CA2024305C - Control of overload in communications networks - Google Patents
Control of overload in communications networksInfo
- Publication number
- CA2024305C CA2024305C CA002024305A CA2024305A CA2024305C CA 2024305 C CA2024305 C CA 2024305C CA 002024305 A CA002024305 A CA 002024305A CA 2024305 A CA2024305 A CA 2024305A CA 2024305 C CA2024305 C CA 2024305C
- Authority
- CA
- Canada
- Prior art keywords
- gap
- value
- threshold
- fraction
- measuring
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/12—Avoiding congestion; Recovering from congestion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/26—Flow control; Congestion control using explicit feedback to the source, e.g. choke packets
- H04L47/266—Stopping or restarting the source, e.g. X-on or X-off
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q3/00—Selecting arrangements
- H04Q3/0016—Arrangements providing connection between exchanges
- H04Q3/0029—Provisions for intelligent networking
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q3/00—Selecting arrangements
- H04Q3/0016—Arrangements providing connection between exchanges
- H04Q3/0062—Provisions for network management
- H04Q3/0091—Congestion or overload control
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q2213/00—Indexing scheme relating to selecting arrangements in general and for multiplex systems
- H04Q2213/13522—Indexing scheme relating to selecting arrangements in general and for multiplex systems traffic management
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q2213/00—Indexing scheme relating to selecting arrangements in general and for multiplex systems
- H04Q2213/13561—Indexing scheme relating to selecting arrangements in general and for multiplex systems congestion - inc. overflow
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q2213/00—Indexing scheme relating to selecting arrangements in general and for multiplex systems
- H04Q2213/13563—Indexing scheme relating to selecting arrangements in general and for multiplex systems call gapping, e.g. to prevent congestion
Abstract
Abstract This invention relates to a method and apparatus for throttling request messages to an overloaded database and for throttling overload traffic to a particular destination. The database sends a gap value to the sources of messages or traffic;
following any message or call to the destination, no messages will be sent or calls to the destination completed for one gap interval. In a departure from the prior art, the gap interval, a function of the number of request messages flushed or overload calls to the destination is not recurrently calculated anew, but is instead recurrently adjusted from its previous value. Advantageously, this arrangement permits a relatively stable value of the gap to be maintained and reduces overhead for preliminary processing of request messages that cannot be handled.
following any message or call to the destination, no messages will be sent or calls to the destination completed for one gap interval. In a departure from the prior art, the gap interval, a function of the number of request messages flushed or overload calls to the destination is not recurrently calculated anew, but is instead recurrently adjusted from its previous value. Advantageously, this arrangement permits a relatively stable value of the gap to be maintained and reduces overhead for preliminary processing of request messages that cannot be handled.
Description
2 0 2 '~
CONTROL OF OVERLOAD IN COMMUNICATIONS Nl~TWORKS
Technical Field This invention relates to a method and apparatus for controlling overload in communications networks.
5 Problem Centralized databases have become critical to the operation of modern communications networks. For many of the services offered by these networks suchas "800" or "Freephone" service, a call cannot be completed without obtaining the information for routing the call to its destination from such a centralized database.
10 When an 800 call is placed, the call is routed from a local switching system to a toll switching system and the toll switching system sends a query message to one of aplurality of centralized databases. One or more signal transfer points are interposed between the toll switch and the plurality of databases to route the request message to the appropriate one of these centralized databases based on a deffination group. The 15 centralized database responds with the information necessary for routing the call. If the centralized database fai1s to respond within a prespecified interval, then the call is routed to overflow tone, the calling customer is annoyed and will usually re-attempt the call shordy. These re-attempts cause further load to be placed on the centralized database thus aggravating an exisdng overload condition. It is therefore 20 important to ensure that overloads to centralized databases are carefully controlled in order to avoid creating this aggravated condition.
In the prior art systems, a database handles overload in approximately the following manner: the database Iceeps track of the number of request messages received and maintains a parameter of the approximate number of messages which 25 can be processed per unit time. If more messages are received than can be processed per unit time, the excess messages are discarded or "flushed". The process of flushing messages takes an appreciable amount of real time, however, so that in conditions of extreme overload, a large fraction of the real time of the database is actually spent in flushing messages which will not be processed. In addition, 30 measures are taken to try to control messages at the sources which send the query messages. These sources are the various toll switches of the communication network that received the traffic and generate the request messages to the database. In the prior art, when the database is in overload and receives a message from a particular toll switch, the message being a request for roudng to a particular destination group, 35 the database as part of its response message sends an indication to the toll switch to .
- .
: - , :
2a2~3~
reduce the message load for traffic to that destination group in accordance withparameters supplied in the response message. These parameters are used by the toll switch to look up a gap and a duration time interval which are used as follows.
Following the acceptance of a call, no more calls to that destination group will be S accepted for one gap interval. Thereafter, another call will be accepted and then no more calls will be accepted for of another gap interval. These gap intervals areapplied during the length of time of the duration interval unless changed by a subsequent message. These controls are applied individually per destination group, rather than collectively to all destination groups served by the toll switch, because in 10 many cases these overloads are generated by flurries of traffic to a particular destination, especially flurries of traffic to a destination which itself is overloaded and which cannot accept more calls. Because only overloaded databases request control, the toll switches will only control messages destined for overloaded databases and will not affect messages destined for other databases which are not 15 overloaded. Further, among the destination groups served by the overloaded databases, there will be a tendency to control those destinations which themselves are in overload most strongly, since in a gapping control a much larger fraction of calls are discarded from heavy traffic sources than from light traffic sources. Thus, only that portion of the traffic which cannot bo adequately served is subjected to the 20 overload control.
Since a database can be subjected to a wide variety of overload condidons, some mild, some extreme, some focused from a few sources, some generated by a large number of sources, the particular gap and duration intervalrequired to properly control traffic can differ widely from one overload to the next.
25 To cover the potential range from 0.1 second to 600 seconds the toll switches are equipped with a zero size gap plus lS gap and duration interval sizes which differ by approximately a factor of 2. Since choice of the wrong gap and duration intervals will result in either undercontrol or overcontrol, it is important that the database be able to select the best values for the given situation. In the prior art, the choice was 30 made based on a table lookup using the fraction of flushed messages as tho index.
A problem with this overall arrangement is that in order to retain a large gap it is necessary to handle a situation in which the database is subject to high overload from a large numbor of sources; when this happens, the database must reject large numbers of calls, thus wasting much of its time and reducing the number 35 of request messages which it can process. Moreover, the arrangement tends to bo highly unstable, swinging from excessively heavy overload control to excessively ~243~
CONTROL OF OVERLOAD IN COMMUNICATIONS Nl~TWORKS
Technical Field This invention relates to a method and apparatus for controlling overload in communications networks.
