Method for the control of flows within an ATM switch with distributed architecture

Abstract

A method for the control of flows particularly within an ATM switch transmission network with a distributed architecture. The method distributes a specified number n of shapers to each input line card, a specific shaper being dedicated to the VBRnrt (Variable Bit Rate non real time) category flows as a function of the totalized mean bit rate, and n−1 other shapers being adjusted as a function of the available bit rate (AVCR) of the output ports.

1. A method for utilizing an ATM switch with distributed architecture and storage at input to control flows within the ATM switch, the ATM switch including input line cards and output ports connected to respective of the input line cards, the method comprising:
 distributing, to each of the input line cards, a specified number n of shapers, one shaper being dedicated to a VBRnrt (variable bit rate non-real time) category flow and being adjusted as a function of a totalized mean bit rate of the VBRnrt category flow, and n−1 other shapers being dedicated to ABR (Available Bit Rate) and UBR (Unspecified Bit Rate) category flows;  
 associating, with each ABR and UBR connection, a tree whose root is constituted by its respective input line card and whose leave  leaves are constituted by respective of the output ports; and  
 assigning each tree to one of the n shapers chosen as a function of the nominal bit rate of the leaves of the tree being assigned, the chosen shaper being adjusted as a function of the available bit rate of the leaves of all the trees assigned to the chosen shaper.  
 
     
     
       2. A method according to  claim 1 , further comprising assigning, to each shaper, a first queue working in PDU application frame mode and a second queue working in cell mode. 
     
     
       3. A method according to  claim 2 , further comprising reserving the first queue to multipoint-to-point and multipoint-to-multipoint connections. 
     
     
       4. A method according to  claim 1 , further comprising assigning each shaper to a class defined by each respective shaperts  shaper's range of operation, a respective shaper of a range being in the same class as another respective shaper if the respective trees associated with each of the respective shapers has  have at least one common leaf. 
     
     
       5. A method according to  claim 4 , further comprising assigning an order of speaking to the shapers of one and the same class, limiting a right to speak to within a specified time interval T, divided into three elementary time intervals T 1 , T 2 , T 3   T 1 ,T 2 ,T 3  such that T=T 1 +T 2 +T 3   T=T 1 +T 2 +T 3 , the beginning of a transmission starting in the interval T 1   T 1 , the end of a transmission taking place in the interval T 1 +T 2   T 1 +T 2 , the turn to speak in the next interval T being determined as a function of wishes expressed by each shaper during the interval T 3   T 3  and in favoring the  a shaper that has suspended a transmission of a PDU application frame. 
     
     
       6. A method according to  claim 5 , further comprising designating a master line card for the transmission of a SYNC SIGNAL token with a periodicity T, in a general broadcast to all the line cards with a shaper of rank i, in which i is an integer, that possesses at least one tree that is associated with the shaper of rank i, bringing about the transmission of a speech token by one of the line cards whose shaper of rank i has finished transmitting before the end of the interval T 1 +T 2  to a next line card with the turn to speak in the class considered and in bringing about the transmission, by the line card that has received the SYNC SIGNAL token at the end of the period T during the time interval T 3   T 3 , of a collection token towards the master line card in which an indicator is recorded indicating a need to transmit during the following time interval T. 
     
     
       7. A method according to  claim 6 , further comprising carrying out communications in multipoint-to-multipoint mode by shifting a frame relaying function of a broadcasting server in an ATM layer of line cards by modification of input translations of point-to-point connections (Ai), which go from VLAi→(VLA′i, LG) to VLAi→(VLM,L 1 ,L 2 ,L 3 , . . . , Lq), where VLAi designates a logic channel identifier associated with (Ai) on an input artery, VLA′i designates a logic channel identifier associated with a same connection in a management module, LG designates an identity of a switch fabric junction of a management module, VLM designates a broadcasting index of the point-to-multipoint connection M, and L 1 , L 2 , . . . , Lq designate an identity of switch fabric functions concerned by the connection M. 
     
