Node architecture for communication networks

• US 4,821,265 A
• Filed: 04/06/1987
• Issued: 04/11/1989
• Est. Priority Date: 04/06/1987
• Status: Expired due to Fees

• Current Assignee: Racal-Datacom, Inc. (Thales SA)
• Original Assignee: Racal Data Communications Incorporated
• Inventors: Albal, Nandakishore A., Goyal, Praduemn K.
• Primary Examiner(s): Olms, Douglas W.
• Assistant Examiner(s): Chin, Wellington
• Application Number:
  US07/035,130
• Time in Patent Office: 736 Days
• Field of Search: 370/60, 370/85, 370/94, 370/110.1
• US Class Current: 370/469
• CPC Class Codes: H04Q 11/0457 Connection protocols
• CPC Class Codes: H04Q 2213/05 Software aspects
• CPC Class Codes: H04Q 2213/107 Control equipment for part of the connection

Abstract

An architecture for a switching node of a communication network includes a dedicated hardware Layer 1 processing portion and a Layer 2 and 3 processing portion based on multiple programmed general purpose processors. An array of such processors are used with an arbitration scheme for selecting which of the processors is to be used for any given Layer 2 or Layer 3 process. The architecture allows the node'"'"'s capacity to be expanded by simply adding more processors to handle increased traffic.

Claims

1. A node for a communication system for processing a D-Channel ISDN message, comprising in combination:

interface means for receiving an ISDN signal from a transmission line;

demultiplexing means for demultiplexing said ISDN signal into a D-Channel message and for performing Physical Layer processes on said D-Channel message;

a processor bank including a plurality of processors which are interconnected by a bus for receiving said D-Channel message after said D-Channel message is processed by said demultiplexing means and for performing a Data Link Layer process on said D-Channel message; and

arbitration means for determining which of said plurality of processors is to perform said Data Link Layer process on said D-Channel message.

2. The apparatus of claim 1, wherein said arbitration means further comprises means for establishing a priority number for each of said processors in said processor bank, and wherein a processor having highest priority number is selected to perform said Data Link Layer process.

3. The apparatus of claim 2, further comprising computing means for computing said priority based upon a number of previously unsuccessful attempts by the processor to access said bus.

4. The apparatus of claim 3, wherein said computing means further comprises means for computing a priority number PN representing said priority as follows:

space="preserve" listing-type="equation">PN=CON{β

F,N}

β

is a constant;

F is a number of bus access failures; and

N is a processor number.

5. The apparatus of claim 2, further comprising an arbitration bus and wherein each of said processors applies its priority number to said arbitration bus so that the processor with the highest priority is selected.

6. The apparatus of claim 1, wherein said processor bank includes a plurality of processors which are capable of performing either Data Link Layer or Network Layer processes and wherein said arbitration means further comprises means for determining which of said plurality of processors is to perform said Network Layer process.

7. The apparatus of claim 1, wherein each of said processors further comprises a computer and gateway means for interfacing said computer to a common set of communication buses so that communication on said buses is essentially independent of characteristics of said computer.

8. The apparatus of claim 1, wherein said processor bank further comprises a plurality of processors for computing a Network Layer process and wherein said arbitration means determines which of said processors performs said Network Layer process.

9. The apparatus of claim 4, wherein β

is a different value for two of said processors.

10. The apparatus of claim 1, further comprising a plurality of queues for containing a plurality of segments of D-channel information to be processed by said processor bank.

11. The apparatus of claim 10, further comprising queue arbitration means for determining which of said plurality of queues is to be serviced by said processor bank.

12. The apparatus of claim 11, wherein said queue arbitration means include means for computing a priority number for each of said queues and wherein said processor bank processes the information from the queue with the highest of said priority numbers.

13. The apparatus of claim 12, wherein each of said queues attempts to access said bus and wherein said computing means computes the priority number PN for each queue as follows:

space="preserve" listing-type="equation">PN=CON{(α

L+β

F),(N)}

Where;

α and

β

are constants,

L=a queue length,

F=the number of unsuccessful attempts to access said bus, and

N=a processor number associated with each of said processors.

14. The apparatus of claim 13, wherein said arbitration means performs a bit by bit arbitration on the priority numbers starting with the most significant bit and proceeding through the least significant bit.

15. The apparatus of claim 1, wherein each of said plurality of processors include a microprocessor.

16. The apparatus of claim 15, wherein each of said processors includes a gateway means for providing a common communication format with said demultiplexing means.

17. The apparatus of claim 4, wherein said microcomputer includes a nonvolatile memory, and wherein β

is stored in said nonvolatile memory.

18. The apparatus of claim 17, wherein said nonvolatile memory includes a read only memory.

19. The apparatus of claim 18, further including a plurality of circuit slots each associated with one of said plurality of processors as an indication of the physical location of installation of each of said processors and wherein N is unique for each of said slots.

20. A method for processing a D Channel ISDN message, comprising the steps of:

receiving an ISDN signal from a transmission line;

demultiplexing said ISDN signal to produce a D Channel message;

performing a Physical Layer process on said D Channel message;

selecting one of a plurality of available processors interconnected by a bus to perform a Data Link Layer process on said D Channel message.

21. The method of claim 20, further comprising the step of performing said Data Link Layer process after said Physical Layer process has been completed.

22. The method of claim 21, further comprising the step of selecting one of said plurality of available processors to perform a Network Layer process on said D Channel message after said Data Link Layer process has been completed.

23. The method of claim 22, further comprising the step of performing said Network Layer process after said Data Link process.

24. The method of claim 20, wherein said selecting step includes the step of arbitrating among said plurality of processors, of which at least two are attempting to access said bus.

25. The method of claim 24, wherein said arbitration step includes computing a priority number designated PN for each of said plurality of processors.

26. The method of claim 25, wherein said priority number PN is:

space="preserve" listing-type="equation">PN═

CON{β

F,N}

where;

β

is a constant;

F is a number of bus access failures; and

N is a processor number associated with each one of said plurality of processors.

27. The method of claim 26, wherein said plurality of processors include at least a first processor having computing power greater than that of a second processor and wherein β

is greater for said first processor than said second processor.

28. The method of claim 26, wherein β

is not equal to 1.