5 Problem Centralized databases have become critical to the operation of modern communications networks. For many of the services offered by these networks suchas "800" or "Freephone" service, a call cannot be completed without obtaining the information for routing the call to its destination from such a centralized database.
10 When an 800 call is placed, the call is routed from a local switching system to a toll switching system and the toll switching system sends a query message to one of aplurality of centralized databases. One or more signal transfer points are interposed between the toll switch and the plurality of databases to route the request message to the appropriate one of these centralized databases based on a deffination group. The 15 centralized database responds with the information necessary for routing the call. If the centralized database fai1s to respond within a prespecified interval, then the call is routed to overflow tone, the calling customer is annoyed and will usually re-attempt the call shordy. These re-attempts cause further load to be placed on the centralized database thus aggravating an exisdng overload condition. It is therefore 20 important to ensure that overloads to centralized databases are carefully controlled in order to avoid creating this aggravated condition.
In the prior art systems, a database handles overload in approximately the following manner: the database Iceeps track of the number of request messages received and maintains a parameter of the approximate number of messages which 25 can be processed per unit time. If more messages are received than can be processed per unit time, the excess messages are discarded or "flushed". The process of flushing messages takes an appreciable amount of real time, however, so that in conditions of extreme overload, a large fraction of the real time of the database is actually spent in flushing messages which will not be processed. In addition, 30 measures are taken to try to control messages at the sources which send the query messages. These sources are the various toll switches of the communication network that received the traffic and generate the request messages to the database. In the prior art, when the database is in overload and receives a message from a particular toll switch, the message being a request for roudng to a particular destination group, 35 the database as part of its response message sends an indication to the toll switch to .
- .
: - , :
2a2~3~
reduce the message load for traffic to that destination group in accordance withparameters supplied in the response message. These parameters are used by the toll switch to look up a gap and a duration time interval which are used as follows.
Following the acceptance of a call, no more calls to that destination group will be S accepted for one gap interval. Thereafter, another call will be accepted and then no more calls will be accepted for of another gap interval. These gap intervals areapplied during the length of time of the duration interval unless changed by a subsequent message. These controls are applied individually per destination group, rather than collectively to all destination groups served by the toll switch, because in 10 many cases these overloads are generated by flurries of traffic to a particular destination, especially flurries of traffic to a destination which itself is overloaded and which cannot accept more calls. Because only overloaded databases request control, the toll switches will only control messages destined for overloaded databases and will not affect messages destined for other databases which are not 15 overloaded. Further, among the destination groups served by the overloaded databases, there will be a tendency to control those destinations which themselves are in overload most strongly, since in a gapping control a much larger fraction of calls are discarded from heavy traffic sources than from light traffic sources. Thus, only that portion of the traffic which cannot bo adequately served is subjected to the 20 overload control.
Since a database can be subjected to a wide variety of overload condidons, some mild, some extreme, some focused from a few sources, some generated by a large number of sources, the particular gap and duration intervalrequired to properly control traffic can differ widely from one overload to the next.
25 To cover the potential range from 0.1 second to 600 seconds the toll switches are equipped with a zero size gap plus lS gap and duration interval sizes which differ by approximately a factor of 2. Since choice of the wrong gap and duration intervals will result in either undercontrol or overcontrol, it is important that the database be able to select the best values for the given situation. In the prior art, the choice was 30 made based on a table lookup using the fraction of flushed messages as tho index.
A problem with this overall arrangement is that in order to retain a large gap it is necessary to handle a situation in which the database is subject to high overload from a large numbor of sources; when this happens, the database must reject large numbers of calls, thus wasting much of its time and reducing the number 35 of request messages which it can process. Moreover, the arrangement tends to bo highly unstable, swinging from excessively heavy overload control to excessively ~243~
light overload control because a large gap sharply reduces the number of requestmessages, leading to a much lower flush fraction and gap, leading to another surge in traffic, and so on, with the result that many more request messages are rejected than under opdmum operation. A problem of the prior art, therefore, is that the 5 arrangements for controlling overload of request messages received at centralized databases are such that the centralized database does not process messages up to its capacity.
Solution In accordance with the principles of this invendon, an advance is made 10 over the prior art by having an arrangement wherein the database adjusts the present value of a gap up or down in accordance with the present fracdon of flushed request messages instead of setdng an absolute value of the gap. Advantageously, such anarrangement permits the database to dynamically adjust the gap va1ue in small increments, constandy searching for the optimum value of the gap. Advantageously, 15 it is not necessary for the database to waste substandal resources in rejecdng a high fracdon of incoming messages in order to maintain a high value of a gap necessary to reduce dhe incoming message volume to a value dhat can be handled by dhe database; most calls that need to be rejected are rejected by the toll switch without ever sending a message to the database. By rejecdng calls at the switch, dhe signal 20 transfer points are also less heavily loaded since they do not need to route large numbers of messages which will subsequently be rejected at the database.
In accordance with one aspect of the invention, the solected gap size is such as to cause some fraction of message traffic to be rejectcd at the database if the database is not in fact capable of processing all of the traffic offered to the toll 25 switches. By selecting that gap size, the amount of traffic unnecessarily rejected at the toll switch is minimized. The opdmum choice of gap is the largest gap which sdll rejects some messages at the database since with this gap no calls are unnecessarily denied at the toll switch. Thus, whenever the amount of request messages rejected at the database goes to 0, the gap slze, if not already 0, is 30 decreased by one increment.
In accordance with another aspect of the invention, the parameter reflecting the fraction of messages flushed at the database before increasing the gap æize is a funcdon of the rado between adjacent gap sizes. Thus, when adjacent values of the gap differ by only a small fraction, the gap is increased when only a 35 small fracdon of request messages are flushed.
;' ., , ' -`` 202~
In accordance with another aspect of the invention, the time between successive potential gap changes is proportional to the absolute va1ue of the gap size currently in effect. Thus, for larger values of the gap, the gap cannot be increased or decreased as frequently as when the gap value is small. In accordance with another S aspect of this invention, the proportionality factor before accepting a decision to increase the gap is higher than the proportionality factor for accepting a decision to decrease the gap. This permits rapid increase of the gap size for a sudden overload, while at the same time avoids overshoot, and allows for a smoother recovery fromoverload. This helps to eliminate large swings in the gap value and provide better 10 performance.
In accordance with another aspect of the invention, the duration interval is chosen proportional to and several times longer than the gap interval. This avoids the problem of a source inadvertently becoming uncontrolled because the durationtimer expired before a subsequent response message was received.
