use d _x as the answer to a recent query; download the cache item then d _x to the client's local cache if if d _x has been invalidated; T _l = T _i ; if (L _data ≠ φ) then query (Q _i ); L _{data ≠} = φ;

In Table 1, when a user named _{j j} receives an validation report and data, if the user receives the IR last time and the user does not receive the IR for more than L * w hours, the entire cache entry is deleted.

Next, if dx is included in the IR and the update time of dx in the IR is greater than the update time in the cache, dx is invalidated.

"T _l = T _i ;" Means to set the timestamp of the currently received IR to the timestamp of the last received IR.

Table 2 below shows a protocol when a user receives UIR _{i, k} .

if IR _i not received then break; wait for the next IR; invalidate each data item <d _x , t ^c _x > do if (d _x ∈ UIR _{i, k} ) then d _x in the for cache; if (L _data ≠ φ) then query (Q _i ); L _{data ≠} = φ;

Table 3 below shows a protocol when a client receives a query for d _x .

if (T _c -T _t <T _threshold ) if d _x is valid data in the cache then immediately uses the value in the cache for question and answer; send a query for d _x to the else server; else L _data = L _data ∪ d _x ;

The meaning of "if (T _c -T _t <T _threshold )" indicates a case where the current time minus the last IR received time is smaller than the threshold value.

Table 4 below shows the procedure of query (Q) described in Tables 1 to 3.

for each query d _x ∈ L _data do L _data is sent to the server;

6 is a block diagram of a mobile network system according to the invention.

The client 610 includes a cache access unit 612, a cache 613, an active cache controller 614, a client receiver 615, and a request unit 616.

When the query 611 for arbitrary data occurs in the client 610, the cache access unit 612 searches the cache 613 for the data in which the query 611 occurs.

The active cache controller 614 may next perform a request by the client receiver 615 if the elapsed time from the time when the previous update list is received from the server 600 exceeds the threshold value when the query data exists in the cache 613. Lets you receive a list of updates. At this time, the client 610 waits until the next update list is received. In this case, the threshold value is greater than zero and smaller than the IR or UIR period, and may be arbitrarily determined by those skilled in the art. Preferably, the update list may be broadcast by the server 600 at predetermined intervals.

When using update history as a threshold, the update history is calculated by the weighted average of the latest update time and the past update history of the specific data, as shown in Equation (1).

In this case, the previous update list is the most recent update list based on the current time point, and the next update list is the update list received at the next period based on the current time point.

The active cache controller 614 reads the data present in the cache 613 and uses it in response to the query 610 when the elapsed time from the time of receiving the previous update list is less than or equal to the threshold.

Preferably, the active cache control unit 614 may read the data present in the cache 613 if the data in which the query 611 occurs in the next update list does not exist.

The client receiver 615 receives the data and the update list transmitted from the server 600.

The request unit 616 requests the server 600 for the data in which the query 611 has occurred if the data in which the query 611 has occurred exists in the next update list.

Preferably, the request unit 616 may request the server 600 for the data generated by the query 611 when the data generated by the query 611 does not exist in the cache 613.

Preferably, server 600 may include an on-demand data manager 601, a broadcast manager 602, and a database 603. At this time, the on-demand data manager 601 receives a request of the request unit 616. At this time, the broadcast manager 602 searches the database 603 according to the request of the request unit 616, and transmits the retrieved data to the client receiver 615. At this time, the database 603 stores a plurality of data for the client 610.

7 is a flowchart of an active cache control method of a mobile network system according to the present invention.

Assume a mobile network system that includes a client and server with a built-in cache.

The server broadcasts an update list every predetermined period. The update list may include an update history calculated by a weighted average of recent update history and past update history of specific data, as shown in Equation (1).

First, when a query for arbitrary data occurs in the client, the cache searches for data in which the query occurs (step 710).

In operation 720, it is determined whether the data in which the query occurs in the cache exists.

If the query has data in the cache, if the elapsed time from receiving the previous update list from the server exceeds the threshold (step 730), the next update list is received from the server (step 740). If it is less than the threshold value, the data existing in the cache is read (step 760).

When using update history as a threshold, the update history is calculated by the weighted average of the recent update time and the past update history of the specific data, as shown in Equation 1.

On the other hand, if there is no data in which the query occurs in the cache, the server requests the data in which the query occurs (step 770).

When the elapsed time exceeds the threshold value and receives the next update list from the server, if the data having a query exists in the next update list (step 750), the server requests the data having the query (step 770).

On the other hand, if there is no data in which the query occurs in the next update list (step 750), the data existing in the cache is read (step 760).

Preferably, the update list may be broadcast by the server at predetermined intervals.

Preferably, the update list may include either an validation report or an updated validation report that notifies invalidation of specific data prior to the current time.

Preferably, the server includes a database for storing a plurality of data, can search the database at the request of the client, and send the retrieved data to the client.

8 is a graph showing a conventional cache control method and the average response time according to the present invention.

