JP5808694B2 - E-mail traffic regulation device - Google Patents

Description

  The present invention relates to an electronic mail traffic regulation apparatus, and more particularly, to an electronic mail traffic regulation apparatus capable of preventing bottleneck congestion and proxy server and mail server overload regardless of the location of the bottleneck.

  Various techniques have been proposed for preventing network line congestion and server overload due to e-mail traffic.

  Patent Document 1 discloses a system that monitors the total size and total number of e-mails and changes a mail server in charge of distribution based on load conditions.

  Patent Document 2 discloses a system that counts the number of mails sent for each IP address of a transmission source user and stops access to the mail server by a user whose number of mails exceeds the threshold, that is, mail transmission.

  According to Patent Document 3, the demand for reply mail to received mail, that is, outgoing demand is reduced by restricting or delaying mail arrival notification depending on the congestion state of the network or server, and further restricting outgoing transmission on the terminal side. A system is disclosed.

JP 2009-177432 A JP 2006-310902 A JP 2010-161500 A

  Congestion of network lines and server overload due to concentration of e-mail traffic (hereinafter sometimes collectively referred to as congestion) causes many users to send and receive e-mail at the same time, such as at the end of the year and after the earthquake. This is likely to occur.

  Since Patent Document 1 is a technique for changing a mail server in accordance with an increase in load, even if congestion occurs in a bottleneck line, this cannot be solved.

  In Patent Document 2, when congestion occurs, the traffic is regulated by stopping the user's mail transmission. However, congestion due to the concentration of e-mail traffic can also occur due to the received traffic alone. It is difficult to suppress enough. Moreover, since the regulation of mail transmission is easily recognized by the user of the transmission terminal, the user's service usage feeling is greatly impaired. Further, even if the congestion caused by the received traffic is to be solved by the restriction of the transmission traffic, it cannot be restricted that the transmission source is another domain managed by another network operator.

  In Patent Literature 3, by restricting or delaying incoming notifications associated with mail reception, the demand for outgoing mail related to delivery of received mail and replies to received mail is suppressed / reduced, but congestion and overload are captured quantitatively. Therefore, even if the incoming notification can be restricted or delayed, an appropriate control value that does not cause congestion or the like cannot be calculated.

  An object of the present invention is to solve all of the problems of the prior art described above, and to provide an electronic mail traffic regulation device capable of preventing bottleneck congestion and proxy server and mail server overload regardless of the location of the bottleneck. It is in.

  In order to achieve the above object, according to the present invention, a wireless terminal establishes a passive fetch session via a proxy line and a proxy server with a received mail server in response to an incoming mail notification, and acquires the received mail. The received mail traffic regulating device of the electronic mail system is characterized in that the following measures are taken.

  (1) Allowable bandwidth setting means for setting the allowable bandwidth of the proxy line, transfer data amount detecting means for detecting the transfer data amount per fetch session related to mail reception, allowable bandwidth of the proxy line and transfer per fetch session Based on the amount of data, an acceptable rate calculating means for calculating an acceptable mail reception rate for the proxy line, and an incoming notification to a terminal having a destination of the received mail stored in the received mail server based on the acceptable mail reception rate An incoming call notification rate determining means for determining a rate for delivering the incoming mail, and an incoming call notification control means for delivering the incoming call notification to the terminal having the destination of the received mail based on the determined incoming call notification rate.

  (2) A means for calculating an occurrence rate of active fetch that is established regardless of mail arrival notification is further provided, and the incoming notification notification determination means determines the distribution rate of the incoming notification as the allowable rate of mail reception and the active rate. The decision was made based on the fetch rate.

  (3) The incoming notification rate determination means determines the distribution rate of the incoming notification as a rate obtained by subtracting the active fetch generation rate from the allowable rate of mail reception.

  (4) A means for acquiring an occurrence rate of a non-response terminal that does not establish a passive fetch session with respect to an incoming notification is further provided, and the incoming notification notification determining means determines the distribution rate of the incoming notification as an allowable rate for receiving the mail The decision was made based on the response terminal's rate of occurrence.

