Method and apparatus for automatic rate adaptation in a DOCSIS upstream

• US 7,315,967 B2
• Filed: 07/31/2003
• Issued: 01/01/2008
• Est. Priority Date: 07/31/2003
• Status: Active Grant

• Current Assignee: Google Technology Holdings LLC (Alphabet Inc.)
• Original Assignee: Terayon Communications Systems Incorporated (Motorola Solutions, Inc.)
• Inventors: Rakib, Selim Shlomo, Azenko, Yehuda
• Primary Examiner(s): TU, CHRISTINE TRINH LE
• Application Number:
  US10/632,660
• Publication Number:
  US 20050039103A1
• Time in Patent Office: 1,615 Days
• Field of Search: None
• US Class Current: 714/704
• CPC Class Codes: H03M 13/1515 Reed-Solomon codes
• CPC Class Codes: H03M 13/353 Adaptation to the channel
• CPC Class Codes: H04L 1/0002 by adapting the transmission rate
• CPC Class Codes: H04L 1/0057 Block codes H04L1/0061, H04L1/0064 take precedence
• CPC Class Codes: H04L 1/0061 Error detection codes
• CPC Class Codes: H04L 1/0072 Error control for data other than payload data, e.g. control data
• CPC Class Codes: H04L 1/20 using signal quality detector
• CPC Class Codes: H04L 2001/0093 Point-to-multipoint

Abstract

A process for automatically monitoring the packet error rate of a logical channel in a cable television DOCSIS data transmission system and automatically altering the bit rate on the channel or sending a message to a cable operator recommending a change in burst profile and bit rate. The preferred embodiment of the process comprises determining the dominant noise type and SNR of the channel and selecting a set of burst profiles best adapted for the dominant noise type and selecting an initial burst profile based upon the SNR. Then the packet error rate of the channel is monitored and compared to one or more thresholds to decide whether a change in bit rate is required to bring the packet error rate within limits. If a change is bit rate is indicated, automatically selecting a new burst profile with an appropriately changed bit rate and automatically generating and sending a UCD message defining the channel'"'"'s new burst profile.

Claims

1. A process for automatically altering the data rate on a logical channel with changing noise conditions, comprising:

1) resetting a flawed packet counter for said logical channel;

2) resetting a total packets received counter for said logical channel;

3) receiving a packet on said logical channel, and incrementing the total packet received counter;

4) processing error detection data in said packet to determine if there is an error in the packet, and, if so, incrementing said flawed packet counter;

5) comparing the count in said total packet counter to a number representing the desired number of packets to be received before a determination of packet loss percentage is made;

6) if the number of packets received is less than the desired number, returning to step 3;

7) if the number of packets received is equal to or exceeds the desired number of packets received, calculating a packet loss percentage by dividing the number of flawed packets by the total number of packets received and comparing the packet loss percentage to one or more packet loss thresholds wherein the comparison includes

A) first comparing said packet loss percentage to a first threshold and calculating said first threshold by evaluating the expression;

packet_loss_required*(1+TH1)

where packet_loss_required is a programmable number representing the desired maximum packet loss percentage and TH1 is a constant representing a first hysteresis threshold;

B) if said packet loss percentage is greater than or equal to said first threshold, making the determination that a reduction in data rate is required and proceeding to step 9;

C) if said packet loss percentage is not greater than or equal to said first threshold, comparing said packet loss percentage to a second threshold and calculating said second threshold by evaluating the expression;

packet_loss_required*(1+TH2)

where packet_loss_required is a programmable number representing the desired maximum packet loss percentage and TH2 is a constant representing a second hysteresis threshold;

D) if said packet loss percentage is less than or equal to said second threshold, making the determination that an increase in data rate is required and proceeding to step 9;

8) determining if a change in data rate throughput is required based upon the comparison of the packet loss percentage to the one or more packet loss thresholds; and

9) if a change in data rate throughput is required, generating a signal indicating the need for a change in data rate for said logical channel.

2. The process of claim 1 further comprising the steps of determining in step 3 whether an incoming packet is transmitted in response to a contention grant, and, if said packet is transmitted in response to a contention grant, discarding said packet and not incrementing the packet received counter or the flawed packet counter, and further comprising a step 10 comprising the substeps of either selecting a new burst profile with an appropriate data rate throughput and generating and transmitting a upstream channel descriptor message setting new data rate throughput parameters of said selected burst profile, or selecting a new burst profile and generating a visible or audible notification to a system operator informing of the need for a change in data rate throughput and suggesting the new burst profile that should be used, and returning to step 1 for a new logical channel.

3. The process of claim 1 further comprising a step performed before any processing to determine if an operator has turned automatic rate adaptation on for all logical channels.

4. The process of claim 1 further comprising a step performed before any processing to determine if an operator has turned automatic rate adaptation on for the particular logical channel to be processed.

5. The process of claim 1 further comprising the step of generating an upstream channel descriptor message having descriptor data that defines a data rate that is different from the existing data rate of said logical channel and appropriate to the currently existing noise conditions, and transmitting said upstream channel descriptor message at least to all cable modems that may be assigned to use said logical channel.

