KR20100061894A - Low-power sensing method utilizing baseband processor - Google Patents

Abstract

The present invention utilizes a baseband processor, which is essentially installed in a portable communication device, to acquire movement data of X, Y, and Z axes 24 hours from an accelerometer that senses device movement. This paper describes low-power sensing technology that saves more than 25% of power consumption. In the existing portable communication devices, the use of a separate micom, rather than a baseband processor, is used to manage mounted sensors, resulting in a problem of increased product cost and insufficient PCB space. By using the low power sensing technology according to the embodiment of the present invention, the pedometer function is driven at low power by using the sleep mode of the baseband processor without using a microcomputer for motion data acquired from the acceleration sensor.

Description

Low-Power Sensing Method Using Baseband Processor

The present invention provides a low power sensing technology utilizing a baseband processor. In other words, it is basically a technology necessary for a kind of portable healthcare system.

  Currently, in Korea, various information is provided to users through the acceleration sensor installed in portable communication devices such as mobile phones. The most commonly used fields are menu movement, games, and sports well-being (pedometer, heart rate monitor, timer).

  Overseas, the Nike sneakers + Apple iPod `` Sports Kit '' jointly manufactured by Nike and Apple are the representative solutions for the pedometer using sensors. A pressure sensor with a wireless transmitter is inserted under Nike's sneaker insole to send out step information when the user walks or runs. The user collects the sensor's information with Apple's iPod, which must be equipped with a separate receiving accessory to receive radio signals. The information collected by the iPod is displayed on the iPod LCD screen or delivered to the user in the sound of announcements. The development of health care solution using the sensors installed in this way is becoming more and more active at home and abroad.

The present invention utilizes the sleep mode of the baseband processor in a mobile machine to drive the sensing function at low power without a separate microcomputer. The present invention provides a basic framework of a mobile SW solution that can be utilized as a health care / exercise helper by measuring the number of steps accurately while reducing current consumption by using a baseband processor.

   The accelerometer applied to mobile phones in the above technology is a three-axis accelerometer called an accelerometer made with MEMS technology. The measurable range has a measuring range of -3 g to +3 g based on the gravitational acceleration value (g). The same kind of three-axis acceleration sensor is applied to the present invention.

    The acceleration sensor measures the change in acceleration in real time, and when requested by the processor, converts the measured acceleration value into a value corresponding to an integer value and delivers it to the processor. The processing of the data is left to the processor and various analysis algorithms process the change in acceleration into motion data. The mobile phone's baseband processor spends most of its time in sleep mode and low power mode, and then periodically goes to run mode to check only incoming calls from the base station and then go back to sleep.

    The current pedometer-related solution is equipped with a separate 16-bit microcomputer in addition to the mobile phone baseband processor so that the microcomputer collects and analyzes the acceleration sensor data. The disadvantages of this method are that the power consumption by the microcomputer alone is not small, and the cost of the product due to the addition of the microcomputer is increased and there is a problem of insufficient PCB space. In particular, the space shortage problem is regarded as very negative to face up against the design trend to make mobile phones smaller and thinner. In addition, since the microcomputer low computing power cannot run complex advanced algorithms, the information of collected accelerometers is not properly processed into useful advanced information.

    In order to solve this problem, the present invention allows a mobile phone baseband processor to collect measurement data of the acceleration sensor, thereby requiring the baseband processor from the acceleration sensor to sleep in a very slow sleep clock in sleep mode. Make information gather This is an advanced technology that requires specialized knowledge and development experience for mobile phone processors and operating systems.

In the present technology, data derived from the acceleration sensor can be processed in a sample unit (Real-Time Process), but since it is difficult to process it directly in the low power mode, a batch process is also performed in parallel. That is, the number of samples of data to be processed at one time does not need to be constant, and the collected data may be transmitted until processing is required. In addition, if there is no change in the number of steps for a long time, lowering the sampling rate from the chip can reduce the power consumption of the entire system. In other words, lowering the sampling rate does not significantly reduce the power consumption of the acceleration sensor, but the baseband processor has a longer time in sleep mode, which is a great help in lowering system power consumption. However, once you're processing your data for gait analysis, don't change Sampling-Rate. To date, after testing at 33.33Hz, various setting values have been determined. If this is changed, more test for adjusting the setting value is needed.

 In general, baseband processors (eg, ARM) used in mobile devices support low power mode (Sleep mode), and it is possible to read data periodically from the acceleration sensor. Therefore, the baseband processor keeps storing the data while in the low power mode, and after exiting the low power mode after a predetermined time, the collected data is processed by the BATCH method to calculate the number of steps.

