US20130121144A1

US20130121144A1 - Multiplexing apparatus and discard method

Info

Publication number: US20130121144A1
Application number: US13/732,093
Authority: US
Inventors: Hiroshi Ueno
Original assignee: Juniper Networks Inc
Current assignee: Juniper Networks Inc
Priority date: 1998-12-28
Filing date: 2012-12-31
Publication date: 2013-05-16
Also published as: JP2000196621A; HK1028511A1; US7154856B1; US20070086467A1; US20110007634A1; US8369320B2; US7826357B2; CN1258975A; CN1147092C

Abstract

An ATM multiplexing apparatus of the present invention is the apparatus for selectively performing cell discard processing in the case of congestion on the basis of a use state of the same connection formed by cells from the side of an ATM switching unit and subscribers without installing UPC units, and the ATM multiplexing apparatus, which is connected to the ATM switching unit and each of plural subscribers through ATM communication lines and performs multiplexing processing to ATM cells sent from the plural subscribers, comprises: detection means 118 for detecting a level of a congestion state corresponding to the received ATM 157 from the subscribers; and discard means (111, 115 and 117) for selectively discarding the received ATM cells from the subscribers on the basis of a communication state determined by the received ATM cells 155 from the ATM switching unit and the received ATM cells from the subscribers and a level value of a warning signal 129 indicating the detected congestion state.

Description

CROSS REFERENCE TO RELATED APPLICATIONS

The present invention claims priority from Japanese Patent Application No. 10-374032 filed Dec. 28, 1998, the contents of which are incorporated herein by reference.

BACKGROUND OF THE INVENTION

1. Field of the Invention
The present invention relates to an ATM (Asynchronous Transfer Mode) multiplexing apparatus, and particularly, to an ATM multiplexing apparatus for performing cell discard processing on the basis of a communication state of the connection between ATM cells from the side of an ATM switching unit and subscribers.
2. Description of Related Art
An ATM network communication system has become popular as service for integrally conducting data communication including voice and text data, further image data requiring high-speed transmission. In the ATM network communication system, cell discard processing is performed in order to protect an ATM transmission band from an ATM cell caused by abnormality etc. of subscriber terminals. The cell discard processing is carried out mainly by an ATM switching unit and an ATM multiplexing apparatus.
With regard to public known arts related to the present invention, technique on a control method of monitoring the packet flow rate is disclosed in Japanese Unexamined Patent Publication No. 46344 of 1997. This control method provides a plurality of usage parameter control mechanisms at the entrance of an ATM network. In each of the usage parameter control mechanism, the transfer rate defined in ABR (Available Bit Rate) communication service is set as a monitor parameter. Thus, monitor operations according to the use state of the ATM network are executed. Also, technique on an ATM communication network comprising ATM switching units connected in a multistage manner is disclosed in Japanese Unexamined Patent Publication No. 261244 of 1997. In this ATM communication network, if congestion has occurred in any of the ATM switching units, an ATM switch of the fore-stage of the ATM switching unit in which the congestion has occurred is notified of the congestion. The notified ATM switching unit controls the discard state of cells. Further, technique on a packet control switching unit and a cell transfer control method are disclosed in Japanese Unexamined Patent Publication No. 13427 of 1998. In this cell transfer control method, the ATM switching unit acting as a node unit stores priority on cell discard corresponding to a connection identifier. Cell discard processing is controlled on the basis of discard conditions determined by the relationship of the stored cell priority to the number of stack cells in the node every connection.
FIG. 11 shows a conceptual diagram of a conventional ATM network system. An ATM switching unit 201 is installed in a central station. ATM multiplexing apparatus 301 and 303 are largely installed at remote sites nearest to subscribers. The ATM multiplexing apparatus 301 and 303 desirably have a cheap and simple configuration where possible.
Functions necessary for the ATM multiplexing apparatus 301 and 303 include a UPC (Usage Parameter Control) function. An object of the UPC function is to prevent congestion in an ATM network by checking whether the ATM cell flow from the subscribers violates a contract band or not. In order to efficiently share a band of an ATM transmission line with all the subscriber signals, the UPC to input signals of each subscriber line before multiplexing needs to be performed. However, installation of this UPC units 309 and 311 increases the cost per subscriber line. On the other hand, when SVC (Switched Virtual Connection) service for performing dynamic signal setting is adopted, traffic setting information set by signaling needs to be passed between the ATM switching unit and the ATM multiplexing apparatus. Passing of the traffic setting information can be managed by installation of a communication line for management, but the amount of processing of the ATM multiplexing apparatus increases.
In the conventional ATM network system as shown in FIG. 11, the ATM multiplexing apparatus 301 provides the UPC units 309 every subscriber line inside the units, so that the cost of installation increases. The UPC unit 311 is installed inside of the ATM multiplexing apparatus 303 to monitor a signal after multiplexing in a lump, so that the cost of installation of the UPC unit 311 decreases. Even in the case of installing any of the ATM multiplexing apparatus 301 and 303, traffic information from a traffic processing control part 211 is transferred from the ATM switching unit 201 to the ATM multiplexing apparatus 301 and 303 through traffic information transfer means 213. The traffic information transfer means 213 means communication between respective unit control parts. Coping with the traffic information transfer means 213 requires the increase in the cost of the ATM multiplexing apparatus 301 and 303.
Under such circumstances, the ATM multiplexing apparatus and the selective cell discard method for performing cell discard processing efficiently in the case of congestion on the basis of a use state of the connection formed by ATM cells received from the side of the ATM switching unit and ATM cells received from the side of the subscribers without installing the UPC units have been desired. It is desirable to detect a congestion state corresponding to the ATM cell flow sent from each subscriber. It is desirable to sequentially update the use state of the connection on the basis of available information on the ATM cells received from the ATM switching unit or the subscribers. The discard processing is desirably performed on the basis of a logical decision using a plurality of data providing the use state (communication state) of the connection and level information of the congestion state.