In accordance with another aspect of the invention, if the toll switches that generate the messages to the database are only prepared to work with a small number of discrete gaps separated by a large multiple, then the value of the gap that is returned in messages from the database is selected to be one of two adjacent gap values and the fraction of the dme during which a lower gap value and a higher gap 20 value is returned is adjusted in an attempt to have an average gap value that is close to optimum. This permits the database to work effectively with a larger number of discrete values of the gap even though the toll switches use a much sma11er number of values, and therefore, allows for a more continuous range of gap control. Theideal gap size in overload is an almost continuously varying number. In accordance 25 with one specific embodiment of the invention, the gap value is adjusted every ten seconds. The toll switches accept one of 15 gap values in addition to the gap value 0. The database calculates one of 75 gap values and then sends the adjacent values of gap representing a pa ticular value by a proportionate fraction of the time. For example, if two adjacent gap values at the toll switches are one second and two 30 seconds and the calculated gap value at the database is 1.6 seconds, then the one second value is sent out 40% of the time and the two second value is sent out 60% of the time.
In accordance with another aspect of the invention, at the time when an adjustment of the gap value is calculated, in addition to the three basic values, 35 namely, increase the gap value by one increment, keep the present gap value, or decrease the gap value by one increment, there are additional ranges in which the gap . - :., ; , .
: . . . .. : , .
: ~ , ---` 2~2~3~
value is increased by two or more. This is especially desirable for the arrangement described above which has 75 different gap values since with such a large number of different gap values, it may be desirable to increase by more than one increment per ten seconds.
S Brief Description of the Drawin~
FIG. 1 shows the message interconnection among a group of toll switches and database for requesting data and for controlling a gap used for rejecting traffic;
FIG. 2 shows how a gap is used to reject traffic as a functdon of dme;
10 and E~IGS. 3 and 4 are flow diagrams of programs for controlling the gap.
Detailed Description FIG. 1 is an overall block diagram of the prior art arrangement for communicadng among toll switches 104,...,105 via one or more signal transfer 15 points (STP) 103 to one or more databases 101,...,102. The toll switches which receive incoming calls from local switches such as local switch 106, send request messages 110 to a database such as database 101. The request message 110 comprises an identifier of the toll switch 111, a destinadon group 112, the rest of the number dialed by the customer 113, and a call identifier 114 to associate a 20 subsequent response message with the proper call. The desdnadon group index 112 is used by the STP to route the request message to the appropriate database if more than one database is required for handling the total volume of request messages. In this case,the request message is routed by the signal transfer point to database 101, which receives the request message and responds with response message 12û~
25 Database 101 comprises a control processor 150, controlled by a program 151 stored in memory of processor 150, and a data link controller 152 for transmitting and receiving data messages to and from toll switches 104,...,105 via STP 103. Response message 120 includes a toll switch idendfier 121, necessary in this case to route the message back via the STP to the appropriate toll switch, routing data 122 supplied by 30 the database to route the call properly, the call identifier 123 equivalent to the call idendfier 114 of the request message, a gap value index 124, and a duration value index 125 to be used by the toll switch for reguladng traffic to the destinadon group idendSed by desdnadon group idendfier 126. Although each response message specifies a new gap value provided via the gap value index 124, since the database 35 may flush any given message, the duradon index 125 is also provided to maintain a gap, in the absence of new messages, for a period much longer than one gap interval.
. : . , : .
. . ' . ' ' ~ , -- ~ ' ' ' '"' , : ~ ' ' ' ' ;
.. .
~2~
FIG. 2 shows how the gap and duration values are used in the toll switch. Assume that at dme 210 a response message is received by toll switch 105from database 101 as a result of a database overload, giving non-zero index values to allow the toll switch 105 to deterrnine the size of the gap and the length of the S duration. Overload control will then be invoked starting at time 210 and lasting until time 215, times 210 and 215 being separated by the length of the duration as specified by the duration index, unless another response message is received before the duration interval expires; if another response message is received, new values of gap and duration are applied. The gap is turned on initially on receipt of the 10 response message. Assume that during the time of gap 201 that a call arrives at time æl for the destinadon group for which the overload control response messagehas been received. That call will then be rejected; the toll switch will return overflow signal and the call will not be completed. After the gap 201 is completed, the next incoming call to that destination group will be accepted. The first incoming 15 call to that destination group on FIG. 2 is the call arriving at time 222. Following the processing of accepted call 222, another gap 202 is automatically inserted. Any call such as the call 223 occurring during gap 202 will be rejected. The response message for accepted call 222 will provide a new value of the gap and duration index from which new values of the gap and duration may be calculated. These new 20 values will be used and effectively the duration time will start from the time that the response message for call 222 is received. Effectively, a gap is generated after every accepted call during overload as well as after every response message containing a non-zero value is received. Both the value of the gap and the value of the duration are updated with each response message for that destination group. PIG. 2 is also in 25 accordance with the principles of the prior art.
PIG. 3 is a flow diagram of a program, part of program 151, used to implement applicants' invention. Block 300 shows that the gap index is initialized to zero whenever program 151 is rebooted. Block 305 shows that the database is periodically computing the fraction of re~uest messages currently being flushed, i e., 30 not being processed but being rejected. Every ten seconds, as indicated by block 310, the system checks whether sufficient time has elapsed since the last time the gap index was recalculated for the database. If test 311 indicates that sufficient time has not elapsed, the program exits (block 312). If sufficient time has elapsed, then the flush fraction is read (action block 314). If this flush fraction exceeds a 35 parameter C, then the gap index is incremented by two. If the flush fraction is between C and a lower parameter B, then the gap index is incremented by one. If the .' ' . - , ..
- , .
, -, 2~3~
flush fraction is between B and a lower parameter A, then the gap index is Tetained.
If the flush fraction is less than A, the gap index is decremented by one (unless the gap index is already 0). The wait timer is then reset (action block 316) to ensure that enough time will elapse before the next recalculation of the gap index, and the 5 program exits (block 317).
In order to have optimum performance for this algorithm, for higher values of the gap, the gap index should be recalculated less frequently. Therefore, test 311 is arranged so ~at it checks the value of the gap index and converts this into a wait time before making a comparison with the present value of the wait timer.10 The conversion from the gap index to the wait time is such that the wait time is approximately proportional to the actual gap. In the specific implementation written by the inventors, the wait time is approximately three times the value of the gap.
The gap has a large range of values from .1 to 600 seconds. The toll switch selects a gap value on the basis of a 4-bit index supplied by the database in lS field 124 of response message 120. The database also keeps track of the length of the gap by maintaining its own gap index, which in addition to the 16 possible values that are transmitted to toll switch, has four intermediate values between each pair of such va1ues. This index is used by the database in test 311 to determine whether the elapsed time exceeds the waiting time for a particular gap va1ue and is used to 20 generate the specific gap index that is transmitted in response messages to the toll switch. FIG. 4, a flow diagram of another part of program 151, illustrates this latter process. Every response message includes a gap and a duration index. During periods of light load, these gap and duration indexes are always 0~ At any particular instant, the same gap index and same duration index are sent from a given database 25 to any toll switch on any response message for any destination group, seNed by that database. Block 400 of ~IG. 4 illustrates that the database gap index (which includes four intermediate values between each of the 16 values that may be sent to a toll switch) is converted into one of the 16 values that will be sent to toll switches during the next 200 milliseconds. ~or example, if the database gap index were 3.4, then30 during a one second inteNal the value of the gap index of three would be sent during three 200 millisecond intervals and the value four would be sent during the other two 200 millisecond intervals. Block 40S shows that given the gap index, a duration index is selected such that the duration inteNal will be approximately 8 times the gap inteNal. As illustrated in block 410, a switch gap index and duration index are 35 transmitted with each response message.