6. The process of claim 1 further comprising the steps:

if step 8 determines that a change in data rate throughput is not required, returning to step 1 and begin processing for a new logical channel; and

if step 9 determines that a change in data rate is required, selecting a new burst profile with a data rate which is changed appropriately for the current noise conditions on said logical channel, and generating and transmitting an upstream channel descriptor message that contains data that controls cable modems to transmit using the changed data rate.

7. The process of claim 1 further comprising the steps:

picking a new logical channel and repeating the process if the packet loss percentage is determined to be greater than said second threshold.

8. A process for automatically altering the data rate on a logical channel with changing noise conditions, comprising:

1) resetting an average noise filter or accumulator or memory location for said logical channel;

2) resetting a total packets received counter for said logical channel;

3) receiving a packet on said logical channel, and incrementing the total packet received counter if said packet is not transmitted in response to a contention grant and discarding said packet if it is transmitted in response to a contention grant;

4) processing a received packet to determine the average noise if the packet has not been discarded and updating a running average noise value using the average noise in the packet;

5) comparing the count in said total packet counter to a number representing the desired number of packets to be received before a determination of channel signal-to-noise ratio is made;

6) if the number of packets received is less than the desired number, returning to step 3; and

7) if the number of packets received is equal to or exceeds the desired number of packets received, determining the SNR of the logical channel, determining if the SNR of the logical channel is less than the value SNR_REQUIRED*(1−

TH1) where SNIR_REQUIRED is a desired SNR for the logical channel and TH1 is a burst profile in use representing a first SNIR threshold, reducing the data rate of the logical channel if the SNR of the logical channel is lower than or equal to the value SNR_REQUIRED*(1−

TH1), and increasing the data rate of the logical channel if the SNR of the logical channel is greater than SNR_REQUIRED*(1−

TH2).

9. The process of claim 8 wherein if the SNR of the logical channel is lower than or equal to the value SNR_REQUIRED*(1−

TH1), the data rate is reduced by selecting a next burst profile down from a currently selected burst profile for the logical channel being processed, said next burst profile down being selected from a table of burst profiles arranged in order of descending data rates.

10. The process of claim 8 wherein if the SNR of the logical channel is lower than or equal to the value SNR_REQUIRED*(1−

TH1), the data rate is increased by selecting a next burst profile up from a currently selected burst profile for the logical channel being processed, said next burst profile up being selected from a table of burst profiles arranged in order of ascending data rates.

11. The process of claim 8 further comprising:

generating a signal which can be seen or heard by an operator suggesting a change in data rate for the channel such that said operator can manually select a new data rate and cause a new upstream channel descriptor message to be generated and sent if the operators chooses to change the data rate.

12. A process for automatically altering the data rate on a logical channel with changing noise conditions comprising:

1) resetting a total uncorrectable codeword count for said logical channel for an interval;

2) resetting a total codewords received count for said logical channel for said interval;

3) receiving a burst on said logical channel, and adding the total codewords received in the burst to the running total of codewords received for said interval;

4) receiving information regarding the number of uncorrectable codewords in said received burst and adding said number of uncorrectable codewords to a running total of uncorrectable codewords for the logical channel over said interval;

5) determining in any way whether said interval has been completed;

6) if said interval has not been completed, returning to step 3;

7) if said interval has been completed, calculating a byte error rate of said logical channel and comparing the byte error rate to one or more byte error rate thresholds, the byte error rate calculated by evaluating the expression

ByteErrorRate=[(T+1)*n_error]/[(k+2T)*n_total]

where T is the number of maximum correctable bytes in an Reed-Solomon (hereafter R-S) codeword n_error is the total number of uncorrectable R-S codewords for said interval from predetermined interval usage code (IUC) burst type(s), n_total is the total number of received R-S codewords from predetermined IUC burst type(s) received within said time interval, k is the number of information bytes in an R-S codeword of a specific IUC; and

(k+2T) stands for;

the R-S codeword length in bytes;

8) determining if a change in data rate throughput is required based upon the comparison of the byte error rate to the one or more byte error rate thresholds; and

9) if a change in data rate throughput is required, generating a signal indicating a need for a change in the data rate of said logical channel.

13. The process of claim 12 wherein step (9) comprises reducing the data rate by selecting the next burst profile down from the currently selected burst profile for the logical channel being processed, said next burst profile down being selected from a table of burst profiles arranged in order of descending data rates and generating and transmitting downstream and upstream channel descriptor message containing the parameters for transmission on said logical channel including parameters which set said new data rate.

14. The process of claim 12 wherein step (9) comprises increasing the data rate by selecting the next burst profile up from the currently selected burst profile for the logical channel being processed, said next burst profile up being selected from a table of burst profiles arranged in order of ascending data rates and generating and transmitting downstream and upstream channel descriptor message containing the parameters for transmission on said logical channel including parameters which set said new data rate.

15. The process of claim 12 wherein step (9) comprises generating a signal which can be seen or heard by an operator suggesting a change in data rate for the channel such that said operator can manually select a new data rate and cause a new upstream channel descriptor message to be generated and sent if the operators chooses to change the data rate.