SUMMARY OF THE INVENTION

An object of the present invention is to provide an ATM multiplexing apparatus and the cell discard method for selectively performing cell discard processing on the basis of a level value of a congestion state corresponding to each subscriber and a communication state of the connection owned by each subscriber, and to provide an ATM multiplexing apparatus and the cell discard method for performing cell discard processing on the basis of a receiving interval of ATM cells from an ATM switching unit or subscribers, and to provide criteria of discard decision logic providing a basis of discard processing.
In order to achieve such an object, an ATM (Asynchronous Transfer Mode) multiplexing apparatus of the present invention is characterized by the ATM multiplexing apparatus for being connected to an ATM switching unit and each of plural subscribers through ATM communication lines and performing multiplexing processing to ATM cells sent from the plural subscribers, the ATM multiplexing apparatus comprising: detection means for detecting a congestion state corresponding to received ATM cells from the subscribers and outputting a warning signal of a level value according to the congestion state; and discard means for selectively discarding the received ATM cells from the subscribers on the basis of a use state (communication state) of the same connection determined by received ATM cells 155 from the ATM switching unit and the received ATM cells from the subscribers and the level value of the congestion state indicated by the warning signal from the detection means.
The above communication state is determined by header information and its received notification included in the received ATM cells from the ATM switching unit or header information and its received notification included in the received ATM cells from the subscribers.
The detection means desirably comprises storage means for storing the received ATM cells from the subscribers, and comparison means for generating the warning signal on the basis of the degree of occupancy in the storage means of the stored ATM cells and the preset threshold.
Further, it is desirable that a plurality of detection means be provided and priority related to storage of the received ATM cells from the subscribers be preset to each of the detection means. Also, each of the detection means desirably comprises storage means for storing the received ATM cells from the subscribers and presetting the threshold corresponding to the priority, and comparison means for generating the warning signal corresponding to the priority on the basis of the degree of occupancy in the storage means of the stored ATM cells and the preset threshold.
Moreover, the discard means desirably comprises switching unit monitor means for receiving (monitoring) the ATM cells from the ATM switching unit and outputting header information and received notification of the received ATM cells from the ATM switching unit as first header information and first received notification, subscriber monitor/selection means for receiving (monitoring) the ATM cells from the subscribers and outputting header information and received notification of the received ATM cells from the subscribers as second header information and second received notification and selectively discarding the received ATM cells from the subscribers on the basis of a discard command, and discard control means for updating status data indicating the communication state on the basis of the first header information and the first received notification or the second header information and the second received notification and generating the discard command for commanding discard of the received ATM cells from the subscribers on the basis of the updated status data and the level value of the warning signal.
The discard control means desirably comprises a status table for storing the status data, which is formed of status flags and received time of the ATM cells from the ATM switching unit or the subscribers, on the basis of an address corresponding to the first header information or the second header information, a table control part for responding to the first header information and the first received notification or the second header information and the second received notification and generating a control signal for commanding update of the status data, an elapsed time decision part for outputting a time-out signal on the basis of a comparison between the elapsed time from the received time to the present time of the status data read out responding to the control signal and the preset reference elapsed time, and a discard decision part for updating the readout status data on the basis of a signal related to either the first or second received notification and the time-out signal and the second header information and generating the discard command on the basis of the updated status data and the level value of the warning signal.
The discard command is generated by the updated status data and the level values of a plurality of the warning signals, and may include information specifying any of a plurality of the provided detection means as the target storage of the received ATM cells from the subscribers. The updated status data (updated status flags) is generated on the basis of criteria of the preset state transition. The discard command is generated on the basis of criteria of the preset discard decision logic. The status flags are formed by connection data “ACTIVE” or “SUSPEND” providing a two-way connection state established by the received ATM cells from the ATM switching unit and the received ATM cells from the subscribers, position data “BOM” or “COM” providing a position of the received ATM cells from the subscribers in a packet signal corresponding to an ATM adaptation layer (AAL5), discard data “DISCARDING” providing whether cell discard processing corresponding to the packet signal is performed or not when the ATM cells from the subscribers constitute the packet signal, and an AAL data “AAL5” indicating whether the received ATM cells from the ATM switching unit and the subscribers constitute the packet signal or not.
A discard method of ATM cells according to the ATM multiplexing apparatus of the present invention comprises the steps of receiving ATM cells sent from subscribers and detecting a congestion state of the received ATM cells from the subscribers, updating data providing a communication state of the connection determined by the received ATM cells from the subscribers and received ATM cells from an ATM switching unit responding to receipt of the ATM cells from the subscribers or the ATM switching unit, deciding whether discard processing of the received ATM cells from the subscribers is performed or not on the basis of the updated data and a level value of a signal indicating the congestion state, and selectively performing the discard processing on the basis of the decision result.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of an asynchronous transfer mode (hereinafter called ATM) multiplexing apparatus and an ATM cell discard method of the present invention will be described in detail with reference to the accompanying drawings below.