  (5) The incoming notification rate determining means determines the distribution rate of the incoming notification as a rate obtained by adding the generation rate of the non-response terminal to the allowable rate of receiving mail.

  (6) Fetch rate related to mail reception for each proxy line in a multi-stage configuration in which proxy lines are connected to the incoming mail server via the proxy server in multiple stages, and the proxy server and the incoming mail server are connected to each other. And the delivery rate of the incoming notification is determined based on the allowable rate of mail reception and the maximum fetch rate obtained for each proxy line.

  (7) In a configuration in which proxy lines are connected to the incoming mail server via a proxy server in multiple stages, and the proxy server and incoming mail server of each stage are connected to each other, the congestion level of each proxy line is monitored, Incoming call notifications are distributed preferentially to terminals connected via proxy lines with lower congestion.

  (8) In a multi-stage configuration, a means for monitoring the number of staying mails of each received mail server is further provided, and the incoming notification delivery means sets the incoming notification delivery rate for each received mail server to a received mail server having a higher mail staying rate. When all the received mail that is forwarded according to each delivery rate is relayed through each proxy line at a rate corresponding to the fetch generation rate, the proxy line with the maximum fetch rate is The mail reception rate is determined to be within the allowable rate.

  (9) In a multi-stage configuration, a means for monitoring the mail hold time of each received mail server is further provided, and the incoming notification delivery means sets the incoming notification delivery rate for each received mail server to a received mail server with a longer mail hold time. When all the received mail that is forwarded according to each delivery rate is relayed through each proxy line at a rate corresponding to the fetch generation rate, the proxy line with the maximum fetch rate is The mail reception rate is determined to be within the allowable rate.

  According to the present invention, the following effects are achieved.

  (1) The session establishment rate (allowable rate) that can be allowed in the proxy line without causing congestion at the bottleneck and without overloading the proxy server or mail server is the maximum bandwidth of the proxy line or one fetch. It depends only on the amount of transferred data per session, and does not depend on the throughput of the radio section or aggregation section. Therefore, regardless of the location of the bottleneck, the bottleneck congestion and the overloading of the proxy server and the mail server can be prevented only by regulating the mail reception rate of the proxy line within the allowable rate.

  (2) When regulating the mail reception rate of the proxy line by controlling the delivery rate of the mail arrival notification, not only the passive fetch session established in response to the mail arrival notification but also waiting for the arrival notification Since the active fetch session established by the terminal user at an arbitrary timing is also considered, the mail reception rate of the proxy line can be more reliably regulated within the allowable rate.

  (3) When restricting the mail reception rate of the proxy line by controlling the delivery rate of mail arrival notifications, the wireless terminal may be turned off or the location may be out of reach of radio waves Since the presence of a non-response terminal that cannot receive an incoming call notification or does not respond even if it can receive an incoming call notification is considered, the mail reception rate of the proxy line can be more reliably regulated within the allowable rate.

  (4) In a multi-stage configuration in which proxy lines are connected to the incoming mail server via a proxy server in multiple stages, and the proxy server and incoming mail server of each stage are connected to each other, the distribution rate of incoming notifications is the proxy line. Since it is determined based on the maximum value of the fetch rate obtained for each, even if the fetch rate is different between the proxy lines at each stage, the use band of the proxy line can be limited to the allowable band.

  (5) In a configuration in which proxy lines are connected to the incoming mail server via a proxy server in multiple stages, and the proxy server and incoming mail server at each stage are connected to each other, the proxy line passes through a proxy line with lower congestion. Incoming notifications are distributed preferentially to connected terminals, so that the mail reception rate of each proxy line can be equalized.

  (6) In a multi-stage configuration, the delivery rate of the incoming mail server with the larger number of staying mails is higher within the range that the mail receiving rate of the proxy line with the maximum fetch rate falls within the allowable rate. Effective use and uniformization of the number of staying mails between receiving mail servers can be realized.

