Method of smart saving high-density data and memory device

• US 10,162,551 B2
• Filed: 06/28/2016
• Issued: 12/25/2018
• Est. Priority Date: 08/27/2015
• Status: Active Grant

• Current Assignee: STMicroelectronics SRL (STMicroelectronics NV)
• Original Assignee: STMicroelectronics SRL (STMicroelectronics NV)
• Inventors: Leo, Marco, Rosingana, Paolo, Castellano, Marco, Locardi, Alessandro Giuliano
• Primary Examiner(s): Yi, David
• Assistant Examiner(s): Paperno, Nicholas A.
• Application Number:
  US15/195,352
• Publication Number:
  US 20170060467A1
• Time in Patent Office: 910 Days
• CPC Class Codes: G01D 9/00 Recording measured values
• CPC Class Codes: G06F 3/0604 Improving or facilitating administration, e.g. storage management
• CPC Class Codes: G06F 3/0638 Organizing or formatting or addressing of data
• CPC Class Codes: G06F 3/0673 Single storage device
• CPC Class Codes: H03M 7/3091 Data deduplication

Abstract

A signal interface has a compression unit and a data memory. The compression unit is configured to input an input datum from signal data generated by at least one sensor and further configured to identify the presence or absence of at least one repetition condition in the input datum. If the presence of the at least one repetition condition of the input datum is identified, the compression unit encodes the input datum in a compressed way to generate a compressed datum and saves the compressed datum in the data memory. If the presence of the at least one repetition condition of the input datum is not identified, the compression unit saves the uncompressed input datum in the data memory.

Claims

1. A method to manage signal data, comprising:

acquiring an input datum from signal data generated by a plurality of different sensors, wherein the different sensors further include a plurality of different types of sensors, wherein each different sensor has its own data rate and based on a set of different data rates of the plurality of different sensors, realizing a least common multiplier of the set of different of data rates, wherein the input datum represents sensor data readings contemporaneously collected from the plurality of different types of sensors, wherein each input datum is formed as a multi-bit block of combined sensor data generated by the plurality of different types of sensors, and wherein each input datum is further processed as a single piece of data;

advancing a counter each time a new input datum is acquired;

if the counter has reached the least common multiplier;

saving the new input datum in the memory; and

resetting the counter;

if the counter has not reached the least common multiplier;

identifying a presence or an absence of at least one repetition condition of the new input datum;

if the presence of the at least one repetition condition of the new input datum is identified;

encoding in a compressed way the new input datum to generate a compressed datum; and

saving the compressed datum in a memory; and

if the absence of the at least one repetition condition of the new input datum is identified;

saving the new input datum in the memory.

2. The method according to claim 1, wherein the at least one repetition condition includes one or more of a repetition frequency, a difference from a preset value, and a difference from a previous value.

3. The method according to claim 1, wherein the act of encoding in the compressed way comprises:

deriving a difference between the input datum and a reference datum.

4. The method according to claim 3, wherein the reference datum is a previous input datum.

5. The method according to claim 4, wherein the act of deriving the input datum comprises:

saving the input datum as an updated reference datum.

6. The method according to claim 4, wherein the act of saving the compressed datum in the memory comprises:

saving a previous input datum as the reference datum.

7. The method according to claim 3, wherein the reference datum comprises a preset value.

8. The method according to claim 3, wherein identifying the presence or absence of at least one repetition condition of the input datum comprises:

calculating a difference between the input datum and the reference value;

comparing the difference with a threshold; and

determining whether or not the difference is lower than the threshold.

9. The method according to claim 3, wherein the input datum is a multidimensional datum having a number of dimensions, and wherein identifying the presence or absence of at least one repetition condition of the input datum comprises:

calculating a difference between each dimension of the input datum and a respective reference value;

comparing each difference with a respective threshold; and

determining whether or not any of the differences is lower than its respective threshold.

10. The method according to claim 9, wherein saving the compressed datum in the memory comprises:

saving the differences if all of the calculated differences are lower than their respective thresholds.

11. The method according to claim 1, further comprising:

saving in the memory a compression tag indicating whether the compressed datum is saved in the memory or the input datum is saved in the memory.

12. The method according to claim 11, wherein the compression tag is arranged to take a plurality of different values including a first value of the compression tag that represents saving the compressed datum in the memory, and a second value of the compression tag that represents saving the input datum in the memory.

