System and method for synchronizing digital bits in a data stream

• US 8,891,714 B2
• Filed: 01/12/2012
• Issued: 11/18/2014
• Est. Priority Date: 06/30/2006
• Status: Expired due to Fees

• Current Assignee: CSR Technology Incorporated (Qualcomm, Inc.)
• Original Assignee: CSR Technology Incorporated (Qualcomm, Inc.)
• Inventors: Raman, Sundar, Brown, James M.
• Primary Examiner(s): PANWALKAR, VINEETA S
• Application Number:
  US13/349,102
• Publication Number:
  US 20120236973A1
• Time in Patent Office: 1,041 Days
• Field of Search: 375/130, 375/140, 375/142, 375/145, 375/147, 375/149, 375/150, 375/316, 375/340, 375/343, 375322-326, 375/354, 375/355, 375/359, 375/360, 375/362, 375/365, 375/367, 708/1, 708/3, 708/5, 708/100, 708/200, 708/207, 708/422, 708/800, 708/801, 708/813, 704/200, 704/201, 704/211, 704/216, 704/218, 704/231, 704/236, 704/237, 704/258, 704/263
• US Class Current: 375/355
• CPC Class Codes: G01S 19/24 Acquisition or tracking or demodulation of signals transmitted by the system synchronisation aspects of direct sequence spread spectrum modulation H04B1/7073
• CPC Class Codes: G01S 19/30 code related G01S19/246 takes precedence
• CPC Class Codes: H04B 1/7075 with code phase acquisition
• CPC Class Codes: H04B 2201/70715 with application-specific features

Abstract

A system for synchronizing a receiver of a bit stream to the bit stream include a correlator to remove the PN code modulation and to generate a stream of time sequence values (samples) from the received bits. Multiple accumulators are included, each accumulator corresponding to a different offset from a first time period. The accumulators add a number of values equal to a number of samples in a bit period. Multiple magnitude calculators receive sums from the corresponding accumulators and calculates respective magnitudes. Multiple non-coherent summers are then used to add the magnitudes for each offset in each bit period for all of the received bits. The outputs of the non-coherent summers are analyzed to find the highest value, to identify the location of the bit transition in the bit period.

Claims

1. A method for identifying a main peak value in an array of peak values where a plot of the peak values against offsets corresponding to each peak value shows a peak value having a largest magnitude between peak values on either side of the largest peak value and substantially linear patterns formed by the peak values right and left of the largest peak value, the method comprising:

generating, by a processor, a right line by a first order curve fit using at least one peak value to the right of the largest peak value and the largest peak value, the right line having a right slope and a right intercept;

generating, by the processor, a left line by a first order curve fit using at least one peak value to the left of the largest peak value and the largest peak value, the left line having a left slope and a left intercept;

identifying, by the processor, a predicted peak value by identifying an intersection point on a timeline formed by the offsets at which the right line intersects the left line; and

identifying, by the processor, the largest peak value as the main peak value if the predicted peak value is within a predefined interval around the largest peak value.

2. The method of claim 1 where the array of peak values is generated as an ordered set of peak values, the first element in the array containing the largest peak value, the second element in the array containing the second largest peak value, the third element in the array containing the third largest peak value, and so on for all of the elements in the array, the method further comprising:

testing the peak values by;

determining if the second and third largest peak values are valid peak values by checking if the second and third largest peak values correspond to positions one offset to the right and left of the offset corresponding to the largest peak value;

determining if the fourth and fifth largest peak values are valid peak values by checking if the fourth and fifth largest peak values correspond to two offsets to the right and left of the offset corresponding to the largest peak value;

determining if the sixth and seventh largest peak values are valid peak values by checking if the sixth and seventh highest peak values correspond to three offsets to the right and left of the offset corresponding to the main peak;

if only the second and third peak values are valid peak values, selecting a two-point analyzer method that uses one peak value to the right of the largest peak value and the largest peak value to define a right slope of the right line and one peak value to the left of the main peak value and the largest peak value to define a left slope of a left line;

if only the second, third, fourth and fifth peak values are valid peak values, selecting a four-point analyzer method that uses two peak values to the right of the largest peak value to generate the right line and two peak values to the left of the largest peak value to generate the left line; and

if the second, third, fourth, fifth, sixth and seventh peak values are valid peak values, selecting a six-point analyzer method that uses three peak values to the right of the largest peak value to generate the right line and three peak values to the left of the largest peak value to generate the left line.