  (7) In a multi-stage configuration, the delivery rate of the incoming mail server with the longer mail hold time is higher within the range where the proxy line mail reception rate with the maximum fetch rate falls within the allowable rate. Efficient use and uniform mail hold time between incoming mail servers can be realized.

1 is a block diagram showing a configuration of an electronic mail system to which the present invention is applied. It is the figure which showed the transmission / reception sequence of an email. It is a figure for demonstrating the fundamental view of this invention. It is the figure which showed that the allowable rate of a proxy line does not depend on the throughput of a radio area even when a radio area becomes a bottleneck. It is a functional block diagram of a 1st embodiment of the present invention. It is a functional block diagram of a 2nd embodiment of the present invention. It is a functional block diagram of a 3rd embodiment of the present invention. It is a functional block diagram of a 4th embodiment of the present invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing a network configuration of an electronic mail system to which the present invention is applied, and illustrations of components unnecessary for the description of the present invention are omitted.

  The first transmission mail server (group) 11 temporarily stores an electronic mail transmitted from the transmission terminal MNs1 via a wide area network such as the Internet. The second transmission mail server (group) 12 temporarily stores an electronic mail transmitted from the transmission terminal MNs2 via a RAN (Radio Access Network). Each mail server 11, 12 transmits the temporarily stored electronic mail to the received mail server (group) 15 corresponding to the destination address via the wide area network 13 such as the Internet and the GW / IX 14.

  When a new electronic mail is received by the received mail server (group) 15, the incoming call notification distribution device 17 distributes the incoming call notification to the destination receiving terminal MNd via the RAN. The e-mail stored in the received mail server 15 is received via the proxy server 16, the proxy line 18 and the RAN 19 in response to a received mail request (passive fetch) from each receiving terminal MNd that has received the incoming communication. Delivered to the terminal MNd.

  In addition to the passive fetch by the terminal user who responded to the incoming communication from the incoming notification delivery device 17, the terminal user performs the mail receiving operation at an arbitrary timing regardless of the incoming communication. Even when a received mail request (active fetch) is transmitted from each receiving terminal MNd to the received mail server 15, if the received mail addressed to the receiving terminal happens to be stored, it can be a trigger for distribution.

  FIG. 2 is a diagram showing an e-mail transmission / reception sequence in the above e-mail system. An e-mail transmitted from the transmission terminal MNs1 (MNs2) at time t1 is transmitted to the transmission mail server 11 (12) at time t2. After being received and stored, it is immediately sent to the received mail server 15. At time t3, the incoming mail server 15 requests the incoming notification delivery device 17 to deliver incoming notifications related to received mail. At time t4, the incoming call notification delivery apparatus 17 delivers incoming communication to the destination wireless terminal MNd in response to the request.

  At time t5, the incoming call notification is received by the destination receiving terminal MNd. At time t6, a reception mail request (passive fetch) is transmitted from the reception terminal MNd to the reception mail server 15. At time t7, the received mail server 15 transmits the stored received mail to the receiving terminal MNd in response to the received mail request. At time t8, the electronic mail is received by the receiving terminal MNd. Depending on the contents of the received mail, a reply mail may be transmitted from the receiving terminal MNd to the transmitting terminal for the received mail at time t9.

  In such an e-mail system, by restricting the incoming notification delivery request to the incoming notification delivery device 17 in the receiving server 15 or the delivery of incoming notifications to terminals having respective destinations in the incoming notification delivery device 17, Incoming mail traffic caused by the passive fetch operation can be suppressed.

  Next, the basic concept of the present invention will be described using the network configuration of FIG. 1 as an example, and then the specific configuration and operation of the present invention will be described.

  As schematically represented in FIG. 3, the e-mail receiving node is a wireless terminal, and a TCP / IP is connected to the receiving mail server group 15 via the wireless line, the RAN, the proxy line 18 and the proxy server 16. When a received mail is received by establishing a session (fetch session), the practical throughput is governed by the throughput of the bottleneck portion. Therefore, as shown in FIG. 6A, even if the throughput of the proxy line 18 is 100 Mbps, for example, if the wireless section has a bottleneck throughput of 50 kbps, for example, the practical throughput is controlled by the wireless section throughput. Will be limited to 50kbps. Although FIG. 3 shows an example in which the Proxy 16 is terminated, the same applies to a configuration in which the received mail server (group) 15 in the subsequent stage is terminated.

