Application of wireless sensor network technology in human body parameter collection

Physiological parameter acquisition refers to the measurement, processing and transmission of certain physiological parameters of a person when he is stationary or active in a specific environment. This technology has a wide range of applications, such as medical treatment, health care, sports, military and clothing comfort evaluation. The implementation method of physiological parameter acquisition is to place sensors on the corresponding parts of the human body and transmit the data collected by the sensors to the terminal for processing in a wired or wireless manner. In use, the wired transmission method is sometimes limited, so it is necessary to study the method of using wireless data transmission. For example, in the application of clothing comfort evaluation, the main methods are: measuring the temperature and humidity of the human body surface in a real environment; measuring the temperature and humidity of the real or fake body surface in a simulated environment (artificial climate chamber), and giving subjective and objective evaluations based on the measured values. When conducting dynamic human wearing experiments in a real environment, wireless data acquisition and transmission methods will bring great convenience. Wireless sensor network technology has developed rapidly. The application of this technology to human physiological parameter acquisition has been applied in China. For example, reference [2] proposed the application of wireless sensor networks based on Zigbee technology in remote home monitoring, and reference [3] proposed the application of wireless sensor technology in medical monitoring. This paper will analyze and design human body parameter collection from the aspects of wireless sensor network node hardware and software platform selection, network architecture, MAC layer protocol, node low power design, etc., combined with specific applications.

1 System Design

1.1 Network Architecture and Protocols

The network topology mainly describes the connection mode of network nodes. The human physiological parameters are collected in a small range, the sensor nodes are concentrated, and there is generally no need for communication between nodes. Based on these characteristics, a star network topology can be used. In a star topology, each branch node is connected to the central node in a point-to-point manner. When there is a large amount of communication between the central node and the branch nodes, the star topology is the most effective. Its advantages are that it can concentrate resources, the network is easy to manage, the coverage is concentrated, the routing algorithm is relatively simple, and the problematic nodes can be easily isolated by the central node without affecting the performance of other nodes.

In wireless sensor networks, the MAC (Medium Access Control) protocol determines how wireless channels are used and allocates limited wireless communication resources between sensor nodes. It has a significant impact on the performance of sensor networks and is one of the key network protocols to ensure low power consumption of wireless sensor networks. In human physiological parameter collection applications, sensor nodes periodically collect data and are often powered by small button batteries. The central node generally has sufficient power, so the CSMA/CA (Carrier Sense Multiple Access with Collision Avoidance) control protocol is used. The central node does not turn off the RF module and continues to receive. The sensor node samples and sends data regularly. When it needs to send data, it turns on the RF module and first listens to the channel. If the channel is occupied, it backs off for a random period of time and then continues to listen. If the channel is not occupied at this time, it sends data, and after confirmation, it turns off the RF module. The MCU enters sleep mode and waits for the next timing trigger.

1.2 RF transceiver devices and network operating system

The sensor node is required to be small in size, low in power consumption, and capable of data processing and wireless transceiver. Here, the CC2430EM evaluation board of TI/Chipcon is used as the wireless transceiver module, and the sensor circuit is designed according to the needs. The P0.1~P0.7, P1.0~P1.7 and P2.0~P2.2 lines of the CC2430 evaluation board can be connected to the sensor circuit for control. The chip CC2430 embeds the CC2420 RF transceiver and 8051 MCU, contains a 12-bit analog/digital converter, 4 timers, a sleep mode timer of a 32 kHz crystal oscillator, hardware supports CSMA/CA, and the power supply voltage is 2.0 V~3.6 V. The current consumption in power-down mode is only 0.5 μA, and it can be awakened by an external interrupt or real-time clock. The wireless sensor network can be regarded as a computer system composed of multiple CPUs with limited energy, storage space and processing power. These CPUs are independent of each other and work together. In order to facilitate software development and maintenance and improve software development efficiency, the TinyOS embedded operating system is used. TinyOS is a component-based operating system designed specifically for embedded systems. It is mainly used in wireless sensor networks and is implemented in the nesC language. It uses lightweight thread technology, active message communication technology, event-driven mode, and component programming. Using this operating system can improve CPU utilization. Under the scheduling of TinyOS, all tasks associated with communication events can be processed quickly when events are generated. When the processing is completed and there are no other events, the CPU will enter a sleep state.

1.3 Node Design

Figure 1 is a schematic diagram of the human physiological parameter collection network structure. To improve portability, the PC can be replaced with an embedded device. Each sensor node in the figure samples and sends data at a fixed time. After receiving the data, the central node sends it to the PC through the serial port for display and processing.

There are two types of nodes: sensor nodes and central nodes. In the TinyOS environment, the programming language is nesC. To support the component-based programming model, the nesC language introduces the concepts of interface and component. Interfaces are function declarations with similar or related functions. They are named as commands or events according to the calling direction and are implemented in components that provide or use the interface. Components include accessories and modules. Accessories are responsible for connecting different components through interfaces, and modules provide the code implementation required by the program. For sensor nodes, it is necessary to design hardware driver components for different sensors, and design configuration components and module components of the application layer in combination with existing intermediate components (such as system components, power management components, A/D conversion components, timer components, radio frequency components, etc.). The following takes the temperature and humidity sensor node as an example to illustrate the design method. The temperature and humidity sensor uses SHT10 from SENSIRION, Switzerland.