In addition, while monitoring the acceleration sensor installed in the mobile phone for 24 hours, it does not reduce the standby time of the mobile phone much.This technology enables the mobile phone to detect the 24-hour motion and transmit the motion information to the server wirelessly in real time. As it becomes possible, it is possible to grasp the user's movement at a specific point in time or in a pattern for a certain period of time, and thus it can be used as a base technology for implementing motion-based or LBS (Location Based Service) based applications and services.

The technical problem to be achieved by the present invention is that the acceleration sensor control using the microcomputer used in the current portable communication device causes a lot of power consumption, and due to the limitations of the microcomputer hardware, it is difficult to provide an advanced service, without a separate microcomputer It uses only a band processor to reduce power consumption, while providing enhanced measurable information such as steps through improved acceleration sensor control. The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned will be clearly understood from the following description.

    In order to achieve the above object, the low-power sensing technology using the baseband processor according to an embodiment of the present invention is to enable the baseband processor to drive the sensing function at low power without a separate microcomputer in a portable communication device. In the present invention, the low power mode (Sleep mode) of a baseband processor (eg, ARM), which is generally used in a portable communication device, is supported. In this case, it is possible to periodically read data from the acceleration sensor. Therefore, the baseband processor keeps storing the data while in the low power mode, and after exiting the low power mode after a predetermined time, the collected data is processed by the BATCH method to calculate the number of steps.

According to the low power sensing technology using the baseband of the present invention as described above has the following effects.

First, it is possible to reduce the power consumption of a mobile phone by controlling it through a baseband processor without using a microcomputer.

Second, it can solve the problem of product cost increase and lack of PCB space due to the addition of micom.

Third, by using the surplus power through the processing of the acceleration sensor information through the baseband processor to extract the user's exercise information and transmit it to the server can provide a variety of health care services.

  In order to achieve the above object, the low-power sensing technology using the baseband processor according to an embodiment of the present invention is to enable the baseband processor to drive the pedometer function at low power without a separate microcomputer. The present invention provides a framework for implementing a mobile SW solution that can be utilized as a health care / exercise helper by measuring the number of steps accurately while reducing current consumption by using a baseband processor.

    Specific details of other embodiments are included in the detailed description and drawings.

    Advantages and features of the present invention and methods for achieving them will be apparent with reference to the embodiments described below in detail with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but can be implemented in various different forms, and only the embodiments make the disclosure of the present invention complete, and the general knowledge in the art to which the present invention belongs. It is provided to fully inform the person having the scope of the invention, which is defined only by the scope of the claims.

   Hereinafter, the present invention will be described with reference to the drawings for explaining a low power sensing device and a solution using a baseband processor according to embodiments of the present invention.

   First, the basic H / W configuration of the low power shunting device will be described. 1 is a basic H / W configuration diagram of the solution. According to Fig. 1, the MCU side can turn on / off the function of the accelerometer chip through the EN terminal, and the only other connection is to get power and acceleration data. You can choose one of SPI / I2C / ADC according to the target to be implemented and the model of acceleration sensor to be used. Mobile devices such as cell phones can receive / process data directly from baseband processors (MSM, Agere, etc.). In the case of Micom, it was necessary to manage power in the past, but in the present invention using a baseband processor, there is no need for cost reduction and power saving.

   2 is a breakdown of the overall configuration by layer. The part marked in red and the application part are provided by the solution and calculate the number of steps by collecting data directly from the hardware such as the acceleration sensor using the baseband processor installed in the mobile phone. In general, the baseband processor of the mobile phone is divided into Idle, Sleep, and Run modes. Most of the time, the device operates in Sleep mode and wakes up at about 1 second intervals for UI update. In the present invention, the baseband processor operates at a low speed of 32Khz even in the sleep mode, and continuously collects data even in the sleep mode and stores it in the RAM. The stored data contains the X, Y and Z axes of the acceleration sensor.

   3 shows the correlation between data collection and power. In Idle mode or Run mode (high speed over 150MHz), the data is collected and the pedometer steps are calculated for the stored data when in sleep mode.

   4 illustrates a correlation between data collection and power for a baseband processor without a sleep mode function. Some baseband processors do not provide Sleep mode, and there is a Deep-Sleep mode (also called Stop mode, also known as Power-Off mode), and in Deel-Sleep mode, the CPU does not communicate with the data from the acceleration sensor. It is impossible to acquire. Therefore, the processor wakes up the processor 20-30 times per second (minimum 17 times), collects data from the acceleration sensor and stores it in RAM, and immediately puts the processor into Deel-Sleep mode to minimize power consumption. In the Idle or Run mode of these processors, the pedometer steps are calculated on the data already stored along with the data collection operation.