  Similarly, as shown in FIG. 5B, if the throughput of the proxy line 18 is 5 Mbps, for example, even if the throughput of the wireless section is 10 Mbps, the practical throughput is controlled by the throughput of the proxy line 18 and is 5 Mbps. It will be limited to.

  Here, congestion due to the increase in mail traffic is not only when the total amount of mail traffic flowing through the bottleneck exceeds the capacity of the bottleneck, but also as shown in (c) of FIG. This may also be caused by an increase in the number of received mail requests from the receiving terminal to the received mail server 15 and, as a result, the number of simultaneous connection sessions of the proxy server 16 and the received mail server 15 increases and a new session cannot be established.

  Moreover, when the number of sessions connected to the server increases, the processing speed per session decreases, so the completion time of each session is delayed, and the number of simultaneous connections increases and the processing speed per session decreases. The server will return an error or become unresponsive when establishing a session.

  In particular, in TCP, once congestion occurs, throughput may be further reduced due to frequent retransmission processing and window size reduction. Therefore, in order to prevent congestion due to mail traffic, it is necessary not only to keep the total amount of bottleneck mail traffic below the line capacity, but also to reduce the number of simultaneous connection sessions to the proxy server 16 and the mail server 15.

  Here, assuming that the upper limit of the bandwidth of the proxy line 18 is BW [bps] and the throughput of the wireless section is TP_air [kbps], the maximum number N of simultaneous connection sessions allowable to the proxy server 16 is expressed by the following equation (1). Is done.

  If the average size of received mail is size [bit] and the number of mails transferred per fetch session is “1”, the hold time T (data transfer time) per fetch session is expressed by the following equation (2). expressed.

  Further, the number of sessions that can be newly established every unit time (1 sec in the present embodiment), that is, the allowable mail reception rate λ [session / sec] is expressed by the following equation (3).

  Therefore, it can be seen that the allowable mail reception rate λ of the proxy line 18 depends only on the bandwidth upper limit BW and the transfer data amount size per fetch session, and does not depend on the throughput TP_air in the wireless section. Further, when the bottleneck is not the wireless section but the proxy line 18 or when the bottleneck is moved from the wireless section to the proxy line 18, it is the same as described above and does not depend on the throughput of the wireless section.

  FIG. 4 is a diagram mainly showing an example of the relationship between the allowable mail reception rate λ of the proxy line and the throughput TH_air of the wireless line when the wireless line becomes a bottleneck. Although the bandwidth upper limit BWmax is 90 [Mbps], the allowable bandwidth upper limit BWreal is set to 80 [Mbps] in anticipation of burst fluctuations in the throughput and the extent of restrictions. That is, the allowable band BW = BWreal. The average mail size is 15 [kByte], and one received mail is acquired by one fetch operation.

  In this case, if the throughput TP_air of the wireless section is, for example, 10 [kbps], the allowable simultaneous connection session number N is 8000 sessions from the above equation (1). Also, the hold time T per fetch session is 12 [sec] from the above equation (2). Then, the allowable mail reception rate λ is 666.67 [session / sec] from the above equation (3), which is constant regardless of the throughput TP_air in the wireless section as long as the allowable bandwidth upper limit BWreal of the proxy line is 80 [Mbps]. It is.

  In the present invention, based on the above consideration, the allowable bandwidth upper limit BW of the proxy line is set in advance, and the transfer data amount (average size of received mail) size per fetch session is determined from the history information or the monitoring result in real time. Obtaining and calculating the allowable mail reception rate λ based on these. Then, based on the allowable mail reception rate λ, for example, by controlling the distribution rate of the incoming notification by the incoming notification distribution device 17 so that the distribution rate of the incoming notification matches the allowable mail reception rate λ, Restrict the bandwidth used for proxy lines below the upper limit of allowable bandwidth.