The component connection of the temperature and humidity sensor node is shown in Figure 2. The small rectangular box contains the interfaces used and provided by each component, and the arrow direction indicates the direction of the command call. Because all sensor nodes have functions such as timed sampling, data processing, data transmission and reception according to the protocol, and RF module and MCU power management, a component SHT10_C is designed separately for sensor-related operations. In the top-level accessories, the functions of each sensor node are realized by connecting with different types of sensor components. Other functions that are not related to sensor operations and related to applications are combined to design an intermediate component SensorNet_C. By connecting with the lower-level accessories McuSleepC, CsmaC and CC2430ActiveMessageC, MCU power management, CSMA protocol and RF module control are realized. Accessories SHT10_C and SensorNet_C only provide the connection relationship between components, and their implementation is completed by SHT10_M and SensorNet_M respectively. When calling the command Send.send() to send data, a clean channel assessment is first performed. If the channel is occupied, it is necessary to retreat for a period of time. The backoff time is obtained by multiplying a random number between 1 and 31 generated by a random function and a given initial value (here set to 160 μs, i.e. 10 symbols). The initial value can be adjusted according to the number of nodes in the network. If there are fewer nodes, the initial value is smaller, otherwise it is appropriately increased. The top-level configuration component program of the temperature and humidity sensor node is as follows:

configuration SensorNodeSHT_C {
}
implementation {
components MainC； /*TinyOS2 main module, used here for associated system startup*/
components new TimerMilliC() as TimerC；
components SensorNet_C； /*RF module, CSMA protocol and MCU power management control configuration components*/
components SHT10_C； /*Temperature and humidity sensor SHT10 configuration components*/
components SensorNodeSHT_M； /*Top-level module components*/
SensorNodeSHT_M.Boot->MainC.Boot；
SensorNodeSTH_M.Timer->TimerC；
SensorNodeSTH_M.RFControl->SensorNet_C；
SensorNodeSTH_M.AMPacket->SensorNet_C；
……
SensorNodeSTH_M.Send->SensorNet_C；
SensorNodeSTH_M.SHT->SHT10_C； /*Implement the connection of interface STH*/
}

In the accessory SensorNet_C, the RF function provided in the module component CC2430ActiveMessageP is used by connecting the accessory CC2430ActiveMessageC. CC2430ActiveMessageP is the top component in the RF stack and provides a single-hop communication implementation method. The MCU power management function in the module component McuSleepP is used by connecting the accessory McuSleepC to implement functions such as MCU sleep, timing, and startup to reduce node power consumption. The events McuSleepControl.beforeSleep() and McuSleepControl.afterWakeup() need to be implemented in the upper-level component SensorNet_M that uses the McuSleepP component to protect and restore the state before and after sleep.

The temperature and humidity sensor SHT10 has a two-wire serial interface and outputs a calibrated digital signal. The interface SHT10 is defined as follows, and the commands defined by the interface need to be implemented in the module component SHT10_M.

interface SHT{
command error_t read()； /*Implemented in sensor module component STH10_M*/
event void readDone(error_t result，uint16_t temperature，uint16_t humidity)； /*Implemented in top-level module component SensorNodeSHT_M*/
}

The central node receives radio frequency data and then sends the data to the PC through the serial port. Its top-level configuration file is as follows:

configuration CenterNode_C{
}
implementation{
components CenterNode_M; /*Top module components*/
components MainC;
components CC2430ActiveMessageC as ActiveMessageC;
components ABSC; /*Serial communication control components*/
CenterNode_M.Boot->MainC.Boot;
CenterNode_M.RFControl->ActiveMessageC ;
CenterNode_M.AMPacket->ActiveMessageC;
CenterNode_M.Packet->ActiveMessageC;
CenterNode_M.Receive->ActiveMessageC.Receive;
CenterNode_M.ABS->ABSC;
}

2 Calculation of sensor node power consumption

When collecting physiological parameters of the human body, especially in outdoor applications, low power consumption is very important for battery-powered sensor nodes. The following application method in reference [6] is used to calculate the power consumption of the sensor node in this design. A 3.3 V DC power supply is used, and a 10 Ω resistor with a 1% accuracy is connected in series with the CC2430EM. The current in each time period of a sampling cycle is calculated by measuring the voltage drop across the resistor. Assume that the sampling period T=10 s, and the CC2430 RF output power is set to 100%. Figure 3 shows the voltage change waveform across the resistor obtained when the sensor node sends data using the oscilloscope TPS2024, with a horizontal direction of 2.50 ms/div and a vertical direction of 100 mV/div. Table 1 shows the current consumption during the activity period.

Low power mode time within a sampling period T (A segment, G segment, CC2430 in PM2 mode):

TPM2=T-Ton=10000 ms-20.45 ms=10 451.65 ms

Current consumption in low power mode (CC2430 in PM2 mode):

0.000 5 mA×10 451.65 ms=5.225 8 mA·ms

Total current consumption within a sampling period T:

472.1 mA·ms+5.225 8 mA·ms)/(3 600 000 ms/h)=1.326×10-4 mAh/10 s

Current consumption per hour: 1.326×10-4×360=0.047 74 mA

Assuming a 60 mAh button battery is used, the usable time is: 60 mAh/0.047 74 mA=1 256 h≈52 days

This paper uses nesC language to design the central node and sensor node programs based on TinyOS, realizes the star-shaped wireless sensor network using CSMA/CA protocol, and takes temperature and humidity data collection as an example to design a sensor node using the temperature and humidity digital sensor SHT10. When the sampling period is 10 s and powered by a 60 mAh battery, the sensor node can work continuously for 52 days. This design can meet the needs of wireless, portable, and low-power collection of human physiological parameters. The use of modular programming language nesC improves development efficiency and facilitates expansion. The research and design methods introduced in this paper can be used in related applications.