   5 is a general flowchart of the S / W system of the present invention. The orange and turquoise areas are provided by this solution, and the turquoise areas are required for porting, depending on the model of the baseband processor being implemented. In the present invention, a real-time process is possible in units of samples coming out of the acceleration chip. However, since the process is difficult in a low power mode, a batch process is also performed in parallel. That is, the number of samples of data to be processed at one time does not need to be constant, and the collected data may be transmitted until processing is required. If there is no change in the number of steps over a long period of time, you can reduce the system-wide power consumption by lowering the sampling rate, which gets the data from the acceleration sensor. However, once you're processing your data for gait analysis, don't change Sampling-Rate. To date, after testing at 33.33Hz, various setting values have been determined. If this is changed, more test for adjusting the setting value is needed.

   In general, most baseband processors (e.g. ARM) used in mobile devices support low power mode (Sleep mode), and it is possible to read data periodically from the acceleration sensor. Therefore, the baseband processor continues to store data while in the low power mode, and after exiting from the low power mode after a predetermined time, the accelerometer data collected therein is processed by BATCH to calculate the number of steps. In the case of a baseband processor that provides Deep-Sleep mode instead of the low-power Sleep mode, Deep-Sleep mode wakes up the processor 20 to 30 times per second (at least 17 times) to acquire data from the acceleration sensor.

   6 is a simple diagram and interface configuration example of a driver that plays a role of reading acceleration values from SPI, I2C, and ADC in a designated memory. Accelerometers offer a variety of interface methods, including ADC conversion as an analog interface or SPI / I2C as a digital interface. The left part of FIG. 6 shows a connection with a baseband processor (eg, MSM6025) via an SPI interface. The necessary PINs are connected by GPIO, and the necessary SPI protocol is already implemented in the driver part. In other words, porting is rarely necessary except for GPIO matching the actual H / W configuration and GPIO numbers.

   The software module that directs data collection from the acceleration sensor and is involved in the extraction of steps or movement patterns is the movement quantity / motion pattern extractor. The motion quantity / motion pattern extractor collects data at a specified period and processes the collected data at once. This part is divided into two types when the CPU is in low power mode and when it is not. In low-power mode, it only periodically collects data using the CPU's RTC or the OS's timer function. That is, when entering the low power mode, the main clock applied to the CPU is changed to the sleep clock, and the data is collected at regular intervals by looping using an RTC or a timer before powering down the CPU. If not, the power consumption is very low since the main clock is very low). After data has been collected as needed, or after a certain period of time, it will exit the low-power mode (return the main clock back) and process the collected data.

How to get out of low power mode depends on the implementation. For baseband processors for CDMA communication, wake-up interrupts occur periodically and wake up automatically. The motion quantity / motion pattern extractor creates a task (pedo task) for the pedometer to handle steps or motion pattern extraction when it is not in the low power mode (these tasks only operate when the power is not in the low power mode). Pedo Task collects data at regular intervals using Timer provided by the OS, and performs batch processing for gait analysis of data stored up to now (including data collected in low power mode). After that, the data is read and processed one by one while the Pedo Task is running.

   In summary, regardless of the low power mode, the data is collected at regular intervals using a loop or OS Timer. When the Pedo Task is awake, the data collected up to that point is analyzed. Through this process, the steps are actually detected based on the data collected through the motion / movement pattern extractor. This part works independently of the other parts. The data collected from the acceleration sensor chip reflects the change in acceleration that is not felt by real people, so it is necessary to remove this component. Therefore, the collected data is filtered to a frequency band suitable for human step, and then the step is searched according to various settings. This part operates independently of H / W. No porting is necessary.

   FIG. 7 shows an example of operation with the MSM6025 in a CDMA network, and shows an operating state and power consumption when the MSM enters a sleep mode. Once every 1.28 seconds (Wake-Up Interrupt) MSM enters the IDLE state, it changes the Main Clock from Sleep Clock to TCXO and the Pedo Task continues its operation for a while. At this time, the stored data is processed.

The operation sequence is as follows.

1.MSM == SLEEP

The RTC is used to loop and read and store data at regular intervals.

2. MSM == IDLE

After processing the data stored in step 1, read the data one by one and continue processing.

3. Continue from 1

   FIG. 8 shows a flowchart of an actual implementation when the solution is applied to a mobile phone using Qualcomm's MSM6025 chip operating in an actual CDMA network. When configured as shown in Figure 7 shows the power consumption as shown in Table 1 below.

Configuration Average actual value MSM6025 (Sleep) 3.51 mA MSM6025 (Sleep) + KXPS5 (ON, NO Sampling) 4.55 mA MSM6025 (Sleep) + KXPS5 (ON, 17Hz Sampling) 4.82 mA

The above configuration is Sampling at 17Hz when the MSM is in the Sleep state. This is done by SPI communication through GPIO. According to Table 1 above, the current consumption in each situation is as follows.