.
'~ , .
. . .. .. .
- ~ , -.. .
2 ~
The basic principles of this invention as illustrated herein, i.e., that of adjusting a gap index and therefore indirectly adjusting a gap instead of calculating the absolute value of a gap can also be used for other types of over10ad control. One of the big problems that faces a communications network administrator is that there S are occasional flurries of traffic to destinadons which are unprepared to receive such flurries. The extra traffic to these desdnadons is unproductive and ties up network resources. In order to keep such traffic from the network, it is desirable to detect the presence of such excess traffic and stop such excess traffic from leaving the toll switches. For such situations, the database can detect whether excess traffic to any 10 specific destinatdon is being generated in the network. The database can keep track of how many request messages have been encountered for a particular destination and how many such request messages should be present during periods of high loadbut not overload. If the number of request messages exceeds this high load value, then the database can invoke gap control for that particular destination by sending a 15 message to each of the toll switches requesting that the toll switch apply gap control to the pardcular 10-digit number destination for which the overload has been detected. The administradon of the gap for that 10-digit number is then performed in the database in accordance with the principles of this invendon as described with respect to FIG. 3.
It is to be understood that the above description is only of one preferred embodiment of the inventdon. Numerous other arrangements may be devised by one skilled in the art without departing from the spirit and scope of the invendon. The invention is thus limited only as defined in the accompanying claims.
...
,
Solution In accordance with the principles of this invendon, an advance is made 10 over the prior art by having an arrangement wherein the database adjusts the present value of a gap up or down in accordance with the present fracdon of flushed request messages instead of setdng an absolute value of the gap. Advantageously, such anarrangement permits the database to dynamically adjust the gap va1ue in small increments, constandy searching for the optimum value of the gap. Advantageously, 15 it is not necessary for the database to waste substandal resources in rejecdng a high fracdon of incoming messages in order to maintain a high value of a gap necessary to reduce dhe incoming message volume to a value dhat can be handled by dhe database; most calls that need to be rejected are rejected by the toll switch without ever sending a message to the database. By rejecdng calls at the switch, dhe signal 20 transfer points are also less heavily loaded since they do not need to route large numbers of messages which will subsequently be rejected at the database.
In accordance with one aspect of the invention, the solected gap size is such as to cause some fraction of message traffic to be rejectcd at the database if the database is not in fact capable of processing all of the traffic offered to the toll 25 switches. By selecting that gap size, the amount of traffic unnecessarily rejected at the toll switch is minimized. The opdmum choice of gap is the largest gap which sdll rejects some messages at the database since with this gap no calls are unnecessarily denied at the toll switch. Thus, whenever the amount of request messages rejected at the database goes to 0, the gap slze, if not already 0, is 30 decreased by one increment.
In accordance with another aspect of the invention, the parameter reflecting the fraction of messages flushed at the database before increasing the gap æize is a funcdon of the rado between adjacent gap sizes. Thus, when adjacent values of the gap differ by only a small fraction, the gap is increased when only a 35 small fracdon of request messages are flushed.
;' ., , ' -`` 202~
In accordance with another aspect of the invention, the time between successive potential gap changes is proportional to the absolute va1ue of the gap size currently in effect. Thus, for larger values of the gap, the gap cannot be increased or decreased as frequently as when the gap value is small. In accordance with another S aspect of this invention, the proportionality factor before accepting a decision to increase the gap is higher than the proportionality factor for accepting a decision to decrease the gap. This permits rapid increase of the gap size for a sudden overload, while at the same time avoids overshoot, and allows for a smoother recovery fromoverload. This helps to eliminate large swings in the gap value and provide better 10 performance.
In accordance with another aspect of the invention, the duration interval is chosen proportional to and several times longer than the gap interval. This avoids the problem of a source inadvertently becoming uncontrolled because the durationtimer expired before a subsequent response message was received.
In accordance with another aspect of the invention, if the toll switches that generate the messages to the database are only prepared to work with a small number of discrete gaps separated by a large multiple, then the value of the gap that is returned in messages from the database is selected to be one of two adjacent gap values and the fraction of the dme during which a lower gap value and a higher gap 20 value is returned is adjusted in an attempt to have an average gap value that is close to optimum. This permits the database to work effectively with a larger number of discrete values of the gap even though the toll switches use a much sma11er number of values, and therefore, allows for a more continuous range of gap control. Theideal gap size in overload is an almost continuously varying number. In accordance 25 with one specific embodiment of the invention, the gap value is adjusted every ten seconds. The toll switches accept one of 15 gap values in addition to the gap value 0. The database calculates one of 75 gap values and then sends the adjacent values of gap representing a pa ticular value by a proportionate fraction of the time. For example, if two adjacent gap values at the toll switches are one second and two 30 seconds and the calculated gap value at the database is 1.6 seconds, then the one second value is sent out 40% of the time and the two second value is sent out 60% of the time.
In accordance with another aspect of the invention, at the time when an adjustment of the gap value is calculated, in addition to the three basic values, 35 namely, increase the gap value by one increment, keep the present gap value, or decrease the gap value by one increment, there are additional ranges in which the gap . - :., ; , .
: . . . .. : , .
: ~ , ---` 2~2~3~
value is increased by two or more. This is especially desirable for the arrangement described above which has 75 different gap values since with such a large number of different gap values, it may be desirable to increase by more than one increment per ten seconds.
S Brief Description of the Drawin~
FIG. 1 shows the message interconnection among a group of toll switches and database for requesting data and for controlling a gap used for rejecting traffic;
FIG. 2 shows how a gap is used to reject traffic as a functdon of dme;
10 and E~IGS. 3 and 4 are flow diagrams of programs for controlling the gap.