3. The method of claim 2 wherein, if the two point analyzer method is used, performing the steps of:

calculating a slope of the right line;

calculating a slope of the left line;

calculating a probability of the largest peak value being the main peak value using;

probability of main peak=5*(slope of left line+slope of light line)/largest peak; and

comparing the probability of the largest peak value with a minimum probability.

4. A method for identifying a main peak value in an array of peak values where a plot of the peak values against offsets corresponding to each peak value shows a peak value having a largest magnitude between peak values on either side of the largest peak value and substantially linear patterns formed by the peak values right and left of the largest peak value, the method comprising:

generating, by a processor, a second array of peak values by;

calculating, by the processor, respective differences between the largest peak value and each other peak value in the array of peak values and combining the respective difference to generate an average difference value, Diff_AVE, between successive peak values in the array of peak values;

generating, by the processor, a set of ideal right peak values by calculating an ideal right peak value for each offset n from the offset of the largest peak value, where n is an integer ranging from 1 to N, N is a largest integer displacement of the right peak values being processed and N=2, 4 or 6, using;

p_right_—

n=largest peak value−

n*Diff_AVE;

generating, by the processor, a set of ideal left peak values by calculating an ideal left peak value for each offset n from the offset of the largest peak value using;

p_left_—

n=largest peak value−

n*Diff_AVE;

calculating, by the processor, a sum of the differences, Diff_sum, using;

Diff_Sum=(right_—

1−

p_right_—

1)+(right_—

2−

p_right_—

2)+ . . . +(right_—

n−

p_right_—

n)+(left_—

1−

p_left_—

1)+(left_—

2−

p_left_—

2)+ . . . +(left_—

n−

p_left_—

n),

where right_n is the peak value n offsets to the right of the largest peak value and left_n is the peak value n offsets to the left of the largest peak value;

calculating, by the processor, a sum of squares, SUM_SQ, using;

SUM_SQ=(right_—

1−

p_right_—

1)^2+(right_—

2−

p_right_—

2)^2+ . . . +(right_—

n−

p_right_—

n)^2+(left_—

1−

p_left_—

1)^2+(left_—

2−

p_left_—

2)^2+ . . . +(left_—

n−

p_left_—

n)^2

calculating, by the processor, a standard deviation, STD; and

indicating, by the processor, that the largest peak value is the main peak value if Diff_AVE/STD is not less than a predetermined metric threshold.

5. A method for synchronizing a receiver of a bit stream to the bit stream by identifying a location of a transition in the bit stream, the method comprising:

correlating a time sequence stream of values from a finite sample of received bits with a known sequence stream of values to generate a correlated sequence of values,

accumulating a plurality of offset accumulation values, each corresponding to an offset in a series of time intervals within the correlated sequence, starting with a first time period, each accumulation value accumulating a number of values equal to a number of samples in a bit period, the number of values accumulated by each accumulator of a plurality of offset accumulators being a set of values in the correlated sequence starting with a value at the offset corresponding to the accumulator;

calculating respective magnitude values for each of the accumulation values;

calculating a plurality of non-coherent summation values each corresponding to one of the offsets, the plurality of non-coherent summation values being respective sums of the magnitude values received for each offset in each bit period for all of the N bit periods, where N is a positive integer less than or equal to 40; and

analyzing the total sum of magnitudes for each offset for the highest sum of magnitudes, such that the offset corresponding to the non-coherent summer with the highest sum of magnitudes represents the location of the bit transition in the bit period.

6. The method of claim 5 wherein before analyzing, performing the steps of:

calculating a probability that the largest peak value is the main peak value;

when the calculated probability is greater than a minimum probability performing the analyzing; and

when the calculated probability is less than the minimum probability, halting the synchronizing, accumulating further offset accumulation values and restarting the synchronizing using the accumulated plurality of further offset accumulated values and the accumulated further offset accumulation values.

7. The method of claim 5 further comprising analyzing the sums of magnitudes of the non-coherent summation values to determine whether the location of the bit transition was correctly determined.

8. The method of claim 7 where the analyzing of the sums of magnitudes of the non-coherent summation values includes:

generating an array of peak values from at least a set of the sums of magnitudes where a plot of the peak values against the offsets corresponding to each peak value shows the largest magnitude forming a main peak value between peak values on either side of the main peak value and substantially linear patterns formed by the peak values right and left of the main peak value;

generating a right line by a first order curve fit using at least two peak values to the right of the main peak value, the right line having a right slope and a right intercept;

generating a left line by a first order curve fit using at least two peak values to the left of the main peak value, the left line having a left slope and a left intercept;

identifying a predicted peak value by identifying an intersection point on a timeline formed by the offsets at which the right line intersects the left line, and

verifying the location of the main peak value if the intersection point is within a predetermined offset time limit of the main peak value.