1.MSM6025 consumes average current of 3.51 mA

2. Current consumption when power is supplied to KXPS5: 1 mA (Max.)

3. Current consumed by the MSM6025 to collect and process data: 0.27 mA

   If it is not configured in the above way, but configured in the existing way with MiCom (cell phone), the total current consumption includes MiCom's current consumption in the above 1,2 items. Therefore, the configuration provided by this solution can reduce the current consumption and eliminate the cost of manufacturing MiCom, which is a great help in product competitiveness.

Recently, most mobile phones have been released with a built-in acceleration sensor for the purpose of rotating the LCD screen, etc. The present invention can be mounted on all mobile phones equipped with the acceleration sensor, and the pedometer function can be used in a low power state without hardware modification. Make it.

   It can provide the user with steps, movement speed, and exercise information for a certain period of time or a day, and send the collected data to the server as SMS messages to provide users with advanced and various health-medical services through mobile phones or Internet sites. Can provide.

   Because the technology does not significantly affect cell phone latency, the cell phone can be used as a 24-hour motion detection tool. It is also possible to develop and create various LBS services based on location + motion based on the location tracking function of mobile phones. It is possible to implement a service that automatically detects the fall of the elderly to notify whether the accident and the location of the accident, and those with the visually impaired swipe the phone to make SOS requests.

   In order to realize the real-time LBS service, the GPS function of the mobile phone must always be alive. In the case of GPS, the power consumption problem is very serious. As the acceleration sensor can monitor the movement of the mobile phone, it can provide a solution to real-time LBS function by temporarily turning off the GPS function when there is no mobile phone movement.

   All of the above application services must have a 24/7 motion detection function of the mobile phone, and it is essential that the mobile phone waiting time should not be greatly reduced due to this. However, the present invention provides the answer.

    1 is a basic H / W configuration diagram of the solution.

    FIG. 2 shows the overall composition of each layer, and the red and application parts are provided in the solution.

    Figure 3 shows the correlation between the data collection and power when the Idle, Run, Sleep state of the mobile phone.

    4 illustrates a correlation between data collection and power for a baseband processor without a sleep mode function.

    5 is an overall S / W system flow chart. The orange and turquoise areas are provided by this solution, and the green areas are the ports that need to be ported.

    6 is a simple diagram and interface configuration example of a driver that plays a role of reading acceleration values from SPI, I2C, and ADC in a designated memory.

    FIG. 7 shows an example of operation with the MSM6025 in a CDMA network, and shows an operating state and power consumption when the MSM enters a sleep mode.

    FIG. 8 shows a flowchart of an actual implementation when the solution is applied to a mobile phone using Qualcomm's MSM6025 chip operating in an actual CDMA network.

Claims (7)

    Low-Power Information Sensing Technologies Using Baseband Processors-     A technology that allows mobile phone baseband processors to collect measurement data from accelerometers, allowing baseband processors to collect information from accelerometers in sleep mode with a very slow sleep clock. The H / W and S / W hierarchical diagrams according to claim 1, comprising the above technique-FIGS. 2, 3, 4, 5, 6, 7, and 8    The method of claim 1,    Data storage technology in low power mode-    The baseband processor continues to store the acceleration sensor data while in the low power mode (sleep mode), and then exits the low power mode after a period of time, processing all the data gathered during that time.   The method of claim 1,    Data Acquisition Techniques for Baseband Processors Without Sleep Mode Some of the baseband processors do not have Sleep mode and Deep-Sleep (or Stop or Power-Off) mode, which collects and stores acceleration sensor data while waking the processor 20-30 times per second (minimum 17 times). Immediately after storage, switch to Deep-Sleep mode to minimize power consumption.   The method of claim 1,    Information Sensing Technology-    The data derived from the acceleration sensor can be processed in the unit of sample (Real-Time Process), but it is difficult to process it directly in the low power mode, so the batch process is also performed in parallel. That is, the number of samples of data to be processed at one time does not need to be constant, and the collected data is transmitted until processing is required.   The method of claim 5,    Sampling Cycle Adjustment Technology-    If there is no change in the number of steps for a long time, lowering the sampling rate from the chip reduces the power consumption of the entire system.    The method of claim 1,     Technology that provides user motion information wirelessly to server in real time and always as motion detection tool-      This technology enables 24 hours of motion detection using a mobile phone and wirelessly transmits motion information to the server in real time, allowing users to see the user's movement at a specific point in time or as a pattern for a certain period of time. Used as a base technology for implementing motion-based or LBS (Location Based Service) based applications and services.

Images