Detailed Description FIG. 1 is an overall block diagram of the prior art arrangement for communicadng among toll switches 104,...,105 via one or more signal transfer 15 points (STP) 103 to one or more databases 101,...,102. The toll switches which receive incoming calls from local switches such as local switch 106, send request messages 110 to a database such as database 101. The request message 110 comprises an identifier of the toll switch 111, a destinadon group 112, the rest of the number dialed by the customer 113, and a call identifier 114 to associate a 20 subsequent response message with the proper call. The desdnadon group index 112 is used by the STP to route the request message to the appropriate database if more than one database is required for handling the total volume of request messages. In this case,the request message is routed by the signal transfer point to database 101, which receives the request message and responds with response message 12û~
25 Database 101 comprises a control processor 150, controlled by a program 151 stored in memory of processor 150, and a data link controller 152 for transmitting and receiving data messages to and from toll switches 104,...,105 via STP 103. Response message 120 includes a toll switch idendfier 121, necessary in this case to route the message back via the STP to the appropriate toll switch, routing data 122 supplied by 30 the database to route the call properly, the call identifier 123 equivalent to the call idendfier 114 of the request message, a gap value index 124, and a duration value index 125 to be used by the toll switch for reguladng traffic to the destinadon group idendSed by desdnadon group idendfier 126. Although each response message specifies a new gap value provided via the gap value index 124, since the database 35 may flush any given message, the duradon index 125 is also provided to maintain a gap, in the absence of new messages, for a period much longer than one gap interval.
. : . , : .
. . ' . ' ' ~ , -- ~ ' ' ' '"' , : ~ ' ' ' ' ;
.. .
~2~
FIG. 2 shows how the gap and duration values are used in the toll switch. Assume that at dme 210 a response message is received by toll switch 105from database 101 as a result of a database overload, giving non-zero index values to allow the toll switch 105 to deterrnine the size of the gap and the length of the S duration. Overload control will then be invoked starting at time 210 and lasting until time 215, times 210 and 215 being separated by the length of the duration as specified by the duration index, unless another response message is received before the duration interval expires; if another response message is received, new values of gap and duration are applied. The gap is turned on initially on receipt of the 10 response message. Assume that during the time of gap 201 that a call arrives at time æl for the destinadon group for which the overload control response messagehas been received. That call will then be rejected; the toll switch will return overflow signal and the call will not be completed. After the gap 201 is completed, the next incoming call to that destination group will be accepted. The first incoming 15 call to that destination group on FIG. 2 is the call arriving at time 222. Following the processing of accepted call 222, another gap 202 is automatically inserted. Any call such as the call 223 occurring during gap 202 will be rejected. The response message for accepted call 222 will provide a new value of the gap and duration index from which new values of the gap and duration may be calculated. These new 20 values will be used and effectively the duration time will start from the time that the response message for call 222 is received. Effectively, a gap is generated after every accepted call during overload as well as after every response message containing a non-zero value is received. Both the value of the gap and the value of the duration are updated with each response message for that destination group. PIG. 2 is also in 25 accordance with the principles of the prior art.
PIG. 3 is a flow diagram of a program, part of program 151, used to implement applicants' invention. Block 300 shows that the gap index is initialized to zero whenever program 151 is rebooted. Block 305 shows that the database is periodically computing the fraction of re~uest messages currently being flushed, i e., 30 not being processed but being rejected. Every ten seconds, as indicated by block 310, the system checks whether sufficient time has elapsed since the last time the gap index was recalculated for the database. If test 311 indicates that sufficient time has not elapsed, the program exits (block 312). If sufficient time has elapsed, then the flush fraction is read (action block 314). If this flush fraction exceeds a 35 parameter C, then the gap index is incremented by two. If the flush fraction is between C and a lower parameter B, then the gap index is incremented by one. If the .' ' . - , ..
- , .
, -, 2~3~
flush fraction is between B and a lower parameter A, then the gap index is Tetained.
If the flush fraction is less than A, the gap index is decremented by one (unless the gap index is already 0). The wait timer is then reset (action block 316) to ensure that enough time will elapse before the next recalculation of the gap index, and the 5 program exits (block 317).
In order to have optimum performance for this algorithm, for higher values of the gap, the gap index should be recalculated less frequently. Therefore, test 311 is arranged so ~at it checks the value of the gap index and converts this into a wait time before making a comparison with the present value of the wait timer.10 The conversion from the gap index to the wait time is such that the wait time is approximately proportional to the actual gap. In the specific implementation written by the inventors, the wait time is approximately three times the value of the gap.
The gap has a large range of values from .1 to 600 seconds. The toll switch selects a gap value on the basis of a 4-bit index supplied by the database in lS field 124 of response message 120. The database also keeps track of the length of the gap by maintaining its own gap index, which in addition to the 16 possible values that are transmitted to toll switch, has four intermediate values between each pair of such va1ues. This index is used by the database in test 311 to determine whether the elapsed time exceeds the waiting time for a particular gap va1ue and is used to 20 generate the specific gap index that is transmitted in response messages to the toll switch. FIG. 4, a flow diagram of another part of program 151, illustrates this latter process. Every response message includes a gap and a duration index. During periods of light load, these gap and duration indexes are always 0~ At any particular instant, the same gap index and same duration index are sent from a given database 25 to any toll switch on any response message for any destination group, seNed by that database. Block 400 of ~IG. 4 illustrates that the database gap index (which includes four intermediate values between each of the 16 values that may be sent to a toll switch) is converted into one of the 16 values that will be sent to toll switches during the next 200 milliseconds. ~or example, if the database gap index were 3.4, then30 during a one second inteNal the value of the gap index of three would be sent during three 200 millisecond intervals and the value four would be sent during the other two 200 millisecond intervals. Block 40S shows that given the gap index, a duration index is selected such that the duration inteNal will be approximately 8 times the gap inteNal. As illustrated in block 410, a switch gap index and duration index are 35 transmitted with each response message.
.
'~ , .
. . .. .. .
- ~ , -.. .
2 ~
The basic principles of this invention as illustrated herein, i.e., that of adjusting a gap index and therefore indirectly adjusting a gap instead of calculating the absolute value of a gap can also be used for other types of over10ad control. One of the big problems that faces a communications network administrator is that there S are occasional flurries of traffic to destinadons which are unprepared to receive such flurries. The extra traffic to these desdnadons is unproductive and ties up network resources. In order to keep such traffic from the network, it is desirable to detect the presence of such excess traffic and stop such excess traffic from leaving the toll switches. For such situations, the database can detect whether excess traffic to any 10 specific destinatdon is being generated in the network. The database can keep track of how many request messages have been encountered for a particular destination and how many such request messages should be present during periods of high loadbut not overload. If the number of request messages exceeds this high load value, then the database can invoke gap control for that particular destination by sending a 15 message to each of the toll switches requesting that the toll switch apply gap control to the pardcular 10-digit number destination for which the overload has been detected. The administradon of the gap for that 10-digit number is then performed in the database in accordance with the principles of this invendon as described with respect to FIG. 3.
It is to be understood that the above description is only of one preferred embodiment of the inventdon. Numerous other arrangements may be devised by one skilled in the art without departing from the spirit and scope of the invendon. The invention is thus limited only as defined in the accompanying claims.
...