13. The method according to claim 1, comprising:

reading a first datum from the memory;

verifying whether the first datum read from the memory is a first compressed datum or a first input datum; and

interpreting the first datum based on the verifying.

14. The method according to claim 13, wherein verifying whether the first datum read from the memory is the first compressed datum or the first input datum comprises:

reading a compression tag corresponding to the first datum;

verifying whether or not the compression tag indicates that the first datum is compressed; and

interpreting the first datum according to the compression tag verification.

15. The method according to claim 1, wherein the input datum is an angular-velocity value, an acceleration value, a magnetic-field value, a pressure value, a temperature value, a humidity value, a brightness value, a gravity value, a quaternion, a position value, a motion signal, or a physical quantity.

16. The method of claim 1, wherein identifying a presence or an absence of at least one repetition condition of the input datum comprises:

performing a selective compression based on a speed of change of the signal data.

17. The method of claim 16, wherein performing a selective compression comprises:

calculating a difference between the input datum and a reference value;

comparing the difference with a first threshold;

determining whether or not the difference is lower than the first threshold;

compressing the datum with a first compression level if the difference is lower than the first threshold;

comparing the difference with a second threshold if the difference is higher than the first threshold, the second threshold being higher than the first threshold;

determining whether or not the difference is lower than the second threshold;

compressing the datum with a second compression level, the second compression level being lower than the first compression level, if the difference is lower than the second threshold.

18. The method of claim 16, wherein performing a selective compression comprises:

loading a plurality of input data;

calculating differences between the loaded input data and a reference value;

comparing the differences with a first threshold;

determining whether or not the differences are lower than the first threshold;

compressing the loaded data with a first compression level if all differences are lower than the first threshold;

comparing the difference with a second threshold if any difference is higher than the first threshold, the second threshold being higher than the first threshold;

determining whether or not the differences of at least two loaded data are lower than the second threshold;

compressing the at least two loaded data with a second compression level, the second compression level being lower than the first compression level, if the differences of the at least two loaded data are lower than the second threshold.

34. The method according to claim 1, wherein the plurality of different sensors include at least one accelerometer sensor, at least one gyroscopic sensor, and at least one magnetometer sensor.

35. The method according to claim 1, wherein the plurality of different sensors include at least one pressure sensor, at least one temperature sensor, and at least one humidity sensor.

19. A signal interface, comprising:

a data memory; and

a compression unit, the compression unit configured to;

isolate an input datum from signal data generated by a plurality of different sensors, wherein the different sensors further include a plurality of different types of sensors, wherein each different sensor has its own data rate and based on a set of different data rates of the plurality of different sensors, realizing a least common multiplier of the set of different of data rates, wherein each input datum represents sensor data readings contemporaneously collected from the plurality of different types of sensors, wherein each input datum is formed as a multi-bit block of combined sensor data generated by the plurality of different types of sensors, and wherein each input datum is further processed as a single piece of data;

advance a counter each a new input datum is acquired;

compare the counter to the least common multiplier;

when the comparison indicates the counter has reached the least common multiplier;

save the new input datum in the data memory; and

reset the counter;

when the comparison indicates the counter has not reached the least common multiplier;

verify whether or not the new input datum includes a repetition condition;

when the new input datum includes the repetition condition, encode in a compressed way the new input datum to transform the new input datum into a compressed datum; and

save the new input datum in the data memory in absence of the repetition condition or save the compressed datum in the data memory after verification of the repetition condition.

20. The signal interface according to claim 19, wherein the repetition condition includes a repetition frequency, a difference from a preset value, or a difference from a previous value.

21. The signal interface according to claim 19, wherein the compression unit configured to encode the input datum in the compressed way includes the compression unit configured to:

derive a difference between the input datum and a reference datum wherein the reference datum is a previous input datum or a preset value.

22. The signal interface according to claim 21, wherein the input datum is a multidimensional datum having a number of dimensions, and wherein the compression unit configured to verify whether or not the input datum includes the repetition condition includes the compression unit configured to:

calculate a difference between each dimension of the input datum and a respective reference value;

compare each difference with a respective threshold; and

determine whether or not any of the differences is lower than its respective threshold.

23. The signal interface according to claim 21 wherein the compression unit is configured to:

save in the data memory a compression tag, wherein the compression tag is arranged to take a plurality of different values including a first value of the compression tag that represents the compressed datum is saved in the data memory, and a second value of the compression tag that represents the input datum is saved in the data memory.