  FIG. 5 is a functional block diagram showing the configuration of the main part of the first embodiment of the present invention. Since the same reference numerals as those described above represent the same or equivalent parts, the description thereof will be omitted.

  In the allowable rate calculation device 20, the allowable bandwidth setting unit 201 sets the allowable bandwidth BW of the proxy line 18. The transfer data amount detection unit 202 determines the average per one fetch session related to mail reception based on the data size of each mail stored in the received mail server 15 and the result of monitoring the traffic of the mail session flowing through the proxy line 18. Detects the size of transferred data. The allowable rate calculation device 20 applies the allowable bandwidth BW of the proxy line 18 and the transfer data amount size per fetch session to the above equation (3), thereby determining the mail reception rate λ allowed for the proxy line 18. calculate.

  Note that the average transfer data size per fetch session changes over time depending on the status of received emails, such as the number of emails received by users and the presence or absence of attachments. Therefore, review BW and size at predetermined intervals. The mail reception rate λ may be appropriately updated by applying the revised value to the above equation (3).

  In the incoming notification delivery device 17, the incoming notification rate determination unit 171 sets the incoming notification delivery rate λctl to the terminal having the destination of the received mail stored in the incoming mail server 15 based on the allowable mail reception rate λ. For example, a rate equal to or equivalent to the allowable rate λ for mail reception is set. The incoming call notification control unit 172 delivers the incoming call notification to the terminal having the destination of each received mail at the determined incoming call notification rate λctl.

  According to this embodiment, the allowable value of the mail reception rate that can prevent the bottleneck congestion and the proxy server and the mail server from being overloaded is determined based on the bandwidth upper limit BW of the proxy line and the transfer data amount size per fetch session. Therefore, the bottleneck congestion and the overload of the proxy server and the mail server can be prevented only by regulating the mail reception rate to the allowable value.

  FIG. 6 is a functional block diagram showing the configuration of the second embodiment of the present invention. The same reference numerals as those described above represent the same or equivalent parts, and thus the description thereof is omitted. In this embodiment, the incoming notification delivery rate λctl is determined in consideration of active fetching.

  In the incoming notification delivery device 17, the active fetch rate detection unit 173 detects an active fetch generation rate λact in which the terminal user performs a mail transmission / reception operation at an arbitrary timing regardless of the incoming mail notification.

  Such an active fetch rate λact can be obtained, for example, based on the difference between the number of incoming notifications transmitted by the incoming notification delivery device 17 and the number of fetches (whether passive or active) observed in the received mail server 15. . Further, if the incoming notification delivery apparatus 17 manages the identification information (for example, user ID) of the terminal that has transmitted the incoming notification, the active fetch rate λact is set by comparing this with the terminal identification information of each fetish session. It can be detected more accurately.

  Further, active fetch does not always generate mail traffic, and mail traffic does not occur in active fetch (empty fetch) performed when there is no received mail. Therefore, if an empty fetch is detected based on the number and ratio of active fetches as well as the presence / absence of data downloaded in each fetch session and is subtracted from the active fetch number in advance, active traffic that truly affects mail traffic The fetch rate λact can be calculated.

  Since it can be estimated that the frequency of active fetch operations by terminal users increases as the restrictions on received mail traffic become more severe, models and functions representing the relationship between the contents of received mail traffic restrictions and the active fetch rate λact are prepared in advance. Alternatively, it may be calculated based on the restriction status of received mail traffic.

  The incoming notification rate determining unit 171 determines the incoming notification delivery rate λctl to the terminal having the destination of the received mail stored in the received mail server 15 based on the allowable rate λ and the active fetch rate λact of the mail reception. For example, the difference rate (λ−λact) between the allowable rate λ of mail reception and the active fetch rate λact is determined.

  According to the present embodiment, since the incoming notification delivery rate λctl is set low by the active fetch rate λact in advance, even if an active fetch session occurs, the used bandwidth of the proxy line 18 can be kept within the allowable bandwidth upper limit. It becomes like this.