,
Claims (15)
1. In a first system for controlling a gap in a plurality of second systems using gap control to reject overload traffic, a method of controlling said gap for said second systems comprising:
in the first system, initializing an indication of a gap value;
measuring a fraction of traffic that cannot be processed in the first system;
responsive to said measuring, adjusting the value of the gap upward if the fraction exceeds a first threshold, downward, but not below zero, if the fraction is less than a second threshold and maintaining the value of the gap if the fraction is between the first and second threshold;
transmitting the value of the adjusted gap to the second system;
repeating the measuring, adjusting, and transmitting steps recurrently.
in the first system, initializing an indication of a gap value;
measuring a fraction of traffic that cannot be processed in the first system;
responsive to said measuring, adjusting the value of the gap upward if the fraction exceeds a first threshold, downward, but not below zero, if the fraction is less than a second threshold and maintaining the value of the gap if the fraction is between the first and second threshold;
transmitting the value of the adjusted gap to the second system;
repeating the measuring, adjusting, and transmitting steps recurrently.
2. The method of claim 1 wherein said adjusting step comprises the step of adjusting the value of the gap upward by one amount if the fraction exceeds a first threshold but is less than a the threshold, and adjusting the value of the gap upward by a second amount if the fraction exceeds the third threshold, the third threshold being larger than the first threshold and the second amount being greater than the first amount.
3. The method of claim 1 wherein said step of repeating recurrently comprises repeating the measuring and adjusting steps more frequently for small values of the gap than for larger values of the gap.
4. The method of claim 3 wherein an interval between successive performance of said measuring step is essentially proportional to the value of the gap.
5. The method of claim 4 wherein said interval is at least a predetermined value.
6. The method of claim 1 wherein the second systems have one set of possible values for the gap and the first system has a second, larger set of possible values including intermediate values, wherein the first system transmits at different times two different values of the first set of values of the gap representing a single intermediate value.
7. In a first system for controlling a gap in a plurality of second systems using gap control to reject overload traffic for a particular destination, a method of controlling the gap for rejecting traffic to the particular destination in said second systems comprising:
in the first system, initializing an indication of a gap value when traffic to said particular destination exceeds a predetermined value;
in the first system, measuring a fraction of traffic that cannot be completed to the particular destination;
responsive to said measuring, adjusting the value of the gap upward if the fraction exceeds a first threshold, downward, but not below zero, if the fraction is less than a second threshold and maintaining the value of the gap if the fraction is between the first and second threshold;
transmitting the value of the adjusted gap to the second systems; and repeating the measuring, adjusting, and transmitting steps recurrently.
in the first system, initializing an indication of a gap value when traffic to said particular destination exceeds a predetermined value;
in the first system, measuring a fraction of traffic that cannot be completed to the particular destination;
responsive to said measuring, adjusting the value of the gap upward if the fraction exceeds a first threshold, downward, but not below zero, if the fraction is less than a second threshold and maintaining the value of the gap if the fraction is between the first and second threshold;
transmitting the value of the adjusted gap to the second systems; and repeating the measuring, adjusting, and transmitting steps recurrently.
8. In a database system, means for controlling a gap value for throttling request message traffic from a plurality of sources, comprising:
means for transmitting said gap value to said plurality of sources; and processor means, operating under control of a program for:
initializing an indication of a gap value;
measuring a fraction of traffic that cannot be processed in the database system;
responsive to said measuring, adjusting the value of the gap upward if the fraction exceeds a first threshold, downward, but not below zero, if the fraction is less than a second threshold and maintaining the value of the gap if the fraction is between the first and second threshold;
controlling said means for transmitting to transmit the value of the adjusted gap to the plurality of sources; and repeating the measuring, adjusting, and controlling steps recurrently.
means for transmitting said gap value to said plurality of sources; and processor means, operating under control of a program for:
initializing an indication of a gap value;
measuring a fraction of traffic that cannot be processed in the database system;
responsive to said measuring, adjusting the value of the gap upward if the fraction exceeds a first threshold, downward, but not below zero, if the fraction is less than a second threshold and maintaining the value of the gap if the fraction is between the first and second threshold;
controlling said means for transmitting to transmit the value of the adjusted gap to the plurality of sources; and repeating the measuring, adjusting, and controlling steps recurrently.
9. In the database system of claim 8, the processor means further operating under control of said program for adjusting the value of the gap upward by one amount if the fraction exceeds a first threshold but is less than a third threshold, and adjusting the value of the gap upward by a second amount if the fraction exceeds the third threshold, the third threshold being larger than the first threshold and the second amount being greater than the first amount.
10. In the database system of claim 8, the processor means further operating under control of said program for repeating the measuring and adjusting more frequently for small values of the gap than for larger values of the gap.
11. In the database system of claim 10, the processor means further operating under control of said program for performing said measuring at a repetition interval that is essentially proportional to the value of the gap.
12. In the database system of claim 11, the processor means further operating under control of said program for performing said measuring at a repetition interval that is at least a predetermined value.
13. In the database system of claim 8, the processor means further operating under control of said program for controlling said means for transmitting to transmit at different times two different values of a gap corresponding to a single gap value, whereby a gap value intermediate between said two different values iseffectively transmitted to said plurality of sources.
14. In a database system, means for calculating a gap value for throttling traffic to a particular destination from a plurality of sources, comprising:
means for transmitting messages to said plurality of sources; and processor means operative under the control of a program for:
initializing an indication of a gap value when traffic to said particular destination exceeds a predetermined value;
measuring a fraction of traffic that cannot be completed to the particular destination;
responsive to said measuring, adjusting the value of the gap upward if the fraction exceeds a first threshold, downward, but not below zero, if the fraction is less than a second threshold and maintaining the value of the gap if the fraction is between the first and second threshold;
controlling said means for transmitting to transmit the value of the adjusted gap to the second systems;
repeating the measuring, adjusting, and controlling steps recurrently.
means for transmitting messages to said plurality of sources; and processor means operative under the control of a program for:
initializing an indication of a gap value when traffic to said particular destination exceeds a predetermined value;
measuring a fraction of traffic that cannot be completed to the particular destination;
responsive to said measuring, adjusting the value of the gap upward if the fraction exceeds a first threshold, downward, but not below zero, if the fraction is less than a second threshold and maintaining the value of the gap if the fraction is between the first and second threshold;
controlling said means for transmitting to transmit the value of the adjusted gap to the second systems;
repeating the measuring, adjusting, and controlling steps recurrently.