     
       8. A method according to  claim 6 , further comprising carrying out communication in multipoint-to-multipoint mode by shifting a frame relaying function of a routing emulation device in an ATM layer of the line cards by an examination of a first cell of each PDU frame to extract an IP address of an addressee therefrom, by making a search in a cache table of a pair formed by a logic channel and an outgoing direction before the extracted IP address, and using a translation obtained in all the cells of the PDU frame, the cache table being updated by routing information coming from a routing emulation function that resides in a management module and in transmitting a request for updating of the cache table if the desired IP address is not therein. 
     
     
       9. A method for controlling flow within an ATM switch with distributed architecture and storage at input, the ATM switch including input line cards and output ports connected to respective of the input line cards, the method comprising
 controlling the flow by: 
 distributing, to each of the input line cards, a specified number n of shapers, one shaper being dedicated to a VBRnrt (variable bit rate non-real time) category flow and being adjusted as a function of a totalized mean bit rate of the VBRnrt category flow, and n−1 other shapers being dedicated to ABR (Available Bit Rate) and UBR (Unspecified Bit Rate) category flows;  
 associating, with each ABR and UBR connection, a tree whose root is constituted by its respective input line card and whose leaves are constituted by respective of the output ports; and  
 assigning each tree to one of the n shapers chosen as a function of the nominal bit rate of the leaves of the tree being assigned, the chosen shaper being adjusted as a function of the available bit rate of the leaves of all the trees assigned to the chosen shaper.  
 
 
     
     
       10. A method according to  claim 9 , further comprising assigning, to each shaper, a first queue working in PDU application frame mode and a second queue working in cell mode. 
     
     
       11. A method according to  claim 10 , further comprising reserving the first queue to multipoint-to-point and multipoint-to-multipoint connections. 
     
     
       12. A method according to  claim 9 , further comprising assigning each shaper to a class defined by each respective shaper's range of operation, a respective shaper of a range being in the same class as another respective shaper if the respective trees associated with each of the respective shapers has  have at least one common leaf. 
     
     
       13. A method according to  claim 12 , further comprising assigning an order of speaking to the shapers of one and the same class, limiting a right to speak to within a specified time interval T, divided into three elementary time intervals T 1 , T 2 , T 3   T 1 , T 2 , T 3  such that T=T 1 +T 2 +T 3   T=T 1 +T 2 +T 3 , the beginning of a transmission starting in the interval T 1   T 1 , the end of a transmission taking place in the interval T 1 +T 2   T 1 +T 2 , the turn to speak in the next interval T being determined as a function of wishes expressed by each shaper during the interval T 3   T 3  and in favoring the  a shaper that has suspended a transmission of a PDU application frame. 
     
     
       14. A method according to  claim 13 , further comprising designating a master line card for the transmission of a SYNC SIGNAL token with a periodicity T, in a general broadcast to all the line cards with a shaper of rank i, in which i is an integer, that possesses at least one tree that is associated with the shaper of rank i, bringing about the transmission of a speech token by one of the line cards whose shaper of rank i has finished transmitting before the end of the interval T 1 +T 2  to a next line card with the turn to speak in the class considered and in bringing about the transmission, by the line card that has received the SYNC SIGNAL token at the end of the period T during the time interval T 3   T 3 , of a collection token towards the master line card in which an indicator is recorded indicating a need to transmit during the following time interval T. 
     
     
       15. A method according to  claim 14 , further comprising carrying out communications in multipoint-to-multipoint mode by shifting a frame relaying function of a broadcasting server in an ATM layer of line cards by modification of input translations of point-to-point connections (Ai), which go from VLAi→(VLA′i, LG) to VLAi→(VLM,L 1 ,L 2 ,L 3 , . . . , Lq), where VLAi designates a logic channel identifier associated with (Ai) on an input artery, VLA′i designates a logic channel identifier associated with a same connection in a management module, LG designates an identity of a switch fabric junction of a management module, VLM designates a broadcasting index of the point-to-multipoint connection M, and L 1 , L 2 , . . . , Lq designate an identity of switch fabric functions concerned by the connection M. 
     