  FIG. 7 is a functional block diagram showing the configuration of the third embodiment of the present invention. The same reference numerals as those described above represent the same or equivalent parts, and thus the description thereof is omitted. In the present embodiment, the delivery rate λctl of the incoming notification is determined in consideration of the presence of a terminal that does not respond to the incoming notification delivered from the incoming notification delivery device 17.

  In the incoming call notification distribution device 17, the no-response rate detection unit 174 cannot receive the incoming call notification or receives the incoming call notification because the wireless terminal is turned off or the location is out of service area where radio waves do not reach. The generation rate λoff of the non-response terminal that does not establish the passive fetch session in response to the incoming notification, such as not responding even if it is possible, is detected.

  Such a non-response terminal rate λoff is, for example, managing the identification information of the destination terminal to which the incoming call notification distribution device 17 has transmitted the incoming call notification, and calculating the response rate by comparing this with the identification information of the response terminal. It is obtained by statistically processing history information. Alternatively, the number of unsuccessful incoming notifications by the incoming notification delivery device 17 may be monitored and the history information may be statistically processed.

  The incoming notification rate determination unit 171 sets an incoming notification delivery rate λctl to a terminal having a destination of the received mail stored in the received mail server 15 based on the allowable rate λ of mail reception and the no-response terminal rate λoff. For example, the total rate (λ + λoff) of the allowable rate λ for mail reception and the no-response terminal rate λoff is determined.

  In each of the above embodiments, the case where all the proxy lines 18 are collected in one place, or the configuration in which the proxy 16 and the proxy line 18 and the received mail server 15 are in a one-to-one relationship is taken as an example. As described with reference to FIG. 8, the proxy line 18 is connected to the received mail server 15 via the proxy server 16 in multiple stages, and the proxy server 16 and the received mail server 15 are connected to each other. It is also possible to adopt a multistage configuration connected to the.

  However, in such a multistage e-mail system, the ratio (proxy ratio) at which fetch sessions are distributed to the proxy lines 18 in each stage is not always equal. This is because it is determined in accordance with the contract time and contract contents of the terminal user whether the received mail addressed to each terminal is stored in each stage of the received mail server 15, and the received mail server This is because it is practically difficult to change 15 and the proxy 16 from the viewpoint of the cost required for the change and service continuity.

  Therefore, in such a multistage electronic mail system, for example, the received mail that is the target of the fetch request via the proxy line 18a is not necessarily stored in the received mail server 15a, but the stored contents managed by the business operator Depending on the network connection configuration, the received mail server 15b may store the received mail, and the received mail is returned from the received mail server 15b via the proxy line 18a. Therefore, the fetch rate observed on the proxy line 18a cannot represent the allowable mail reception rate of the reception mail server 15a.

  For example, when the regulated rates of the received mail servers 15a and 15b are both λideal and the allowable bandwidth upper limit of each proxy line 18a and 18b corresponding to the regulated rate λideal is λmax, the proxy lines 18a and 18b Suppose that the ratio of the fetch rate λ observed at is 0.8: 0: 2. In this case, even if both of the received mail servers 15a and 15b are operated at the regulated rate λideal, the bandwidth used for the proxy line 18a can be 1.6λmax and the bandwidth used for the proxy line 18b can be 0.4λmax. That is, the use bandwidth of the proxy line 18a having a high fetch rate λ may exceed the allowable bandwidth upper limit λmax.

  Therefore, in such an electronic mail system, it is necessary to set the delivery rate λctl of the incoming notification in accordance with the proxy line 18a having a high fetch rate λ ratio. In the above example, it is desirable that the incoming notification delivery rate λctl is set to 1 / 1.6 times that of the other embodiments.

  Further, in the electronic mail system according to the multistage configuration described above, the congestion level of each proxy line 18 is monitored, and the incoming call notification distribution device 17 receives an incoming call to a terminal connected via a proxy line with a lower congestion level. The mail reception rate may be calculated based on the permissible bandwidth of the proxy line and the amount of transfer data per fetch session under the policy of delivering notifications with priority. In this way, the mail reception rate of each proxy line 18 can be equalized.