15. In a first system for controlling a gap in a second system, the second system using gap control to reject overload traffic, a method of controlling said gap for said second system comprising:
measuring a fraction of traffic that cannot be processed in the first system; and responsive to said measuring, controlling the generation of information for adjusting a previously set gap value in said second system upward if the fraction exceeds a first threshold, downward, but not below zero, if the fraction is less than a second threshold, and for maintaining said gap value if the fraction is between the first and second thresholds.
measuring a fraction of traffic that cannot be processed in the first system; and responsive to said measuring, controlling the generation of information for adjusting a previously set gap value in said second system upward if the fraction exceeds a first threshold, downward, but not below zero, if the fraction is less than a second threshold, and for maintaining said gap value if the fraction is between the first and second thresholds.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US428,268 | 1989-10-27 | ||
US07/428,268 US5067074A (en) | 1989-10-27 | 1989-10-27 | Control of overload in communications networks |
Publications (2)
Publication Number | Publication Date |
---|---|
CA2024305A1 CA2024305A1 (en) | 1991-04-28 |
CA2024305C true CA2024305C (en) | 1993-09-14 |
Family
ID=23698194
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CA002024305A Expired - Fee Related CA2024305C (en) | 1989-10-27 | 1990-08-30 | Control of overload in communications networks |
Country Status (5)
Country | Link |
---|---|
US (1) | US5067074A (en) |
EP (1) | EP0425202B1 (en) |
JP (1) | JP3033781B2 (en) |
CA (1) | CA2024305C (en) |
DE (1) | DE69027395T2 (en) |
Families Citing this family (51)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0521892A1 (en) * | 1990-03-29 | 1993-01-13 | Micro Technology, Inc. | Method and apparatus for scheduling access to a csma communication medium |
US5175537A (en) * | 1990-03-29 | 1992-12-29 | Micro Technology, Inc. | Method and apparatus for scheduling access to a CSMA communication medium |
JP2555907B2 (en) * | 1990-05-23 | 1996-11-20 | 日本電気株式会社 | Complex network address routing control system |
US5335268A (en) * | 1992-10-22 | 1994-08-02 | Mci Communications Corporation | Intelligent routing of special service telephone traffic |
JPH0793645B2 (en) * | 1993-01-11 | 1995-10-09 | 日本電気株式会社 | Signal connection controller |
ATE160066T1 (en) * | 1993-02-03 | 1997-11-15 | Novell Inc | MULTIMEDIA DISTRIBUTION SYSTEM |
KR960014693B1 (en) * | 1993-12-11 | 1996-10-19 | 조백제 | Load control method for distributed processor in switching system |
KR960016662B1 (en) * | 1993-12-15 | 1996-12-19 | 양승택 | Mixed switching system |
DE69328380T2 (en) * | 1993-12-31 | 2000-10-12 | Ibm | METHOD AND DEVICE FOR COMMUNICATING SEVERAL TRAFFIC CLASSES |
US5500889A (en) * | 1994-06-09 | 1996-03-19 | At&T Corp. | Method and apparatus for prioritizing a telephone call according to a level of service of an originator |
US5715395A (en) * | 1994-09-12 | 1998-02-03 | International Business Machines Corporation | Method and apparatus for reducing network resource location traffic in a network |
JP3560370B2 (en) * | 1994-09-26 | 2004-09-02 | 富士通株式会社 | Data transfer system and data transfer method |
FI97185C (en) * | 1994-11-11 | 1996-10-25 | Nokia Telecommunications Oy | Overload lock in a node in a data communication network |
FI97186C (en) * | 1994-11-11 | 1996-10-25 | Nokia Telecommunications Oy | Overload lock in a node in a data communication network |
EP0735786A3 (en) * | 1995-03-31 | 1998-05-06 | Siemens Aktiengesellschaft | Method for overload defence in a communication network |
KR960043938A (en) * | 1995-05-27 | 1996-12-23 | 김광호 | Message Overload Control Method for Unit Program of Multiprocessor Control System |
US5828729A (en) * | 1995-11-22 | 1998-10-27 | Stentor Resource Centre Inc. | Automated mass calling detection using CCS7 message parameters |
US5933481A (en) * | 1996-02-29 | 1999-08-03 | Bell Canada | Method of controlling call traffic in a telecommunication system |
GB9604675D0 (en) * | 1996-03-05 | 1996-05-01 | British Telecomm | Telecommunications network |
US5878224A (en) * | 1996-05-24 | 1999-03-02 | Bell Communications Research, Inc. | System for preventing server overload by adaptively modifying gap interval that is used by source to limit number of transactions transmitted by source to server |
WO1997045792A1 (en) * | 1996-05-24 | 1997-12-04 | Bell Communications Research Inc. | Apparatus and method for preventing network server overload |
US5905870A (en) * | 1996-09-11 | 1999-05-18 | Advanced Micro Devices, Inc | Arrangement for initiating and maintaining flow control in shared-medium, full-duplex, and switched networks |
DE69816053T2 (en) * | 1997-03-25 | 2004-06-03 | British Telecommunications Public Ltd. Co. | Telecommunications network with congestion control |
US5943232A (en) * | 1997-10-29 | 1999-08-24 | Lucent Technologies Inc. | Autonomous overload control for distributed real time systems |
US6084856A (en) * | 1997-12-18 | 2000-07-04 | Advanced Micro Devices, Inc. | Method and apparatus for adjusting overflow buffers and flow control watermark levels |
US6160875A (en) * | 1997-12-30 | 2000-12-12 | Lg Information & Communications, Ltd. | Method of managing overload of message in the communication system |
US6954458B1 (en) * | 1998-10-20 | 2005-10-11 | Telefonaktiebolaget Lm Ericsson (Publ) | System and method for message transport and segmentation |
US6356629B1 (en) * | 1999-02-02 | 2002-03-12 | Cisco Technology, Inc. | Switched virtual circuit controller setup congestion management strategy |
US6956850B1 (en) | 1999-03-25 | 2005-10-18 | Cisco Technology, Inc. | Call record management for high capacity switched virtual circuits |
US6625121B1 (en) | 1999-04-28 | 2003-09-23 | Cisco Technology, Inc. | Dynamically delisting and relisting multicast destinations in a network switching node |
US6459902B1 (en) | 1999-09-07 | 2002-10-01 | Qualcomm Incorporated | System and method for selectively blocking or dropping calls in a telecommunications network |
US7200408B1 (en) * | 1999-12-15 | 2007-04-03 | Lucent Technologies Inc. | Selective blocking in a communication network |
US8161182B1 (en) | 2000-01-26 | 2012-04-17 | Cisco Technology, Inc. | Managing network congestion using dynamically advertised congestion status |
US7707305B2 (en) | 2000-10-17 | 2010-04-27 | Cisco Technology, Inc. | Methods and apparatus for protecting against overload conditions on nodes of a distributed network |