24. A mobile device, comprising:

a plurality of different sensors configured to supply output signal data, wherein the different sensors further include a plurality of different types of sensors;

a signal interface configured to receive the output signal data;

a data memory associated with the signal interface; and

an application processor configured to retrieve historical data representing at least some of the output signal data supplied by the plurality of different types of sensors, wherein each different sensor has its own data rate and based on a set of different data rates of the plurality of different sensors, realizing a least common multiplier of the set of different of data rates, wherein the signal interface is configured to iteratively process batches of the output signal data, wherein each batch of output signal data represents sensor data readings contemporaneously collected from the plurality of different types of sensors, wherein each batch of output signal data is formed as a multi-bit block of combined sensor data generated by the plurality of different types of sensors, and wherein each batch of output signal data is further processed as a single piece of data, the iterative processing including;

advancing a counter during each iteration;

if the counter has reached the least common multiplier;

storing an uncompressed version of the output signal data as historical data; and

resetting the counter;

if the counter has not reached the least common multiplier;

verifying whether a current batch of output signal data is repetitive of a previous batch of output signal data; and

based on verifying repetitive output signal data, storing in the data memory a compressed version of the output signal data as historical data, or based on not verifying repetitive output signal data, storing an uncompressed version of the output signal data as historical data.

25. The mobile device according to claim 24, wherein verifying whether the current batch of output signal data is repetitive of the previous batch of output signal data includes verifying a repetition frequency, verifying a difference from a preset value, or verifying a difference from a previous value.

26. The mobile device according to claim 24, wherein the plurality of different sensors includes at least one of an inertial sensor, a position sensor, a sensor of derived signals, a motion detector, and an environmental sensor.

27. The mobile device according to claim 24, wherein the output signal data includes at least one of an angular-velocity value, an acceleration value, a magnetic-field value, a pressure value, a temperature value, a humidity value, a brightness value, a gravity value, a quaternion, a position value, a motion signal, or a physical quantity.

28. The mobile device according to claim 24, wherein the mobile device includes features of at least one of a cellular-based device, a tablet, a wearable device, an inertial-navigational device, an automotive system, and a robotic system.

29. The mobile device according to claim 24, wherein the signal interface includes:

a compression unit, the compression unit configured to generate the compressed version of the output signal data, the compressed version of the output signal data derived from a difference between a portion of the current batch of output signal data and a corresponding portion of the historical data.

30. The mobile device according to claim 29, wherein the corresponding portion of the historical data includes at least one preset value.

31. The mobile device according to claim 24, wherein the output signal data is multidimensional signal data having a number of dimensions, and wherein verifying whether the current batch of output signal data is repetitive of the previous batch of output signal data includes:

calculating a difference between each dimension of a portion of the current batch of output signal data and a respective dimension of the corresponding portion of the historical data; and

comparing each difference with a respective threshold.

32. The mobile device according to claim 24, wherein verifying whether the current batch of output signal data is repetitive of the previous batch of output signal data includes:

storing in the data memory a compression tag, the compression tag arranged to take a first value if the compressed version of the output signal data is saved in the data memory as historical data, and arranged to take a second value if the uncompressed version of the output signal data is saved in the memory as historical data.

33. The mobile device according to claim 24, wherein the compressed version of the output signal data occupies less than half as much storage space in the data memory as a corresponding uncompressed version of the output signal data would occupy.

36. A method to manage signal data, comprising:

contemporaneously collecting data from a plurality of different sensors, the plurality of different sensors including a plurality of different types of sensors, wherein each different sensor has its own data rate and based on a set of different data rates of the plurality of different sensors, realizing a least common multiplier of the set of different of data rates;

forming an input datum from at least some of the collected data, the input datum formed as a multi-bit block of combined sensor data generated by the plurality of different types of sensors;

acquiring and processing the input datum as a single piece of data;

advancing a counter each time an input datum is acquired;

if the counter has reached the least common multiplier;

saving the input datum in the memory; and

resetting the counter;

if the counter has not reached the least common multiplier;

identifying a presence or an absence of at least one repetition condition of the input datum;

if the presence of the at least one repetition condition of the input datum is identified;

encoding in a compressed way the input datum to generate a compressed datum; and

saving the compressed datum in a memory; and

if the absence of the at least one repetition condition of the input datum is identified;

saving the input datum in the memory.