9. The method of claim 8 further comprising:

generating an array of peak values from at least a set of the sums of magnitudes where a plot of the peak values against the offsets corresponding to each peak shows the largest magnitude forming a main peak value between peak values on either side of the main peak value and substantially linear patterns formed by the peak values right and left of the main peak value;

generating a second array of peak values by;

calculating respective differences between the largest peak value and each other peak value in the array of peak values and combining the respective difference to generate an average difference value, Diff_AVE, between successive peak values in the array of peak values;

generating a set of ideal right peak values by calculating an ideal right peak value for each offset n from the offset of the largest peak value, where n is an integer ranging from 1 to N, N is a largest integer displacement of the right peak values being processed and N=2, 4 or 6, using;

p_right_—

n=main peak value−

n*Diff_AVE,

where n=an integer number of offsets from the offset of the main peak value;

generating a set of ideal left peaks by calculating an ideal left peak value for each offset n from the offset of the largest peak value using;

p_left_—

n=main peak value−

n*Diff_AVE,

where n=an integer number of offsets from the offset of the main peak value;

calculating a sum of the differences, Diff_sum, using;

Diff_Sum=(right_—

1−

p_right_—

1)+(right_—

2−

p_right_—

2)+ . . . +(right_—

n−

p_right_—

n)+(left_—

1−

p_left_—

1)+(left_—

2−

p_left_—

2)+ . . . +(left_—

n−

p_left_—

n),

where right_n is the peak value n offsets to the right of the largest peak value and left_n is the peak value n offsets to the left of the largest peak value;

calculating a sum of squares, SUM_SQ, using;

SUM_SQ=(right_—

1−

p_right_—

1)^2+(right_—

2−

p_right_—

2)^2+ . . . +(right_—

n−

p_right_—

n)^2+(left_—

1−

p_left_—

1)^2+(left_—

2−

p_left_—

2)^2+ . . . +(left_—

n−

p_left_—

n)^2

calculating a standard deviation, STD; and

verifying the location of the main peak value if Diff_AVE/STD is not less than a predetermined metric threshold.

10. A method for synchronizing a receiver of a bit stream to the bit stream by identifying a bit transition in a bit period, the method comprising:

generating a correlated sequence of values by correlating a time sequence stream of values from a finite sample of received bits with a known sequence stream of values;

accumulating a plurality of offset accumulation values, each corresponding to an offset in a series of time intervals within the correlated sequence, starting with a first time period, by accumulating a number of values equal to a number of samples in a bit period, the number of values accumulated to form each accumulation value being a set of values in the correlated sequence starting with the value at the offset corresponding to the offset accumulation value;

calculating a plurality of power magnitude values from respective ones of the plurality of offset accumulation values;

non-coherently summing the plurality of power magnitude values to provide respective sum-of-magnitude values, each sum-of-magnitude value corresponding to a respective offset; and

analyzing the plurality of sum-of-magnitude values for a highest sum of magnitude value, such that the offset corresponding to the highest sum of magnitude value represents the location of the bit transition in the bit period.

11. The system of claim 10 further comprising analyzing the sums-of-magnitude values in the non-coherent summers to determine whether the location of the bit transition was correctly determined.

12. The method of claim 11 where the analyzing includes:

generating a list of peak values from at least a set of the sum-of-magnitude values where a plot of the sum-of-magnitude values against the offsets corresponding to each peak shows the largest sum-of-magnitude value forming a main peak between peaks on either side of the main peak and substantially linear patterns formed by the peaks to the right and left of the main peak in the list;

generating a right curve by curve-fitting at least two peaks to the right of the main peak;

generating a left curve by curve-fitting at least two peaks to the left of the main peak;

identifying a predicted peak by identifying an intersection point on a timeline formed by the offsets at which the right line intersects the left line; and

indicating that the bit transition location passes if the intersection point is within a peak offset time limit of the bit transition location.

13. The method of claim 10 wherein before performing the analyzing, performing the steps of:

calculating a probability that the largest peak value is the main peak value;

when the calculated probability is greater than or equal to a minimum probability, performing the analyzing; and

when the calculated probability is less than the minimum probability, halting the synchronizing, accumulating further offset accumulation values and restarting the synchronizing using the accumulated plurality of further offset accumulated values and the accumulated further offset accumulation values.