Forward error correction coding in ethernet networks

• US 20050005189A1
• Filed: 04/25/2002
• Published: 01/06/2005
• Est. Priority Date: 04/25/2002
• Status: Active Grant

• Current Assignee: PMC-Sierra Israel Ltd. (Microchip Technology Incorporated)
• Original Assignee: PMC-Sierra Israel Ltd. (Microchip Technology Incorporated)
• Inventors: Haran, Onn, Maislos, Ariel, Khermosh, Lior
• Granted Patent: US 7,343,540 B2
• Time in Patent Office: Days
• US Class Current: 714/4
• CPC Class Codes: H04L 1/0057 Block codes H04L1/0061, H04L1/0064 take precedence
• CPC Class Codes: H04L 1/0084 Formats for payload data

Abstract

A method for improving the bit error rate of Ethernet packets applies forward error correction (FEC) coding to transmitted packets. The FEC coding is systematic block coding, and is applied so that the coded packets can be interpreted by legacy network devices that are not capable of FEC decoding. The transmit and receive state machines of FEC-capable Ethernet nodes are modified to enable the nodes to encode and/or decode the packets with the FEC code, and to adapt the nodes'"'"' respective medium access layer (MAC) and physical layer (PHY) data rates.

Claims

1. A method for sending data packets from a transmitter to a receiver through a shared-medium digital transmission network, each packet ending in a stop symbols field delimiting the end of said each packet, the method comprising the steps of:

dividing data of at least a portion of said each packet into one or more frames;

applying a systematic FEC block code to each frame of said each packet, said systematic FEC block code keeping the data symbols of the frames visible and adding parity-check symbols;

gathering together the parity-check symbols of the frames of said each packet in a parity-check field of said each packet;

modifying said each packet by adding, after the stop symbols field delimiting the end of said each packet, the parity-check field of said each packet, and adding, after the parity-check field of said each packet, a delimiter of the parity-check field of said each packet, thereby creating a plurality of modified packets; and

transmitting the modified packets from the transmitter to the receiver through the Ethernet network.

2. The method of claim 1, wherein the shared-medium digital transmission network is an Ethernet network.

3. The method of claim 2, wherein the data packets comprise a first data packet and a second data packet, and said applying step comprises the step of applying a first FEC block code to each frame of the first packet, and applying a second FEC block code to each frame of the second packet.

4. The method of claim 2, wherein the step of applying a systematic FEC block code comprises the step of applying a Reed-Solomon code to each frame of said each packet.

5. The method of claim 2, wherein the step of applying a systematic FEC block code comprises the step of applying a Reed-Solomon GF(2⁸)−

(255, 255-2t, t) code to each frame of said each packet.

6. The method of claim 2, wherein the data of at least a portion of said each packet comprises a header field, a payload data field, and a frame check sequence field.

7. The method of claim 2, wherein the data of at least a portion of said each packet comprises a preamble field, a start frame delimiter field, a header field, a payload data field, and a frame check sequence field.

8. The method of claim 2, wherein:

the frames comprise frames of a first predetermined length and short frames, the length of each short frame being shorter than the first predetermined length; and

the step of applying a systematic FEC block code to each frame of said each packet comprises the step of virtually zero padding the short frames.

9. The method of claim 8, wherein the step of applying a systematic FEC block code comprises a step for line rate encoding each short frame.

10. The method of claim 2, wherein the transmitter comprises a MAC layer and a PHY layer, and said each packet is followed by an interpacket gap interval, the method further comprising the step of adapting transmission rate of the MAC layer to transmission rate of the PHY layer.

11. The method of claim 10, wherein the step of adapting transmission rate comprises the step of interpacket gap interval stretching in the MAC layer of the transmitter.

12. The method of claim 11, wherein the step of interpacket gap interval stretching comprises the step of extending interpacket gap interval at the end of said each packet in linear proportion to the length of said each packet.

13. The method of claim 2, wherein the receiver comprises a MAC layer and a PHY layer, further comprising the steps of:

receiving the transmitted packets at the receiver; and

adapting reception rate of the MAC layer to the reception rate of the PHY layer.

14. The method of claim 13, wherein the step of adapting reception rate comprises the step of inserting idle symbols from the PHY layer to the MAC layer in place of parity-check symbols of the received packets.

15. The method of claim 14, wherein the step of adapting reception rate further comprises the step of asserting and negating a control signal to allow the MAC layer to transfer data to the PHY layer only when the PHY layer can accept the data.