     
       16. A method according to  claim 14 , further comprising carrying out communications in multipoint-to-multipoint mode by shifting the frame relaying function of a routing emulation device in an ATM layer of the line cards by an examination of a first cell of each PDU frame to extract the IP address of the addressee therefrom, by making a search in a cache table of a pair formed by a logic channel and an outgoing direction before the extracted IP address, and using the translation obtained in all the cells of the PDU frame, the cache table being updated by routing information coming from a routing emulation function that resides in the management module and in transmitting a request for updating of the cache to a management module if the desired IP address is not therein. 
     
     
       17. A method comprising:
   associating a shaper in a line card with a non - real time variable bit rate category flow, associating at least one additional shaper in the same line card with an available bit rate or unspecified bit rate category flow; and        adjusting a bit rate of each at least one additional shaper associated with an available bit rate or unspecified bit rate category flow as a function of an available bit rate of a set of output ports.      
     
     
       18. The method of  claim 17 , further comprising adjusting a bit rate of the shaper associated with the non- real time variable bit rate category flow as a function of a totalized mean bit rate of the non - real time variable bit rate category flow.    
     
     
       19. The method of  claim 17 , further comprising assigning a range of operation to each at least one additional shaper associated with an available bit rate or unspecified bit rate category flow.  
     
     
       20. The method of  claim 19 , wherein each range of operation is assigned initialization of a system that associates shapers in a line card with category flows.  
     
     
       21. The method of  claim 19 , wherein the ranges of operation include the entire range of the nominal bit rates of the set of output ports.  
     
     
       22. The method of  claim 21 , wherein a range of operation of an i+ 1  ranking shaper has a lower limit lower than a lower limit of a range of an i ranking shaper and an upper limit slightly higher than the lower limit of the range of the i ranking shaper so as to obtain a slight overlapping of the range of operation of adjacent shapers.  
     
     
       23. A non- transitory computer - readable medium having computer - executable instructions stored thereon for implementing a method for servicing communications, the method comprising:      associating a shaper in a line card with a non - real time variable bit rate category flow;        associating at least one additional shaper in the same line card with an available bit rate or unspecified bit rate category flow, and        adjusting a bit rate of each at least one additional shaper associated with an available bit rate or unspecified bit rate category flow as a function of an available bit rate of a set of output ports.      
     
     
       24. The non- transitory computer - readable medium of    claim 23   , further comprising adjusting a bit rate of the shaper associated with the non - real time variable bit rate category flow as a function of a totalized mean bit rate of the non - real time variable bit rate category flow.    
     
     
       25. The non- transitory computer - readable medium of    claim 23   , further comprising assigning a range of operation to each at least one additional shaper associated with an available bit rate or unspecified bit rate category flow.    
     
     
       26. The non- transitory computer - readable medium of    claim 25   , wherein each range of operation is assigned at initialization of a system that associates shapers in a line card with category flows.    
     
     
       27. The non- transitory computer - readable medium of    claim 25   , wherein the ranges of operation include the entire range of the nominal bit rates of the set of output ports.    
     
     
       28. The non- transitory computer - readable medium of    claim 27   , wherein a range of operation of an i+ 1  ranking shaper has a lower limit lower than a lower limit of a range of an i ranking shaper and an upper limit slightly higher than the lower limit of the range of the i ranking shaper so as to obtain a slight overlapping of the range of operation of adjacent shapers.    
     
     
       29. A method for servicing ATM communications, the method comprising associating a tree constituted by a data element of an input line card and output ports to a point- to - multipoint connection; and      assigning a shaper to the tree as a function of the nominal bit rate of a leaf of the tree having the lowest nominal bit rate, the shaper for processing cells of the point - to - multipoint connection associated with the tree.      
     
     
       30. The method of  claim 29 , wherein assigning per further includes selecting a shaper from a plurality of shapers, wherein each of the plurality of shapers is associated with a flow category.  
     