  Further, if the rate of fetch rate λ observed on each proxy line 18a, 18b and the rate at which each fetch session is connected to each incoming mail server 15a, 15b are known, each incoming notification delivery rate λctl is received. It can be optimized according to the mail retention rate of the mail servers 15a and 15b.

  For example, the ratio of the fetch rate λ observed on each proxy line 18a, 18b is 0.8: 0: 2, and the ratio at which the fetch session is connected to each received mail server 15a, 15b is 0.5: 0.5. Then, regardless of which receiving server 15 the received mail stays in, 80% of the received mail returned from each received mail server 15a, 15b to each terminal is relayed through the proxy line 18a. 20% relay the proxy line 18b.

  Therefore, if the mail retention rate of each incoming mail server 15a, 15b is, for example, 0.4: 0: 6, the incoming notification delivery rate λctl at each receiving server 15a, 15b is set to 0.4: 0: 6. The distribution rate λctl of the received mail server 15 having a higher mail retention rate is increased. Furthermore, if the mail reception rate when 80% of all received mails transferred according to each distribution rate λctl pass through the proxy line 18a falls within the allowable rate, each proxy line 18a , 18b and the uniform number of staying mails between the receiving mail servers.

  Alternatively, if the mail hold time in each received mail server 15 can be monitored, the delivery rate λctl of the incoming notification can be optimized according to the mail retention rate of each receiving server 15a, 15b.

  For example, the ratio of the fetch rate λ observed in each proxy line 18a, 18b is 0.8: 0: 2, the ratio that each fetch session is connected to each received mail server 15a, 15b is 0.5: 0.5, and If the ratio of the mail hold time in each of the received mail servers 15a and 15b is 0.4: 0: 6, the distribution rate λctl of the incoming notification in each of the received mail servers 15a and 15b is set to 0.4: 0: 6. The delivery rate λctl of the received mail server 15 having a longer holding time is increased.

  Further, if the received mail rate when 80% of all received mails returned in response to each distribution rate λctl pass through the proxy line 18a falls within the allowable rate, each proxy line Effective use of 18a and 18b and equalization of the mail hold time between the received mail servers can be realized.

  DESCRIPTION OF SYMBOLS 11 ... 1st transmission mail server (group), 12 ... 2nd transmission mail server (group), 13 ... Wide area network, 14 ... GW / IX, 15 ... Reception mail server, 16 ... Proxy server, 17 ... Incoming notification delivery apparatus , 18 ... Proxy line, 19 ... RAN, 20 ... Allowable rate calculation device, 171 ... Incoming call notification rate determination unit, 172 ... Incoming call notification control unit, 173 ... Active fetch rate detection unit, 174 ... No response rate detection unit, 201 ... Allowable bandwidth setting unit, 202 ... Transfer data amount detection unit

Claims (9)