16. The method of claim 15, wherein the control signal is a CRS signal.

17. The method of claim 15, wherein the control signal is a COL signal.

18. The method of claim 2, further comprising the step of 8b/10b line coding of the modified packets before said transmitting step, wherein:

said each packet comprises a start symbols field delimiting the beginning of said each packet;

the transmitter comprises a MAC layer and a PHY layer; and

the modifying step comprises the step of extending the start symbols field and the stop symbols field of said each packet to short sequences of 10B that maintain a False_carrier_detect mode.

19. The method of claim 2, further comprising the step of 8b/10b line coding of the modified packets before said transmitting step, wherein:

said each packet comprises a start symbols field delimiting the beginning of said each packet;

the transmitter comprises a MAC layer and a PHY layer; and

the modifying step comprises a step for decreasing error rate of characters in the start symbols and stop symbols fields of said each packet.

20. The method of claim 2, further comprising the step of communicating packets between the transmitter and receiver to register the receiver with the transmitter and to determine the systematic FEC block code.

21. The method of claim 2, wherein the step of applying comprises the step of applying at least one systematic FEC block code to each frame of said packet, the method further comprising the step of selecting a systematic FEC block code for applying to the frames of said each packet based on the length of said each packet.

22. The method of claim 2, wherein the data of at least a portion of said each packet comprises a header field, a payload data field, and a frame check sequence field, the frame check sequence field of said each packet comprising a cyclic redundancy check value of said each packet, the method further comprising the steps of:

receiving the modified packets at the receiver;

reading the frame check sequence field of each modified packet received at the receiver;

using the cyclic redundancy check in the frame check sequence field of said each modified packet received at the receiver to verify integrity of said each modified packet received at the receiver;

if the cyclic redundancy check in the frame check sequence field of said each modified packet received at the receiver indicates that integrity of said each modified packet received at the receiver has been compromised, performing FEC decoding on said each modified packet received at the receiver.

23. A method for receiving data packets sent from a transmitter to a receiver through an Ethernet network, the receiver comprising a MAC layer and a PHY layer, each packet ending in a second stop symbols field delimiting the end of said each packet, said each packet followed by an interpacket gap interval, said each packet comprising a parity-check field at the end of said each packet and before the second stop symbols field, said each packet comprising a first stop symbols field before the parity check field, said each packet comprising data in a data field preceding the first stop symbols field, the parity-check field of said each packet comprising parity-check symbols resulting from encoding frames of at least a portion of the data of said each packet with a systematic FEC block code, the method comprising the steps of:

receiving the packets at the receiver;

FEC decoding each received packet; and

adapting reception rate of the MAC layer to the reception rate of the PHY layer.

24. The method of claim 23, wherein the step of adapting comprises the step of inserting idle symbols from the PHY layer to the MAC layer in place of parity-check symbols of the received packets.

25. The method of claim 24, wherein the packets are encoded with a Reed-Solomon code.

26. A method of improving bit error rate of data packet transmission between a transmitter and a receiver coupled together by an Ethernet network, the method comprising the steps of:

(a) obtaining a bit error rate for non-FEC-encoded packet transmission from the transmitter to the receiver;

(b) comparing the bit error rate obtained to a predefined bit error rate limit;

(c) if the bit error rate obtained exceeds the predefined bit error rate limit, performing the following steps (d) through (h);

(d) dividing the data of at least a portion of each packet intended for transmission from the transmitter to the receiver through the network into one or more frames;

(e) applying a systematic FEC block code to each frame of said each packet, said systematic FEC block code keeping the data symbols of the frames visible and adding parity-check symbols;

(f) gathering together the parity-check symbols of the frames of said each packet in a parity-check field of said each packet;

(g) modifying said each packet by adding, after stop symbols field delimiting the end of said each packet, the parity-check field of said each packet, and adding, after the parity-check field of said each packet, a delimiter of the parity-check field of said each packet, thereby creating a plurality of modified packets; and

(h) transmitting the modified packets from the transmitter to the receiver through the Ethernet network.

27. A method of improving bit error rate according to claim 26, wherein the step of applying a systematic FEC block code comprises the step of applying a Reed-Solomon code to each frame of said each packet.

28. A method of improving bit error rate according to claim 27, wherein the step of obtaining comprises the step of estimating the bit error rate for non-FEC-encoded packet transmission from the transmitter to the receiver based on architectural parameters of the Ethernet network.