GB2375002B (en) * | 2001-04-25 | 2003-07-09 | Lucent Technologies Inc | A method for overload control in a telecommunications network and apparatus therefor |
WO2003047229A1 (en) * | 2001-11-30 | 2003-06-05 | Telefonaktiebolaget Lm Ericsson (Publ) | A method and arrangement for network load regulation |
EP1345354A1 (en) * | 2002-03-13 | 2003-09-17 | Alcatel | Electrical network for data transmission |
US7570584B1 (en) | 2002-03-29 | 2009-08-04 | Cisco Technology, Inc. | Network-wide congestion control of SPVC signaling messages |
US7418492B1 (en) | 2002-06-20 | 2008-08-26 | P-Cube Ltd. | System and a method for testing network communication devices |
US6829338B2 (en) | 2002-10-02 | 2004-12-07 | At&T Corp. | Network having space chattering control for maximizing call throughput during overload |
US7979694B2 (en) * | 2003-03-03 | 2011-07-12 | Cisco Technology, Inc. | Using TCP to authenticate IP source addresses |
US6907115B2 (en) * | 2003-05-30 | 2005-06-14 | Lucent Technologies | Methods and devices for providing early detection, controlled elimination of overload conditions and the return of a controlled level of traffic after such conditions have been substantially eliminated |
US20050050139A1 (en) * | 2003-09-03 | 2005-03-03 | International Business Machines Corporation | Parametric-based control of autonomic architecture |
CN1926889B (en) | 2004-02-25 | 2011-01-26 | 英国电讯有限公司 | Overload control in a communications network |
US7540025B2 (en) * | 2004-11-18 | 2009-05-26 | Cisco Technology, Inc. | Mitigating network attacks using automatic signature generation |
US7568224B1 (en) | 2004-12-06 | 2009-07-28 | Cisco Technology, Inc. | Authentication of SIP and RTP traffic |
US7620733B1 (en) | 2005-03-30 | 2009-11-17 | Cisco Technology, Inc. | DNS anti-spoofing using UDP |
US8315245B2 (en) * | 2005-09-21 | 2012-11-20 | At&T Intellectual Property I, L.P. | Overload call control in a VoIP network |
US8156557B2 (en) * | 2007-01-04 | 2012-04-10 | Cisco Technology, Inc. | Protection against reflection distributed denial of service attacks |
US8341265B2 (en) * | 2009-01-09 | 2012-12-25 | Sonus Networks, Inc. | Hybrid server overload control scheme for maximizing server throughput |
US9058210B2 (en) * | 2010-03-23 | 2015-06-16 | Ebay Inc. | Weighted request rate limiting for resources |
Family Cites Families (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4123794A (en) * | 1974-02-15 | 1978-10-31 | Tokyo Shibaura Electric Co., Limited | Multi-computer system |
DE3311875A1 (en) * | 1983-03-31 | 1984-10-04 | Siemens AG, 1000 Berlin und 8000 München | CIRCUIT ARRANGEMENT FOR TELECOMMUNICATION SYSTEMS, ESPECIALLY TELECOMMUNICATION SYSTEMS, WITH INFORMATION PROCESSING SWITCHGEAR AND DEVICES TO REDUCE OVERLOAD |
DE3311866A1 (en) * | 1983-03-31 | 1984-10-04 | Siemens AG, 1000 Berlin und 8000 München | CIRCUIT ARRANGEMENT FOR TELECOMMUNICATION SYSTEMS, ESPECIALLY TELECOMMUNICATION SYSTEMS WITH INFORMATION-PROCESSING SWITCHGEAR AND DEVICES TO DEFEND OVERLOAD |
US4642758A (en) * | 1984-07-16 | 1987-02-10 | At&T Bell Laboratories | File transfer scheduling arrangement |
JPS61152151A (en) * | 1984-12-26 | 1986-07-10 | Nec Corp | Overload control system |
JPS6413845A (en) * | 1987-07-08 | 1989-01-18 | Nippon Telegraph & Telephone | Voice deputy reply system for incoming congestion subscriber |
US4974256A (en) * | 1989-06-30 | 1990-11-27 | At&T Bell Laboratories | Load balancing and overload control in a distributed processing telecommunications system |
-
1989
- 1989-10-27 US US07/428,268 patent/US5067074A/en not_active Expired - Lifetime
-
1990
- 1990-08-30 CA CA002024305A patent/CA2024305C/en not_active Expired - Fee Related
- 1990-10-19 EP EP90311487A patent/EP0425202B1/en not_active Expired - Lifetime
- 1990-10-19 DE DE69027395T patent/DE69027395T2/en not_active Expired - Fee Related
- 1990-10-26 JP JP2287416A patent/JP3033781B2/en not_active Expired - Lifetime
Also Published As
Publication number | Publication date |
---|---|
EP0425202A3 (en) | 1993-05-05 |
EP0425202A2 (en) | 1991-05-02 |
JP3033781B2 (en) | 2000-04-17 |
CA2024305A1 (en) | 1991-04-28 |
DE69027395D1 (en) | 1996-07-18 |
US5067074A (en) | 1991-11-19 |
DE69027395T2 (en) | 1997-01-23 |
JPH03153149A (en) | 1991-07-01 |
EP0425202B1 (en) | 1996-06-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CA2024305C (en) | Control of overload in communications networks | |
US4845710A (en) | Dynamic buffer supervising system for a data link access protocol control | |
US6363052B1 (en) | Congestion control in network systems | |
US5592672A (en) | System for load balancing between message processors by routing all queued messages to a particular processor selected by a deterministic rule | |
US5778057A (en) | Service control point congestion control method | |
US5541987A (en) | Connection-oriented congestion controller for common channel signaling network | |
US6453351B1 (en) | Traffic control method and network control system | |
EP0610625A2 (en) | Dynamic network automatic call distribution | |
US5896441A (en) | Communication state management system and method in an intelligent network | |
EP0800294B1 (en) | Method to control data flow rate, queuing network node and packet switching network | |
US6259776B1 (en) | System for controlling telecommunication overload traffic | |
CA2033290C (en) | Packet exchange network and method of controlling discard of packets in packet exchange network | |
US5825860A (en) | Load sharing group of service control points connected to a mediation point for traffic management control | |
US6707900B1 (en) | Dynamic load limiting | |
US6473402B1 (en) | Communications link interconnecting service control points of a load sharing group for traffic management control | |
JP4166434B2 (en) | Call distribution | |
US5982751A (en) | Rare probability connection call registration method using payload type indication field information for asynchronous transfer mode switching system | |
JP3317503B2 (en) | Communication links interconnecting service control points of load balancing groups for traffic management control | |
US5946296A (en) | Rare probability connection call registration method for asynchronous transfer mode switching system | |
EP0872124B1 (en) | Load sharing group of service control points connected to a mediation point for traffic management control | |
KR960006469B1 (en) | Overload control method in the intelligent network | |
WO1998043451A1 (en) | Telecommunications networks | |
Hosein | An Improved Automatic Congestion Control Algorithm for Telecommunications Signaling Networks | |
JPH06101747B2 (en) | Communication device buffer management method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
EEER | Examination request | ||
MKLA | Lapsed |