16. A process for automatically altering the data rate on a logical channel with changing noise conditions comprising:

1) resetting a total uncorrectable codeword count for said logical channel for an interval;

2) resetting a total codewords received count for said logical channel for said interval;

3) receiving a burst on said logical channel, and adding the total codewords received in the burst to the running total of codewords received for said interval;

4) receiving information regarding the number of uncorrectable codewords in said received burst and adding said number of uncorrectable codewords to a running total of uncorrectable codewords for the logical channel over said interval;

5) determining in any way whether said interval has been completed;

6) if said interval has not been completed, returning to step 3;

7) if said interval has been completed, calculating a bit error rate of said logical channel and comparing the bit error rate to one or more bit error rate thresholds, the bit error rate calculated by evaluating the expression

BER=[0.5*(T+1)*n_error]/[(k+2T)*n_total]

where T is the number of maximum correctable bytes in an Reed-Solomon (hereafter R-S) codeword, n_error is the total number of uncorrectable R-S codewords for said interval from predetermined interval usage code (IUC) burst type(s), n total is the total number of received R-S codewords from predetermined IUC burst type(s) received within said time interval, k is the number of information bytes in an R-S codeword of a specific IUC; and

(k+2t) stands for;

the R-S codeword length in bytes;

8) determining if a change in data rate throughput is required based upon the comparison of the bit error rate to the one or more byte error rate thresholds; and

9) if a change in data rate throughput is required, generating a signal indicating a need for a change in the data rate of said logical channel.

17. The process of claim 16 wherein step (9) comprises reducing the data rate by selecting the next burst profile down from the currently selected burst profile for the logical channel being processed, said next burst profile down being selected from a table of burst profiles arranged in order of descending data rates and generating and transmitting downstream and upstream channel descriptor message containing the parameters for transmission on said logical channel including parameters which set said new data rate.

18. The process of claim 16 wherein step (9) comprises increasing the data rate by selecting the next burst profile up from the currently selected burst profile for the logical channel being processed, said next burst profile up being selected from a table of burst profiles arranged in order of ascending data rates and generating and transmitting downstream and upstream channel descriptor message containing the parameters for transmission on said logical channel including parameters which set said new data rate.

19. The process of claim 16 wherein step (9) comprises generating a signal which can be seen or heard by an operator suggesting a change in data rate for the channel such that said operator can manually select a new data rate and cause a new upstream channel descriptor message to be generated and sent if the operators chooses to change the data rate.

20. A process carried out in a cable modem termination system comprising the steps:

(1) powering up and measuring the signal-to-noise ratio (SNR) on a channel prior to transmitting any downstream grant messages authorizing cable modems to transmit on said channel;

(2) selecting an initial burst profile from a table of burst profiles appropriate to said channel based upon a mode of operation of said channel, said selection based upon said SNR of said channel;

(3) monitoring SNR of said channel and a rate that impulse noise is detected on said channel;

(4) comparing the amount of impulse noise erasures to one or more erasure thresholds and deciding whether said logical channel is dominated by Average White Gaussian Noise (AWGN) or impulse noise, and selecting an appropriate set of burst profiles for said channel based upon the dominant noise type;

(5) after cable modems start transmitting on said channel, determining a packet error rate for packets transmitted on said channel;

(6) comparing the packet error rate to one or more thresholds and deciding whether a change in bit rate on said channel is advisable to bring said packet error rate within predetermined limits;

(7) if a change is bit rate is indicated by step (6), automatically picking a new burst profile with an appropriately changed bit rate from said set of burst profiles selected in step 4.

21. The process of claim 20 wherein step 4 comprising comparing said amount of impulse noise erasures to an upper limit threshold to determine when a channel has become dominated by impulse noise, and comparing said amount of impulse noise erasures to a tower limit to decide when said channel has become dominated by AWGN, said upper and lower thresholds being separated by an amount sufficient to provide a hysteresis effect.

22. The process of claim 20 wherein step 6 comprising comparing said packet error rate to an upper limit threshold to determine when a channel should have its bit rate lowered, and comparing said packet error rate to a lower limit to decide when said channel should have its A rate increased, said upper and lower thresholds being separated by an amount sufficient to provide a hysteresis effect.

23. The process of claim 20 wherein step 7 further comprises automatically generating and sending an upstream channel descriptor message which includes parameters set by said new burst profile so as to cause cable modems transmitting on said channel to alter their bit rates.

24. The process of claim 20 wherein step 7 further comprises sending a message to said cable operator indicating it would be desirable to manually change said bit rate.

25. The process of claim 20 further comprising the steps:

(8) repeating steps 3 through 7 until either said packet error rate has been brought within acceptable limits or a maximum number of attempts have been made;

(9) if a maximum number of attempts have been made without success in bringing the packet error rate (PER) within limits, jumping to another logical channel by sending new bandwidth allocation map (MAP) messages to all cable modems (CMs) that have been using the logical channel for which the PER cannot be brought within limits instructing said CMs when they may transmit on the new logical channel and, if necessary, sending an upstream channel descriptor message defining the characteristics and burst profile of said new logical channel.