FIG. 1 is a block diagram illustrating an ATM network communication system according to the first embodiment of the present invention

FIG. 2 is a block diagram illustrating an ATM multiplexing apparatus according to the first embodiment of the present invention

FIG. 3 is a block diagram illustrating a discard control part in the ATM multiplexing apparatus according to the first embodiment of the present invention

FIG. 4 is an illustration of the format of a status table in the ATM multiplexing apparatus according to the first and second embodiments of the present invention

FIGS. 5A and 5B are conceptual diagrams illustrating a part of processing of the discard control part in the ATM multiplexing apparatus according to the first and second embodiments of the present invention;

FIGS. 6A and 6B are conceptual diagrams illustrating a part of processing of the discard control part in the ATM multiplexing apparatus according to the first embodiment of the present invention;

FIG. 7 is a flowchart illustrating operations of cell discard processing of the ATM multiplexing apparatus according to the first embodiment of the present invention;

FIG. 8 is a block diagram illustrating an ATM multiplexing apparatus according to the second embodiment of the present invention

FIG. 9 is a block diagram illustrating a discard control part in the ATM multiplexing apparatus according to the second embodiment of the present invention

FIG. 10 is a conceptual diagram illustrating a part of processing of the discard control part in the ATM multiplexing apparatus according to the second embodiment of the present invention; and

FIG. 11 is a block diagram illustrating a conventional ATM network communication system and ATM multiplexing apparatus.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

