Assigning time stamps to received data packets

Patent metadata

Field	Value
Publication number	US-9882799-B2
Application number	US-201514736528-A
Country	US
Kind code	B2
Filing date	Jun 11, 2015
Priority date	Dec 12, 2012
Publication date	Jan 30, 2018
Grant date	Jan 30, 2018

Abstract

A method is provided for assigning a time stamp to a specific data packet which, as with a group of received data packets, is received by a cyclically operating sensor. The specific data packet especially is part of the group of received data packets. The method includes: determining the time of arrival of each data packet of the group of received data packets; assigning a time stamp to each data packet of the group of data packets based on a first reference point in time; determining the time difference between the determined time of arrival and the assigned time stamp for each data packet of the group of data packets to provide a group of time differences; determining negative time differences from among the group of time differences, for the data packets of which it has been determined that the time of arrival preceded the respective assigned time stamp; determining the negative time differences that are not to be classified as outliers; determining a second reference point in time taking into account the negative time differences that are not to be classified as outliers; and assigning a time stamp to the specific data packet based on the second reference point in time.

First claim

What is claimed is:

1. A method for assigning a time stamp to a specific data packet which, as with a group of received data packets, is received by a cyclically operating sensor, wherein the specific data packet is part of the group of received data packets, the method comprising the acts of:
determining an arrival time for each data packet of the group of received data packets;
assigning a time stamp to each data packet of the group of data packets based on a first reference instant;
determining a time difference between the determined arrival time and the assigned time stamp for each data packet of the group of data packets to provide a group of time differences;
determining negative time differences from the group of time differences for the data packets where the arrival time was determined to precede the respective assigned time stamp;
determining the negative time differences which are not to be classified as outliers;
determining a second reference instant taking into account the negative time differences which are not to be classified as outliers; and
assigning a time stamp to the specific data packet based on the second reference instant.

2. The method according to claim 1 , wherein the act of determining the negative time differences which are not to be classified as outliers comprises the following acts of:
determining a threshold value on the basis of the determined negative time differences; and
determining the negative time differences of which the magnitude is smaller than a magnitude of the threshold value.

3. The method according to claim 2 , wherein the threshold value is determined based on one of the following conditions:
a variance of the negative time differences;
a variance of the negative time differences which lie in a quantile of the time differences; and
a variance of the negative time differences of which the magnitude is smaller than a maximum value.

4. The method according to claim 3 , wherein the threshold value is determined based on a multiple of a square root of the variances used to determine the threshold value.

5. The method according to claim 4 , wherein the multiple is an odd-number.

6. The method according to claim 3 , wherein the variances of the negative time differences which lie in the quantile of the time differences lie in a 90 th or 95 th percentile.

7. The method according to claim 2 , wherein the act of determining the threshold value comprises the following acts of:
forming clusters of the negative time differences; and
determining the threshold value based on a boundary between two clusters.

8. The method according to claim 1 , wherein the act of determining the negative time differences which are not to be classified as outliers comprises the following acts of:
forming clusters of the negative time differences; and
determining one or more of the clusters to be classified as outliers.

9. The method according to claim 1 , wherein the act of determining the second reference instant taking into account the negative time differences which are not to be classified as outliers comprises the following acts of:
selecting a single time difference from the negative time differences which are not to be classified as outliers; and
setting the second reference instant to the arrival time of the data packet for which the arrival time of the selected single time difference has been determined.

10. The method according to claim 9 , wherein the act of selecting the single time difference comprises the following acts of:
selecting the negative time difference of which a magnitude is greater than the magnitudes of the remaining negative time differences which are not to be classified as outliers; or
selecting the negative time difference which forms a boundary element of a quantile of the negative time differences which are not to be classified as outliers.

11. The method according to claim 10 , wherein the boundary element of the quantile is a 90 th percentile.

12. The method according to claim 9 , wherein the act of determining the second reference instant taking into account the negative time differences which are not to be classified as outliers further takes into account one or more of the following criteria:
a total number of data packets received by the sensor since the start-up of a system performing the method;
a transit time since the start-up of a system performing the method;
a number of determinations of a reference instant since the start-up of a system performing the method; and
a size of the magnitudes or a statistic measure of the size of the magnitudes of all determined negative time differences since the start-up of a system performing the method.

13. The method according to claim 1 , wherein the act of determining the second reference instant taking into account the negative time differences which are not to be classified as outliers comprises the following acts of:
determining a descriptive value, for the negative time differences or a subgroup of the negative time differences which are not to be classified as outliers; and
setting the second reference instant to the descriptive value.

14. The method according to claim 13 , wherein the descriptive value is a statistical measure.

15. The method according to claim 13 , wherein the descriptive value is one of the following values:
an average of the negative time differences, or a subgroup thereof, which are not to be classified as outliers; or
a median of the negative time differences, or a subgroup thereof, which are not to be classified as outliers.

16. The method according to claim 15 , wherein the subgroup of the negative time differences is a subgroup quantile of the negative time differences.

17. The method according to claim 1 , wherein the group of data packets comprises a minimum number of data packets and/or wherein the specific data packet is included in the group of data packets or is received later as the last-received data packet of the group.

18. A motor vehicle, comprising:
an electronic controller including
a memory storing a program having program code segments; and
an associated processor for executing the program code segments to: determine an arrival time for each data packet of the group of received data packets;

assign a time stamp to each data packet of the group of data packets based on a first reference instant;
determine a time difference between the determined arrival time and the assigned time stamp for each data packet of the group of data packets to provide a group of time differences;
determine negative time differences from the group of time differences for the data packets where the arrival time was determined to precede the respective assigned time stamp;
determine the negative time differences which are not to be classified as outliers;
determine a second reference instant taking into account the negative time differences which are not to be classified as outliers; and
assign a time stamp to the specific data packet based on the second reference instant.

19. A computer product, comprising a non-transitory computer readable medium having stored thereon program code segments that:
determine an arrival time for each data packet of the group of received data packets;
assign a time stamp to each data packet of the group of data packets based on a first reference instant;
determine a time difference between the determined arrival time and the assigned time stamp for each data packet of the group of data packets to provide a group of time differ

Assignees

• Bayerische Motoren Werke Ag

Inventors

• Kloeden Horst
• Klanner Felix
• Bartsch Armin
• Rauch Andreas

Classifications

• H04J3/0658 Primary

  Clock or time synchronisation among packet nodes · CPC title

• H04L43/067

  using time frame reporting · CPC title

• H04L43/106 Primary

  using time related information in packets, e.g. by adding timestamps · CPC title