29. A method of improving bit error rate according to claim 27, wherein the step of obtaining comprises the step of measuring the bit error rate of at least one non-FEC-encoded packet sent from the transmitter to the receiver through the Ethernet network.

30. A method of improving bit error rate of data packet transmission between a first network element and a second network element that is not capable of FEC decoding, the first and the second network elements being coupled by an Ethernet network, the method comprising the step of providing an adaptation device, between the Ethernet network and the second network element, the adaptation device being capable of performing the following steps:

receiving, from the Ethernet network, FEC encoded packets sent from the first network element to the second network element, the packets received from the Ethernet network being FEC encoded so as not to cause MAC layer errors when received at Ethernet-compliant network elements that are not capable of FEC decoding;

detecting boundaries of the packets received from the Ethernet network;

FEC decoding the packets received from the Ethernet network, thereby creating a plurality of FEC decoded packets;

adapting the FEC decoded packets by inserting idle symbols in place of FEC parity-check symbols of the FEC decoded packets, thereby creating a plurality of adapted packets;

transmitting the adapted packets to the second network element;

receiving, from the second network element, packets sent from the second network element to the first network element;

detecting boundaries of each packet received from the second network element;

dividing data of at least a portion of said each packet received from the second network element into one or more frames;

applying a systematic FEC block code to each frame of said each packet, said systematic FEC block code keeping the data symbols of the frames visible and adding parity-check symbols;

gathering together the parity-check symbols of the frames of said each packet in a parity-check field of said each packet;

modifying said each packet by adding, after the stop symbols field delimiting the end of said each packet, the parity-check field of said each packet, and adding, after the parity-check field of said each packet, a delimiter of the parity-check field of said each packet, thereby creating a plurality of modified packets;

transmitting the modified packets from the transmitter to the receiver through the Ethernet network; and

adapting transmission rate of the second network element to the Ethernet network.

31. A method of improving bit error rate according to claim 30, wherein:

the packets received from the Ethernet network are FEC encoded using a Reed-Solomon code; and

said step of applying a systematic FEC block code comprises the step of applying a Reed-Solomon code to each frame of said each packet.

32. A method for sending data packets from a transmitter to a receiver through an Ethernet network, the method comprising:

a step for applying at least one systematic FEC block code to data of the data packets to encode the data packets so that the encoded data packets do not cause MAC layer errors when the encoded data packets are received at non-FEC-capable Ethernet-compliant network elements, thereby creating a plurality of encoded packets; and

sending the encoded packets from the transmitter to the receiver through the Ethernet network.

33. A method according to claim 32, further comprising:

receiving the encoded packets; and

decoding the encoded packets.

34. A network element for sending data packets to a receiver through an Ethernet network, each packet ending in a stop symbols field delimiting the end of said each packet and followed by an interpacket gap interval, the network element comprising:

a framer that divides at least a portion of said each packet into one or more frames;

an FEC encoder that applies at least one systematic FEC block code to the frames, said at least one systematic FEC block code keeping data symbols of the frames visible and calculating parity-check symbols;

a packet modifier that modifies said each packet by adding, after the stop symbols field delimiting the end of said each packet, the calculated parity-check symbols of the frames of said each packet, and by adding, after the calculated parity-check symbols of the frames of said each packet, a delimiter of the parity-check field of said each packet, thereby creating aof modified packets; and

a transmitter that transmits the modified packets to the receiver.

35. A network element according to claim 34, wherein said at least one systematic FEC block code comprises a first systematic FEC block code and a second systematic FEC block code, and the FEC encoder applies the first systematic FEC block code to frames of a first packet of the data packets and the second systematic FEC block code to frames of a second packet of the data packets.

36. A network element according to claim 34, wherein said at least one systematic FEC block code comprises a Reed-Solomon code.

37. A network element according to claim 34, wherein said at least one systematic FEC block code comprises a Reed-Solomon GF(2⁸)−

(255, 255-2t,t) code.

38. A network element according to claim 34, wherein the portion of said each packet divided by the framer comprises a header field, a payload data field, and a frame check sequence field.

39. A network element according to claim 34, wherein the portion of said each packet divided by the framer comprises a preamble field, a start frame delimiter field, a header field, a payload data field, and a frame check sequence field.

40. A network element according to claim 34, wherein:

the framer divides at least a portion of said each packet into frames of a first predetermined length and short frames, the length of each short frame being shorter than the first predetermined length; and

the FEC encoder virtually zero pads the short frames.

41. A network element according to claim 34, further comprising:

a MAC layer;

a PHY layer; and

a rate adaptation mechanism that adapts transmission rate of the MAC layer to transmission rate of the PHY layer.