Referring to FIG. 1 which shows a block diagram of an ATM network communication system receiving ATM multiplexing apparatus of the present invention, this ATM network communication system comprises an ATM switching unit 101 and ATM multiplexing apparatus 109. The ATM switching unit 101 is arranged in a station. The ATM switching unit 101 is connected to ATM transmission lines (ATM communication lines) A.
The ATM switching unit 101 comprises a switch part 103, UPC (Usage Parameter Control) units 105, and a traffic processing control part 107. The switch part 103 switches the direction of signals (ATM cells) from subscribers. The switched signals are connected to an ATM network through the ATM transmission lines A. The UPC units 105 monitor whether multiplexing signals 112 from the ATM multiplexing apparatus 109 violate the contracted traffic or not. The traffic processing control part 107 responds to the signaling connection request (not shown) from the subscribers and performs connection processing of the requested connection.
The ATM multiplexing apparatus 109 are arranged in the side of the subscribers corresponding to the number of UPC units 105 provided in the ATM switching unit 101, and are connected to the ATM switching unit 101 through ATM transmission lines B and are connected to the signals (ATM cells) from the subscribers through ATM transmission lines C.
The ATM multiplexing apparatus 109 comprises a cell header monitor part (switching unit monitor means) 111, a multiplexing part 113, discard control parts (discard control means) 115, and cell filters (subscriber monitor/selection means) 117.
The cell header monitor part 111 monitors ATM cells transmitted from the side of the ATM switching unit to the side of the subscribers and acquires header information. The monitored ATM cells from the ATM switching unit are transmitted to the ATM transmission lines C as ATM cells (hereinafter called received ATM cells) 155 received from the ATM switching unit. Also, the cell header monitor part 111 outputs header information (first header information) and its received notification (first received notification) 125 acquired from the received ATM cells 155 from the ATM switching unit. The first header information includes a VPI (Virtual Pass Identifier) and a VCI (Virtual Channel Identifier).
The multiplexing part 113 performs multiplexing processing to ATM cells from the respective cell filters 117 and generates a multiplexing signal 112. The multiplexing part 113 detects a congestion state if the amount of retention of the received ATM cells from the subscribers exceeds the preset threshold. The detected congestion state is generated as a warning signal 129. The discard control part 115 pairs off with the cell filter 117 and they are provided corresponding to the number of subscribers.
The discard control part 115 updates status data 139 (as described in FIGS. 3 and 4 below) indicating a communication state on the basis of the first header information and the first received notification 125 from the cell header monitor part 111 or the second header information and the second received notification 127. Further, the discard control part 115 responds to the warning signal 129 from the multiplexing part 113, and decides whether the received ATM cells from the subscribers are discarded or not on the basis of a level value of the warning signal 129 and the updated status data, and generates a discard command 131 as the decision result.
The cell filter 117 monitors ATM cells 126 from the subscribers and acquires header information. The monitored ATM cells from the subscribers are temporarily stored in the cell filter 117 as ATM cells (hereinafter called received ATM cells) received from the subscribers (not shown). Also, the cell filter 117 outputs header information (second header information) and its received notification (second received notification) 127 acquired from the received ATM cells from the subscribers. The second header information includes a PTI (Payload Type Identifier) in addition to the VPI and the VCI. Moreover, the cell filter 117 discards the received ATM cells from the subscribers stored temporarily on the basis of the discard command 131 from the discard control part 115, or transfers (passes) the received ATM cells to the multiplexing part 113. FIG. 2 shows the received ATM cells 157 from the subscribers in the case of being transferred.
Thus, the ATM multiplexing apparatus 109 of the present invention selectively discards the received ATM cells from the subscribers. This selective discard processing is performed on the basis of the level value of the warning signal 129 from the multiplexing part 113 and the communication state. The communication state in this case indicates a use state of the same connection determined by the received ATM cells 155 from the ATM switching unit and the received ATM cells from the subscribers.
FIG. 2 shows a detailed block diagram of the ATM multiplexing apparatus 109 in the ATM network communication system of the present invention. The multiplexing part 113 comprises detection means 118 and a readout control part 123.
In the present embodiment, one detection means 118 is provided corresponding to a subscriber line (or a pair of the discard control part 115 and the cell filter 117). The detection means 118 comprises storage means 119 and comparison means 121.
The storage means 119 stores the received ATM cells from the subscribers. The comparison means 121 holds the preset threshold. The comparison means 121 generates the warning signal 129 on the basis of the degree of occupancy in the storage means 119 of the ATM cells stored in the storage means 119 and the preset threshold. The warning signal 129 provides the level value corresponding to the preset threshold. Also, the storage means 119 generates the warning signal 129 indicating a congestion state if the above degree of occupancy exceeds the preset threshold. As described below, if the threshold, for example, is set to “½”, the warning signal 129 is generated as “0” or “1”. Also, if a plurality of the thresholds, for example, are set to “⅝”, “¾” and “⅞”, the warning signal 129 provides the level values of “000”, “001”, “011” or “111”. The readout control part 123 sequentially reads out the ATM cells stored in the respective storage means 119, and transmits the ATM cells to the ATM transmission lines B as the multiplexing signal 112. In the present invention, a readout control method is not limited.
FIG. 3 shows a detailed block diagram of the discard control part 115 as indicated in FIG. 2. Referring to FIG. 3, the discard control part 115 comprises a table control part 132, a status table 137, a first register 141, a second register 143, a discard decision part 147, an elapsed time decision part 151, and a timer counter 153.
The table control part 132 responds to the first header information (VPI, VCI) 125 a and the first received notification 125 b or the second header information (VPI, VCI) 127 a and the second received notification 127 b, and generates a control signal 133. The control signal 133 is a signal for commanding updating of status data 139, and includes a directive signal 133 a of the address corresponding to the first header information 125 a or the second header information 127 a and a sending end signal 133 b. The sending end signal 133 b provides a sending source indicating that the received ATM cells have been received from the side of either the ATM switching unit or the subscribers. Thus, the sending end signal 133 b is a signal associated with either the first received notification 125 b or the second received notification 127 b.
The status table 137 stores the status data 139 on the basis of the address corresponding to the first header information 125 a or the second header information 127 a. The format (the status data 139) of the status table 137 is shown in FIG. 4. The status data 139 provides the above-mentioned communication state. The status data 139 is formed of status flags 139 a and received time (time stamp) 139 b of the ATM cells from the ATM switching unit or the subscribers. Details of the status flags 139 a are described below. Also, the status table 137 outputs the status data 139 corresponding to the address indicated by the directive signal 133 a of the address as status data 135 using time division processing.