In order to analyze the performance of the present invention, a simulation was performed assuming that there is one server and several clients.

It is assumed that data updates occur only on the server and requests for data occur only on the client. And to simplify the simulation, we do not consider the size or bandwidth of the data and the case where the client is disconnected from the server. The parameters used in the simulations are shown in Table 1 below.

Parameter (unit) Contents Parameter (unit) Contents Database size (data item) 1000 UIR replicate times 4 Number of clients 100 HOT data items 5% of DB Broadcast interval (sec) 200 Cold data items 95% of DB Broadcast bandwidth (bits / sec) 10000 Hot data update probability 33% Cache size (data item) 1 to 500 Client's zipf parameter 0.95 Broadcast window (broadcast interval) 4 Query generate time (data item) 1 to 200

As shown in Figure 8, the present invention shows that the average response time is shorter than the conventional IR / UIR used.

Preferably, the update list may be broadcast by the server at predetermined intervals.

Preferably, the threshold may be an update history broadcast by the server.

Preferably, the update list may include either an validation report or an updated validation report that notifies invalidation of specific data prior to the current time.

Preferably, a program for executing the active cache control method of the mobile network system of the present invention may be provided by recording a program for executing in a computer on a computer-readable recording medium.

The invention can be implemented via software. When implemented in software, the constituent means of the present invention are code segments that perform the necessary work. The program or code segments may be stored on a processor readable medium or transmitted by a computer data signal coupled with a carrier on a transmission medium or network.

The computer-readable recording medium includes all kinds of recording devices in which data is stored which can be read by a computer system. Examples of computer-readable recording devices include ROM, RAM, CD-ROM, DVD ± ROM, DVD-RAM, magnetic tape, floppy disks, hard disks, optical data storage devices, and the like. The computer readable recording medium can also be distributed over network coupled computer devices so that the computer readable code is stored and executed in a distributed fashion.

Although the present invention has been described with reference to one embodiment shown in the drawings, this is merely exemplary and will be understood by those of ordinary skill in the art that various modifications and variations can be made therefrom. However, such modifications should be considered to be within the technical protection scope of the present invention. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

As described above, according to the present invention, when a query occurs in a mobile environment, clients can use their caches in a short time, thereby reducing unnecessary response delay time and immediately querying the cache without waiting for the requested data. Because you can get a response to, you can increase the cache availability by increasing the cache utilization rate, and can eliminate the false alarm problem.

Claims (12)

In a mobile network system comprising a client with a cache and a server for broadcasting the update list, Receiving an update list including an update history that is a weighted average of the time at which the data of the server was updated; Retrieving the data from the cache when a query for any data occurs at the client; If the data exists in the cache, if the time elapsed from the time of receiving the previous update list from the server exceeds the update history, the next update list is received from the server, and if the elapsed time is less than or equal to the update history. Reading data present in the cache; And Requesting the data from the server if the data exists in the next update list. The method of claim 1, Requesting the data from the server And if the data does not exist in the cache, requesting the data from the server. The method of claim 1, Requesting the data from the server And if the data does not exist in the next update list, reading data existing in the cache. The method of claim 1, The update list is A method of active cache control of a mobile network system, which is broadcasted by the server at predetermined intervals and includes either an validation report or an updated validation report that notifies invalidation of specific data before a current time. . The method of claim 1, The server is A database for storing a plurality of data, generating a current update history by calculating a weighted average of a recent update time and a past update history of specific data in the database, and generating the current update history at the request of the client. And searching and broadcasting an update list including the retrieved data and the update history at predetermined intervals. delete A non-transitory computer-readable recording medium having recorded thereon a program for executing the method of claim 1. In a mobile network system comprising a client with a cache and a server for broadcasting the update list, A client receiver configured to receive an update list including data broadcast by the server and an update history that is a weighted average of the times when the data of the server is updated; A cache access unit for retrieving the data from the cache when a query for any data occurs in the client; If the data exists in the cache, if the elapsed time from the time when the previous update list is received from the server exceeds the update history, the client receiver receives the next update list, and the elapsed time is the update. An active cache controller which reads data present in the cache if it is less than or equal to a history; And And if the data exists in the next update list, actively controlling a cache including a request unit for requesting the data to the server. The method of claim 8, The request unit And if the data does not exist in the cache, requesting the data from the server. The method of claim 8, The active cache controller If the data does not exist in the next update list, the mobile network system to actively control the cache, characterized in that for reading the data existing in the cache. The method of claim 8, The update list is A mobile network that is broadcast by the server at predetermined intervals and includes either an validation report or an updated validation report that notifies invalidation of specific data before a current time. system. The method of claim 8, The server is A database for storing a plurality of data; An on-demand data manager receiving the request of the request unit; And The current update history is generated by calculating a weighted average of a recent update time and a past update history of specific data in the database, searching the database according to a request of the request unit, and including the retrieved data and the update history. And a broadcast manager that broadcasts an update list every predetermined period.

Images