42. A network element according to claim 34, further comprising:

a MAC layer;

a PHY layer; and

a rate adaptation mechanism that adapts transmission rate of the MAC layer to transmission rate of the PHY layer by interpacket gap interval stretching.

43. A network element according to claim 34, further comprising:

a MAC layer;

a PHY layer; and

a rate adaptation mechanism that adapts transmission rate of the MAC layer to transmission rate of the PHY layer by linearly stretching an interpacket gap interval following said each packet in proportion to the length of said each packet.

44. A network element according to claim 34, further comprising:

a MAC layer;

a PHY layer;

an 8b/10b line encoder;

wherein;

said each packet further comprises a start symbols field delimiting the beginning of said each packet;

the packet modifier comprises a 10B extender that extends the start symbols field and the stop symbols field of said each packet to short sequences of 10B that maintain a False_carrier_detect mode.

45. A network element according to claim 44, further comprising means for implementation of a transmit ordered_set state machine of a physical coding sublayer.

46. A component of a physical coding sublayer of an Ethernet transmitter sending data packets to a receiver through an Ethernet network, each packet ending in a stop symbols field delimiting the end of said each packet, the component comprising:

an FEC encoder that divides at least a portion of said each packet into one or more frames and applies at least one systematic FEC block code to the frames, said at least one systematic FEC block code keeping data symbols of the frames visible and calculating parity-check symbols;

a packet modifier that modifies said each packet by adding, after the stop symbols field delimiting the end of said each packet, the parity-check symbols of the frames of said each packet, and by adding, after the parity-check symbols of the frames of said each packet, a delimiter of the parity-check field of said each packet, thereby creating a plurality of modified packets; and

a data wrapper that adds packet boundary symbols to the modified packets.

47. A component of a physical coding sublayer in accordance with claim 46, further comprising a rate adapter that receives the packets from a reconciliation sublayer of the Ethernet transmitter and adapts transmission rate of a MAC layer of the Ethernet transmitter to transmission rate of a PHY layer of the Ethernet transmitter.

48. A component of a physical coding sublayer in accordance with claim 47, wherein the at least one systematic FEC block code comprises a Reed-Solomon code.

49. A component of a physical coding sublayer in accordance with claim 47, further comprising a line code encoder that applies a line code to the packets with the added boundary symbols.

50. A component of a physical coding sublayer in accordance with claim 49, wherein:

the FEC encoder divides at least a portion of said each packet into frames of a first predetermined length and short frames, the length of each short frame being shorter than the first predetermined length, and

the FEC encoder virtually zero pads the short frames before applying the at least one systematic FEC block code to the frames.

51. A component of a physical coding sublayer in accordance with claim 46, wherein the portion of said each packet divided into one or more frames comprises a header field, a payload data field, and a frame check sequence field.

52. A component of a physical coding sublayer in accordance with claim 50, wherein the FEC encoder comprises means for flexible encoding of (N, N−

2t, t) Reed-Solomon codes with a variable t, t being less than or equal to a first predefined complexity limit.

53. A component of a physical coding sublayer in accordance with claim 46, wherein the portion of said each packet divided into one or more frames comprises a preamble field, a start frame delimiter field, a header field, a payload data field, and a frame check sequence field.

54. A component of a physical coding sublayer in accordance with claim 46, wherein said each packet further comprises a start symbols field, the component further comprising a 10B extender that extends the start symbols field and the stop symbols field of said each packet to short sequences of 10B that maintain a False_carrier_detect mode.

55. A component of a physical coding sublayer of an Ethernet receiver that receives data packets from a transmitter through an Ethernet network, each packet being encoded with at least one systematic FEC block code, the network element comprising:

a line code decoder that applies line code decoding to the data packets, thereby creating line code decoded packets;

a boundary detector that determines the boundaries of the line code decoded packets;

a selector that separates FEC parity-check symbols of the line code decoded packets;

a buffer that replaces the FEC parity-check symbols in each line code decoded packet with idle symbols, and buffers said each line code decoded packet, thereby creating buffered packets; and

an FEC decoder that receives the buffered packets and the FEC parity-check symbols and FEC decodes each buffered packet.

56. A component of a physical coding sublayer of an Ethernet receiver in accordance with claim 55, wherein the FEC decoder is capable of performing steps of a method for decoding of (N, N−

2t, t) Reed-Solomon codes with a variable t, t being less than or equal to a first predefined complexity limit.