The first register 141 temporarily stores the status data 135 read out from the status table 137 and the sending end signal 133 b. The second register 143 temporarily stores the status data updated by the discard decision part 147 and that time. The elapsed time decision part 151 calculates the elapsed time from the received time 139 b to the present time 154. The received time 139 b is included in the status data 135 read out responding to the control signal 133 (directive signal 133 a of the address). The present time 154 is generated by the timer counter 153. A time interval of the timer counter 153 is the order of several hundreds of milliseconds to a second and may be set in apparatus unit. Further, the elapsed time decision part 151 compares the calculated elapsed time with a preset reference elapsed time 149. The elapsed time decision part 151 generates a time-out signal 152 if it is decided that the calculated elapsed time has exceeded the preset reference elapsed time 149. The preset reference elapsed time 149 is held in the elapsed time decision part 151 and is variably set from the outside.
The discard decision part 147 updates the readout status data 135 on the basis of the sending end signal 133 b, the time-out signal 152 and the second header information (PTI) 127 c using the sending end signal 133 b (signal related to any of the first or second received notification) from the first register 141 and the readout status data 135, the time-out signal 152 from the elapsed time decision part 151 and further the second header information (PTI) 127 c as input. In this update processing, each component of the status flags 139 a in the status data 135 read out is updated on the basis of criteria of the state transition shown in FIG. 5. Moreover, the received time 139 b in the status data 135 read out is updated by the present time from the timer counter 153. Each of the updated data is again stored in the status table 137 as the updated status data 145 through the second register 143. The criteria of the state transition will be described below. Further, the discard decision part 147 generates the discard command 131 on the basis of criteria of discard decision logic shown in FIG. 6 when the warning signal 129 indicating the congestion state is inputted in the case of the above update processing. The criteria of discard decision logic will be described below.
Incidentally, for convenience of description, the first and second registers (141 and 143), the discard decision part 147 and the elapsed time decision part 151 are individually shown in the present embodiment, but they are preferably integrated on one LSI (Large Scale Integration).
FIG. 4 shows the format of the status table 137 (status data 139). The status data 139 is formed of status flags 139 a and received time (time stamp) 139 b. Also, in this example, the status flags 139 a are formed of connection data (ACTIVE/SUSPEND) 139 a 1, position data (BOM/COM) 139 a 2, discard data (DISCARDING) 139 a 3, and AAL data (AAL5) 139 a 4. The connection data (ACTIVE/SUSPEND) 139 a 1 indicates a two-way connection state of the same connection established by the received ATM cells 155 from the ATM switching unit and the received ATM cells from the subscribers. The position data (BOM/COM) 139 a 2 indicates the where the received ATM cells from the subscribers are positioned in a packet signal corresponding to an ATM adaptation layer (AAL)5. “BOM” means Beginning Of Message of the packet signal, and “COM” means Continuation Of Message of the packet signal. The discard data (DISCARDING) 139 a 3 indicates whether cell discard processing constituting the packet signal corresponding to the ATM adaptation layer (AAL)5 is performed or not. The AAL data (AAL5) 139 a 4 indicates whether the received ATM cells 155 from the ATM switching unit and the received ATM cells from the subscribers constitute the packet signal corresponding to the above AAL5 or not.
FIGS. 5A and 5R show diagrams of criteria of the state transition. The update processing in the discard decision part 147 as described above is performed on the basis of the criteria of the state transition. FIG. 5A provides the state transition of the connection data (ACTIVE/SUSPEND) 139 a 1. FIG. 5B provides the state transition of the AAL data (PALS) 139 a 4.
Referring first to FIG. 5A, the discard decision part 147 sets a state (connection data 139 a 1) of the connection specified by cell information (VPI, VCI) of the received ATM cells from the subscribers to a “SUSPEND” state when a receiving end signal 133 b related to the second received notification 127 b is inputted (transition 0, including initial state). In the state of transition 0, the discard decision part 147 holds the state of the connection data 139 a 1 read out through the first register 141 with the “SUSPEND” state when the receiving end signal 133 b related to the second received notification 127 b is inputted (transition 1). In the state of transition 0, the discard decision part 147 updates the connection data 139 a 1 to an “ACTIVE” state when a receiving end signal 133 b related to the first received notification 125 b is inputted (transition 2). In the state of transition 2, the discard decision part 147 holds the connection data 139 a 1 with the “ACTIVE” state when the receiving end signal 133 b related to the first received notification 125 b is inputted (transition 3). In the state of transition 2, the discard decision part 147 performs processing based on a time-out signal 152 as shown below when the receiving end signal 133 b related to the second received notification 127 b is inputted. The discard decision part 147 holds the connection data 139 a 1 with the “ACTIVE” state when inputting only the receiving end signal 133 b related to the second received notification 127 b in the “ACTIVE” state (transition 4). Also, the discard decision part 147 updates the connection data 139 a 1 to the “SUSPEND” state when inputting the time-out signal 152 as well as the receiving end signal 133 b related to the second received notification 127 b in the “ACTIVE” state (transition 5).
Next, FIG. 5B will be described. When the packet signal corresponding to the AAL5 is divided into ATM cells, the least significant bit of a PTI field (3 bits) included in a cell header of the last cell is set to “1”. Thus, the discard decision part 147 responds to cell information (PTI) 127 c and updates AAL data 139 a 4 when the received ATM cells from the subscribers constitute the packet signal conforming to the AAL5. That is, the discard decision part 147 responds to the cell information (PTI) 127 c providing“PTI=xx0” (where x is an arbitrary bit value) and sets the AAL data 139 a 4 to a “COM” state (transition 00, including initial state). In the state of transition 00, the discard decision part 147 responds to the cell information (PTI) 127 c providing “PTI=xx1” and updates the AAL data 139 a 4 to a “BOM” state (transition 01). In the state of transition 01, the discard decision part 147 responds to the cell information (PTI) 127 c providing “PTI=xx0” and updates the AAL data 139 a 4 to a “COM” state (transition 02). In the state of transition 02, the discard decision part 147 holds the AAL data 139 a 4 to a “COM” state when inputting the cell information (PTI) 127 c providing“PTI=xx0” (transition 02)
The following two states may be detected by the state transition diagram of FIG. 5A.
The first state is the state in which cell input only from the subscribers is present and response cells from the ATM switching unit (ATM network side) are absent. This first state is the state in which abnormal conditions have occurred in subscriber terminals or flow control such as TCP/IP has not been performed in a high layer, and is a possibility of the unreliable connection in communication lines of only one way. In this case, input cells may be the abnormal cells or the unreliable connection. Thus, there is no serious problem even if the input cells are discarded in congestion. The second state is the state in which arrival intervals of the cells from the subscribers have elapsed longer than the preset time. That is, it is possible that the SVC connection is disconnected once and is switched to other connection.