In the electronic mail traffic regulation device of the electronic mail system, the wireless terminal establishes a passive fetch session by connecting to the incoming mail server from the proxy line via the proxy server in response to the incoming mail notification, and acquires the incoming mail.
An allowable bandwidth setting means for setting the allowable bandwidth of the proxy line;
Transfer data amount detection means for detecting the transfer data amount per fetch session related to mail reception;
An allowable rate calculating means for calculating a mail reception rate allowed for the proxy line based on an allowable bandwidth of the proxy line and a transfer data amount per fetch session;
Means for calculating the rate of occurrence of active fetch that is established independently of the mail arrival notification;
An incoming notification rate determining means for determining a rate at which an incoming notification is distributed to a terminal having a destination of the received mail stored in the received mail server , based on the allowable rate of mail reception and the occurrence rate of active fetch ;
An e-mail traffic regulation device , comprising: an incoming notification delivery unit for delivering an incoming notification to a terminal having the destination of the received mail based on the determined delivery rate of the incoming notification . 2. The e-mail traffic according to claim 1 , wherein the incoming notification rate determining means determines the distribution rate of the incoming notification as a rate obtained by subtracting the active fetch generation rate from the allowable rate of receiving the mail. Regulatory device. In the electronic mail traffic regulation device of the electronic mail system, the wireless terminal establishes a passive fetch session by connecting to the incoming mail server from the proxy line via the proxy server in response to the incoming mail notification, and acquires the incoming mail.
An allowable bandwidth setting means for setting the allowable bandwidth of the proxy line;
Transfer data amount detection means for detecting the transfer data amount per fetch session related to mail reception;
An allowable rate calculating means for calculating a mail reception rate allowed for the proxy line based on an allowable bandwidth of the proxy line and a transfer data amount per fetch session;
Means for obtaining a rate of occurrence of a non-response terminal that does not establish a passive fetch session for the incoming notification;
An incoming notification rate determining means for determining a rate at which an incoming notification is distributed to a terminal having a destination of the received mail stored in the received mail server , based on the allowable rate of mail reception and the generation rate of a non-response terminal ;
An e-mail traffic regulation device , comprising: an incoming notification delivery unit for delivering an incoming notification to a terminal having the destination of the received mail based on the determined delivery rate of the incoming notification . 4. The e-mail according to claim 3 , wherein the incoming call notification rate determining means determines the distribution rate of the incoming call notification to a rate obtained by adding the generation rate of the non-response terminal to the allowable rate for receiving the mail. Traffic control device. A means for detecting an empty fetch in which received mail corresponding to active fetch is not accumulated;
The incoming call notification rate determining means, a delivery rate of said incoming notification, e-mail traffic regulation device according to any one of claims 1, characterized in that reducing the number of the air fetched from an active fetch number 4. A path for connecting the proxy line to the incoming mail server via a proxy server is provided in multiple stages,
Means for connecting the proxy server and the incoming mail server to each other;
Means for calculating a fetch generation rate for proxy line mail reception for each stage,
2. The e-mail traffic regulation device according to claim 1, wherein the incoming notification rate determining unit determines a delivery rate of the incoming notification based on an allowable rate of a proxy line having a maximum fetch rate. A path for connecting the proxy line to the incoming mail server via a proxy server is provided in multiple stages,
Means for connecting the proxy server and the incoming mail server to each other;
Means for monitoring the degree of congestion of each proxy line,
The incoming notification rate determining means determines an incoming notification delivery rate based on a policy of preferentially delivering incoming notifications to a terminal connected via a proxy line having a lower congestion degree. The electronic mail traffic regulation device according to claim 1. A path for connecting the proxy line to the incoming mail server via a proxy server is provided in multiple stages,
Means for connecting the proxy server and the incoming mail server to each other;
Means for calculating the rate of fetches related to proxy line mail reception for each stage;
Means for monitoring the number of staying mails in each of the received mail servers,
The incoming notification rate determination means increases the distribution rate of incoming notifications related to each received mail server, increases the distribution rate λctl of the received mail server having a higher mail retention rate, and transfers all of the data transferred according to each distribution rate. The determination is made such that when the received mail is relayed through each proxy line at a rate corresponding to the fetch generation rate, the mail reception rate of the proxy line having the maximum fetch rate falls within an allowable rate. The electronic mail traffic control apparatus according to 1. A path for connecting the proxy line to the incoming mail server via a proxy server is provided in multiple stages,
Means for connecting the proxy server and the incoming mail server to each other;
Means for calculating the rate of fetches related to proxy line mail reception for each stage;
Means for monitoring the mail hold time in each of the received mail servers,
The incoming notification rate determination means increases the distribution rate of incoming notifications for each received mail server, increases the distribution rate λctl of the received mail server with a longer mail hold time, and transfers all of the data transferred according to each distribution rate. The determination is made such that when the received mail is relayed through each proxy line at a rate corresponding to the fetch generation rate, the mail reception rate of the proxy line having the maximum fetch rate falls within an allowable rate. The electronic mail traffic control apparatus according to 1.