     
       31. The method of  claim 30 , wherein the bit rate of a shaper is computed as a minimum of available bit rates of the leaves belonging to any trees assigned to the shaper, the method further comprising adjusting the bit rate of a shaper based on the associated category flow.  
     
     
       32. The method of  claim 31 , wherein a shaper is associated with a variable bit rate non- real time category flow, and wherein adjusting the bit rate of the shaper associated with a variable bit rate non - real time category flow includes adjusting the bit rate as a function of totalized mean bit rate of the leaves of a tree.    
     
     
       33. The method of  claim 32 , wherein a shaper is associated with an available bit rate or unspecified bit rate category flow and wherein adjusting the bit rate of the shaper associated with the available bit rate or unspecified bit rate category flow includes adjusting the bit rate as a function of the available bit rate of the leaves of a tree.  
     
     
       34. A non- transitory computer - readable medium having computer - executable instructions stored thereon for implementing a method for servicing communications, the method comprising:      associating a tree constituted by a data element of an input line card and output ports to a point - to - multipoint connection; and        assigning a shaper to the tree as a function of the nominal bit rate of a leaf of the tree having the lowest nominal bit rate the shaper for processing cells of the point - to - multipoint connection associated with the tree.      
     
     
       35. The non- transitory computer - readable medium of    claim 34    wherein assigning a shaper further includes selecting a shaper from a plurality of shapers, wherein each of the plurality of shapers is associated with a flow category.    
     
     
       36. The non- transitory computer - readable medium of    claim 35   , wherein the bit rate of a shaper is computed as a minimum of available bit rates of the leaves belonging to any trees assigned to the shaper, the method further comprising adjusting the bit rate of a shaper based on the associated category low.    
     
     
       37. The non- transitory computer - readable medium of    claim 36   , wherein a shaper is associated with a variable bit rate non - real time category flow, and wherein adjusting the bit rate of the shaper associated with a variable bit rate non - real time category flow includes adjusting the bit rate as a function of the totalized mean bit rate of the leaves of a tree.    
     
     
       38. The non- transitory computer - readable medium of    claim 37   , wherein a shaper is associated with an available bit rate or unspecified bit rate category flow and wherein adjusting the bit rate of the shaper associated with the available bit rate or unspecified bit rate category flow includes adjusting the bit rate as a function of the available bit rate of the leaves of a tree.    
     
     
       39. In an ATM communications network including, a first line card having two or more shapers including; a first shaper associated with a first range of operation and a second line card having two or more shapers including a second shaper associated with the first range of operation, wherein the first and second shapers are adapted to send cells to a common output port, a method comprising:
   obtaining an output port request by the first line card;        obtaining an output port request by the second line card; and        arbitrating access to the common output port, wherein arbitrating access occurs between two queues, a priority queue P 1  which works in protocol data unit application frame mode and a queue P 2  which works in cell mode.      
     
     
       40. The method of  claim 39 , wherein the first and second shapers are associated with an available bit rate or unspecified bit rate category flow.  
     
     
       41. The method of  claim 39 , wherein arbitrating access to the common port includes assigning, an order of speaking to the first and second shapers during a time interval favoring the shaper that has suspended a transmission of a protocol data unit application frame.  
     
     
       42. A non- transitory computer - readable medium having, computer - executable instructions stored thereon for implementing a method in an ATM communications network including a first line card having, two or more shapers including, a first shaper associated with a first range of operation and a second line card having, two or more shapers including a second shaper associated with the first range of operation, wherein the first and second shapers are adapted to send cells to a common output port, a method comprising:      obtaining, an output port request by the first line card,        obtaining an output port request by the second line card; and        arbitrating access to the common output port.      
     
     
       43. The non- transitory computer - readable medium of    claim 42   , wherein the first and second shapers are associated with all available bit rate or unspecified bit rate category flow.    
     
     
       44. The non- transitory computer - readable medium of    claim 42   , wherein arbitrating access to the common port includes assigning an order of speaking to the first and second shapers during a time interval favoring the shaper that has suspended a transmission of a protocol data unit application frame.