The second state cannot continue the previous condition (“ACTIVE” state). Thus, the connection data 139 a 1 is updated to the “SUSPEND” state in order to again decide the state. Also, when the connection is not disconnected, it may be decided that there is the connection having low traffic density. For low traffic density, there is a little load on the ATM network even if the high layer executes resending after cell discard. Therefore, there is no serious problem even if discard processing is performed in congestion.
In the case where the received ATM cells from the subscribers constitute the packet signal corresponding to the AAL5, if one cell is discarded, all the packet signal is discarded in the high layer. When cell discard is carried out in a congestion state, packet discard technique for discarding the packet signal corresponding to the AAL5 except for the last cell is used. Hence, in the present embodiment, the communication state is managed using the technique in conjunction with the criteria of the state transition as shown in FIG. 5B. When the cell discard in the packet signal is carried out, the cells corresponding to the same packet signal and the same connection are discarded except for the last cell.
As described above, the ATM multiplexing apparatus 109 of the present invention manages the communication state on the basis of the criteria of the state transition as shown in FIGS. 5A and 58. This sequentially updated communication state (status data 139) provides decision materials in the case of discard decision processing as described below.
FIGS. 6A and 6B show criteria of discard decision logic. Here, the criteria of discard decision in the case that the level value of congestion given by a warning signal 129 is one kind and three kinds (“⅝”, “¾” and “⅞”) is indicated. On the basis of this criteria, the discard decision part 147 performs the discard decision processing using the level value indicated by the warning signal 129 and the updated status flags as decision materials when the warning signal 129 indicating the congestion state is inputted.
Referring to FIG. 6A, decision 1 provides the decision in the case that discard data 139 a 3 after the state transition is “DISCARDING” and the received ATM cells from the subscribers is not the last cell (“PTI=xx0”). In this case, the discard decision part 147 generates the discard command 131 (for example, discard level“1”) intended for the cells except the last cell in spite of the level value of the warning signal 129. When discard processing intended for the BOM cell or COM cell of the packet signal corresponding to the AAL5 is carried out, the discard data 139 a 3 is updated to a “DISCARDING” state. Also, when receiving the last cell (inputting the cell information (PTI) 127 c providing “PTI=xx1”), the “DISCARDING” state is released. Decision 2 provides the decision in the case that the discard decision command 129 is “1” (congestion state) and the connection data 139 a 1 is a “SUSPEND” state. In this case, the discard decision part 147 generates the discard command 131 in spite of the position data 139 a 2.
Referring to FIG. 6B, decision 1′ provides the decision in the case that discard data 139 a 3 is the “DISCARDING” state and the received ATM cells from the subscribers is not the last cell (“PTI=xx0”). In this case, the discard decision part 147 generates the discard command 131 in spite of the level value of the warning signal 129. Decision 2 provides the decision in the case of inputting the warning signal 129 giving the level value “001”. In this case, the discard decision part 147 generates the discard command 131 when the connection data 139 a 1 is the “SUSPEND” state and the position data 139 a 3 is “BOM”. At this time, the discard data 139 a 3 is updated to the “DISCARDING” state. Decision 3′ provides the decision in the case of inputting the warning signal 129 giving the level value “011”. In this case, the discard decision part 147 generates the discard command 131 when the connection data 139 a 1 is the “SUSPEND” state and the position data 139 a 3 is “COM”. Decision 4′ provides the decision in the case of inputting the warning signal 129 giving the level value “111”. In this case, the discard decision part 147 generates the discard command 131 when the position data 139 a 2 is the “BOM” state even if the connection data 139 a 1 is the “ACTIVE” state. At this time, the discard data 139 a 3 is updated to the “DISCARDING” state.
The ATM multiplexing apparatus of the present invention takes communication of the packet signal corresponding to the AAL5 into consideration. The AAL5 is in the most use in data communication on an ATM network. As to a procedure, the initial value of the position data 139 a 2 of the received ATM cells constituting the packet signal corresponding to the AAL5 is set to the “COM” state. Next, responding to the second header information (“PTI=xx1”) 127 e, the position data 139 a 2 is updated to the “BOM” state. At this time, the AAL5 data 139 a 4 is set to “ON”. Further, the discard data 139 a 3 is updated to the “DISCARDING” state only if the AAL5 data 139 a 4 is the “ON” state. Thus, the legality of packet discard processing is assured. Also, in the above discard processing, when the CBR traffic by an AAL1 flows, the position data 139 a 2 is held to the“COM” state. In this case, the cell discard processing is not performed since the AAL5 data is not updated to the “ON” state.
On the other hand, when detection means 109 in the ATM multiplexing apparatus 109 does not detect the congestion state, namely, output capacity in the case of applying multiplexing processing to an ATM signal from the subscribers is not more than output lines connected to the ATM switching unit, the UPC units 105 arranged in the input part of the ATM switching unit 101 monitor the traffic of contract violation in the output capacity after the multiplexing processing, and the input of abnormal cells due to abnormality of the subscriber terminals.
FIG. 7 shows a flowchart illustrating a cell discard method of the ATM multiplexing apparatus according to the first embodiment of the present invention.
Referring to FIGS. 2 and 7, first, in step S101, update processing of status data 139 is executed. This update processing includes the received time 139. Also, the update processing responds to receipt of ATM cells 126 from the subscribers or ATM cells 124 from the ATM switching unit, and is sequentially performed on the basis of the criteria of the state transition shown in FIG. 5 as described above. Next, in step S102, detection processing of a congestion state of the received ATM cells from the subscribers is carried out. In this detection processing, the case that the congestion state is not detected is, for example, the case that the warning signal 129 providing the level value “0” in FIG. 6A as described above is inputted to the discard decision part 147. In this case, processing of step S101 is performed. Then, when the congestion state is detected in step S102, operation is shifted to update processing (step S105) of the status data 139 (status flags 139 a) through receipt (monitor) processing (step S104) of the ATM cells 126 from the subscribers. Further, in step S104, the same update processing as step S101 is performed when receiving (monitoring) the ATM cells 124 from the ATM switching unit. After the update processing of step S105, in step S106, the discard decision part 147 decides whether a pair of data of the level value of the warning signal 129 and the updated status flags 139 a correspond to any of the criteria of discard decision shown in FIGS. 6A and 6B as described above. In step S106, the discard decision part 147 generates the discard command 131 (for example, level value “1”) when it is decided that a pair of data of the level value of the warning signal 129 and the updated status flags 139 a correspond to any of the criteria of the preset discard decision logic. After generation of the discard command 131 (for example, level value “1”), the discard decision part 147 updates or holds the discard data 139 a 3 of the status flags 139 a to “DISCARDING”. Moreover, the discard decision part 147 updates the received time 139 b to the present time 154 and stores them in the status table 137 as the status data 145. Then, in step S107, the cell filter 117 responds to the discard command 131 (for example, “1” or “0”) and selectively discards the received ATM cells from the subscribers stored temporarily.
The ATM multiplexing apparatus 109 in the first embodiment of the present invention selectively discards the ATM cells flowing into the ATM network due to trouble etc. of the subscriber terminals in the case of congestion of the subscriber side. Since each header of the received ATM cells 155 from the ATM switching unit 101 (ATM network) is set by the traffic processing control part 107 of the ATM switching unit 101, header information of the received ATM cells 155 from the ATM switching unit is correct connection information. Among the received ATM cells from the subscribers, the received ATM cells having the header information different from the received ATM cells 155 from the ATM switching unit may highly be abnormal cells caused one-sidedly from the subscriber terminals, so that such cells are priority discarded. By the criteria of the preset discard decision logic, cell discard processing in the case of a short arrival interval of cells decreases in comparison with the case of a long receipt interval of cells, so that packet discard in the case of transmitting a large amount of data through TCP/IP etc. can be reduced. Also, in the cell discard processing in the present invention, connection setting information is unnecessary and it is unnecessary to dynamically receive to process actual connection information held by the ATM switching unit 101, so that control circuits in the ATM multiplexing apparatus can be configured at a low cost, thereby the ATM multiplexing apparatus itself can be configured at a low cost.
Next, an ATM multiplexing apparatus and a cell discard method according to the second embodiment of the present invention will be described. FIG. 8 shows a block diagram of the ATM multiplexing apparatus according to the second embodiment of the present invention. Incidentally, by marking like reference characters on components having the same function as the above-mentioned first embodiment, description of the components is omitted.
Referring to FIG. 8, an ATM multiplexing apparatus 109 according to the present embodiment provides a plurality of detection means corresponding to subscriber lines. Priority related to storage of the received ATM cells from the subscribers is preset to each of a plurality of the detection means. In the present embodiment, two classes of the detection means (118-1, 118-2) are provided to one subscriber. Each of a plurality of the detection means (118-1 or 118-2) provides storage means (119-1 or 119-2) and comparison means (121-1 or 121-2). The storage means (119-1 or 119-2) stores the received ATM cells from the subscribers. Threshold corresponding to the priority is preset in the storage means. The comparison means (121-1 or 121-2) generates a warning signal (129-1 or 129-2) corresponding to the priority on the basis of the degree of occupancy in the storage means of the stored ATM cells and the preset threshold. The warning signal (129-1 or 129-2) provides the level value corresponding to the preset threshold. In the case of providing the storage means in which the priority is respectively set, a use state of the same connection managed by the status flags 139 a can be utilized for priority decision. A readout control part 132 executes cell readout processing from the storage means 119-1 always taking higher priority than cell readout processing from the storage means 119-2.
FIG. 9 shows a block diagram of a discard control part 115′. FIG. 9 corresponds to FIG. 3 in the above-mentioned first embodiment. The difference between FIG. 9 and FIG. 3 is that a discard decision part 147′ inputs the warning signals (129-1, 129-2 shown in FIG. 9) generated from a plurality of the detection means (storage means). The discard decision part 147′ performs discard decision processing on the basis of the level value indicated by the warning signals and the updated status data 139 (status flags) when inputting the warning signals (129-1, 129-2) indicating the congestion state. Update processing of the status data 139 is the same processing as the above-mentioned first embodiment.
The discard decision processing is performed on the basis of criteria of discard decision logic shown in FIG. 10. The criteria shown in FIG. 10 is construed as the decision criteria in the case of inputting the warning signal 129 (level value) of FIG. 6B in the above-mentioned first embodiment to the discard decision part 147 simultaneously.
Referring to FIG. 10, decision 1″ provides the decision in the case that discard data 139 a 3 after the state transition is “DISCARDING” and the received ATM cells from the subscribers is not the last cell (“PTI=xx0”). In this case, the discard decision part 147′ generates the discard command 131′ (discard level “1”) intended for the cells except the last cell in spite of the level values of the warning signals 129-1 and 129-2. Next, decision 2″ provides the decision in the case that each of the level values of the warning signals 129-1 and 129-2 is “1” (congestion state) and the connection data 139 a 1 is a “SUSPEND” state. In this case, the discard decision part 147′ generates the discard command 131′ (discard level “1”) in spite of the position data 139 a 2. Further, decision 3″ provides the decision in the case that each of the level values of the warning signals 129-1 and 129-2 gives “0” and “1”. In this case, the discard decision part 147′ generates the discard command 131′ (pass level “0”) when the connection data 139 a 1 is the “SUSPEND” state in spite of the position data 139 a 3. This discard command 131′ includes indication of the target storage (storage means 119-1). Moreover, decision 4″ provides the decision in the case that each of the level values of the warning signals 129-1 and 129-2 gives“1” and “0”. In this case, the discard decision part 147′ generates the discard command 131′ (pass level “0”) when the connection data 139 a 1 is the “SUSPEND” state in spite of the position data 139 a 3. This discard command 131′ includes indication of the target storage (storage means 119-2). Further, decision 5″ provides the decision in the case that each of the level values of the warning signals 129-1 and 129-2 gives “0”. In this case, the discard decision part 147′ generates the discard command 131′ (pass level “0”) when the connection data 139 a 1 is the “ACTIVE” state in spite of the position data 139 a 3. This discard command 131 includes indication of the target storage (storage means 119-1).
A cell filter 117′ responds to the discard command 131′ from the discard decision part 147′ and selectively performs the cell discard processing. Also, the cell filter 117 transfers the received ATM cells 157 from the subscribers to the storage means 119-1 or 119-2 indicated by the discard command 131′.
By the ATM multiplexing apparatus according to the present embodiment, the cell discard processing with the amount of discard reduced is implemented. Also, by setting the level values of the congestion state different each other, discard processing of the packet signal corresponding to the AAL5 may also be used in the same manner as the above-mentioned first embodiment.
Incidentally, the present invention is not limited to the criteria of state transition and the criteria of discard decision logic (content of decision items) defined in the above-mentioned first and second embodiments, and the criteria are preset variably.
As described above, the ATM multiplexing apparatus according to the present invention implements selective discard processing to the received ATM cells from the subscribers in the case of congestion. By managing a use state of the connection on the basis of information easily acquired from the received ATM cells from the subscribers and the received ATM cells from the ATM switching unit, the selective discard processing is performed.

Claims

1-9. (canceled)

10. A device comprising:

one or more processors to:

obtain information from data received by the device, a first portion of the data being received from a network device and a second portion of the data being received from a subscriber device,

determine, based on the obtained information, a communication state associated with transmitting the data between the network device and the subscriber device,

determine, based on receiving the data, a level of congestion associated with the data, and

selectively discard data from the second portion of the data based on the communication state and the level of congestion.

11. The device of claim 10, where, when obtaining the information, the one or more processors are to:

obtain the information from header portions of the received data.

12. The device of claim 10, where the one or more processors are further to:

store the second portion of the data in a memory, and

where, when determining the level of congestion, the one or more processors are to:

determine an amount of memory used to store the second portion of the data, and

determine the level of congestion based on the amount of memory used to store the second portion of the data.

13. The device of claim 12, where, when determining the level of congestion, the one or more processors are further to:

determine whether the amount of memory used to store the second portion of the data exceeds one or more of a plurality of threshold values, and

determine the level of congestion based on whether the amount of memory used to store the second portion of the data exceeds the one or more of the plurality of thresholds.

14. The device of claim 10, where, when determining the communication state, the one or more processors are to:

determine a state of a connection associated with the first portion of the data,

determine a state of a connection associated with the second portion of the data, and

determine the communication state based on the state of the connection associated with the first portion of the data and the state of the connection associated with the second portion of the data.

15. The device of claim 10, where, when determining the communication state, the one or more processors are to:

determine first information associated with a position of the second portion of the data in a packet signal,

determine second information regarding whether a discard processing of the packet signal has been performed,

determine third information that indicates whether the received data includes the packet signal, and

determine the communication state based on the first information, the second information, and the third information.

16. The device of claim 10, where the one or more processors are further to:

selectively transmit the second portion of the data to the network device as a multiplexed signal.

17. A method comprising:

obtaining, by a device, information from data received by the device;

a first portion of the data being received from a network device, and

a second portion of the data being received from a subscriber device;

determining, by the device and based on the obtained information, a communication state associated with transmitting the data between the network device and the subscriber device;

determining, by the device and based on receiving the data, a level of congestion associated with the data; and

selectively discarding, by the device, data from the second portion of the data based on the communication state and the level of congestion.

18. The method of claim 17, where obtaining the information includes:

obtaining the information from header portions of the received data.

19. The method of claim 17, further comprising:

storing the second portion of the data in a memory; and

where determining the level of congestion includes:

determining an amount of memory used to store the second portion of the data, and

determining the level of congestion based on the amount of memory used to store the second portion of the data.

20. The method of claim 19, where determining the level of congestion includes:

determining whether the amount of memory used to store the second portion of the data exceeds one or more of a plurality of threshold values, and

determining the level of congestion based on whether the amount of memory used to store the second portion of the data exceeds the one or more of the plurality of thresholds.

21. The method of claim 17, where determining the communication state includes:

determining a state of a connection associated with the first portion of the data,

determining a state of a connection associated with the second portion of the data, and

determining the communication state based on the state of the connection associated with the first portion of the data and the state of the connection associated with the second portion of the data.

22. The method of claim 17, where determining the communication state includes:

determining first information associated with a position of the second portion of the data in a packet signal,

determining second information regarding whether a discard processing of the packet signal has been performed,

determining third information that indicates whether the received data includes the packet signal, and

determining the communication state based on the first information, the second information, and the third information.

23. The method of claim 17, further comprising:

selectively transmitting the second portion of the data to the network device as a multiplexed signal.

24. A system comprising:

a device including one or more processors to:

maintain first communication state information, of data packets received from a switching unit,

the switching unit connecting a plurality of subscriber devices to a network,

the first communication state information including information representing one or more of connection data, position data, or discard data, associated with the data packets received from the switching unit,

maintain second communication state information, of data packets sent from a first subscriber device of the plurality of subscribers,

the second communication state information including information representing one or more of connection data, position data, or discard data, associated with data packets received from the first subscriber device,

determine a congestion level corresponding to an amount of congestion of a memory corresponding to the first subscriber device, and

selectively discard data packets received from the first subscriber device based on the congestion level, the first communication state information, and the second communication state information.

25. The system of claim 24, where the device includes a filter component for each of the plurality of subscriber devices.

26. The system of claim 25, where the one or more processors are further to:

receive data packets from the filter component for each of the plurality of subscriber devices, and

selectively transmit a multiplexed version of the data packets to the switching unit.

27. The system of claim 24, where, when determining the congestion level, the one or more processors are further to:

compare the amount of congestion to a threshold amount of congestion; and

determine the congestion level based on comparing the amount of congestion to the threshold amount of congestion.

28. The system of claim 24, where, when determining the congestion level, the one or more processors are to:

determine whether the amount of congestion exceeds one or more of a plurality of threshold amounts of congestion, and

determine the congestion level based on whether the amount of congestion exceeds one or more of the plurality of threshold amounts of congestion.

29. The system of claim 28, where, when determining the congestion level, the one or more processors are further to:

determine that the congestion level comprises a first congestion level when the congestion level does not exceed any of the plurality of threshold amounts of congestion,

determine that the congestion level comprises a second congestion level when the congestion level exceeds only one of the plurality of threshold amounts of congestion, and

determine that the congestion level comprises a third congestion level when the congestion level exceeds each of the plurality of threshold amounts of congestion.