Abstract: Among the many factors affecting the lifecycle of wireless sensor networks, node energy is one of the most important. Node power consumption, deployment environment, and specific applications all significantly impact node energy supply. This paper analyzes current node energy supply systems based on energy mining and proposes an improved wireless network sensor energy supply system that utilizes ambient energy to permanently power wireless network sensor nodes. Keywords: Wireless Sensor Networks; Power Supply; Node; Energy Scavenging Abstract: Among all the factors affecting the lifecycle of Wireless Sensor Networks (WSNs), the most important is the power of the WSN node. Node power consumption, environment, and application all greatly influence the WSN node power supply. This paper analyzes the current WSN node power supply system based on energy scavenging and proposes an improved WSN node power supply system. Powering WSN nodes perpetually using environmentally available energy is now possible. Key Words: WSNs; power supply; node; energy scavenging 1 Introduction Wireless Sensor Networks (WSNs) have broad application prospects in fields such as battlefield environmental monitoring, home security, and smart spaces, and are increasingly attracting the attention of researchers. However, node energy has become a bottleneck for the effectiveness of WSNs. The energy supply system for wireless network sensor nodes should be designed according to their own characteristics. Sensor nodes have low power consumption, but the power consumption varies considerably. If energy mining technology is used to mine energy from the environment, enabling nodes to replenish energy, wireless sensor nodes will avoid the one-way energy reduction process, which will fundamentally solve the energy supply problem of nodes. Energy mining devices can mine various types of energy. The abundance of energy in various environments is different. Under the direct sunlight of the sun, it is 100 mW/cm² [1]. In addition, there is energy in the form of temperature difference, vibration, etc. At present, there are already research and development of systems that use environmental energy to provide energy for wireless sensor network nodes. For example, the PMG7 micro generator launched by the startup Perpetuum can generate up to 5mW/3.3V output power from a 100mg vibration [2]. However, the use of vibration energy means that nodes can only be deployed in areas where vibration often occurs, which limits the deployment environment of nodes. In addition, under intermittent vibration conditions, the system cannot work continuously. This paper proposes a solar energy supply system. The system uses multi-level energy storage, combined with energy management and energy transfer technology, to make reasonable use of the energy collected by solar cells, thereby forming a self-managing energy supply system, realizing the purpose of providing permanent power to wireless nodes and unlimited use of wireless sensor networks. 2 Design and Analysis To better address the energy supply problem of sensor nodes, we propose a solar-based energy supply system, consisting of the following components: an energy mining device, composed of solar panels, responsible for converting solar energy into electrical energy; an energy storage unit, including a primary energy storage unit and a secondary energy storage unit, composed of supercapacitors, responsible for storing the energy collected by the solar cells and powering the wireless network sensor nodes; a backup energy storage unit, composed of lithium batteries, which serves as the system's energy source in emergencies; and a power management and control unit, responsible for monitoring the energy levels of the primary, secondary, and backup energy storage units, controlling these energy storage units to power the system based on their status, and controlling the solar cells to replenish the energy storage units. The overall system structure is shown in Figure 1. The main operating mode of the system is that the primary and secondary energy storage units store the collected environmental energy and power the sensor nodes, while the backup energy storage unit serves as a reliable backup power source in emergencies, enabling the system to operate well under conditions of intermittent environmental energy. The solar cells used in the energy mining device have an efficiency of approximately 16%-17%, and an output power of approximately 16-17 milliwatts per square centimeter under direct sunlight. If the voltage is determined by selecting a configuration that meets the system voltage requirements, the ambient energy can be used as a power source model: Pe(t) = Punit(t) * A (1) where Punit(t) is the energy per unit area per solar panel, and A is the number or area of ​​the solar panels connected in parallel. Although the capacity of the energy storage is limited, Pe(t) should be as high as possible in order to extract as much energy from the environment as possible. In this way, after replenishing the energy storage with sufficient energy, the system can continue to utilize ambient energy. However, the size of the solar panel should be considered in conjunction with factors such as system performance indicators, volume, and cost. The primary and secondary energy storage are composed of supercapacitors. As the main energy source of the node, it needs to store the energy extracted by the energy extraction device, so it should be able to withstand frequent charging and discharging. The capacitor can meet this performance requirement, and a large-capacity supercapacitor can provide sufficient capacity. Therefore, it is most ideal to use capacitors as the primary and secondary energy storage. In order to extend the power supply time of the primary energy storage, the appropriate capacitor capacity should be determined according to the requirements of leakage current, energy consumption level, and system startup time. According to reports, the 22F capacitor has the lowest leakage current. Considering the requirements of system cost and size, a 25F supercapacitor was selected as the energy storage element, and the leakage current was reduced by connecting two supercapacitors in series. The backup energy memory is composed of lithium batteries. The characteristics of lithium batteries are: low leakage current, high energy density, and high single-cell voltage. Therefore, lithium batteries were selected to form the backup energy memory. However, it must be noted that it requires a complex charging circuit to prevent harmful effects on the battery. The power consumption of the sensor node in active mode and sleep mode is very different. The power consumption is determined by three parameters: the percentage of active time to total time D, the current consumed in active mode Ia, and the current in sleep mode Is. In most cases, we are only interested in the average power consumption P (assuming that the wake-up time can be ignored): P = Vsup × (D × Ia + (1 − D) × Is) (2) The above formula shows that Ia, Is, Vsup and D are all factors that affect the power consumption of the system and need to be considered in implementation. 3 Implementation of the scheme The energy supply system is implemented on the sensor node. The system is connected to a node through a 40-pin interface [3], as shown in Figure 2. The system uses a 60 mm × 60 mm solar panel as an energy mining device with an output voltage of 4.4 V; it uses four 25 F supercapacitors to form a two-stage energy storage system, each with a maximum rated voltage of 2.7 V; the two supercapacitors are connected in series as a first-stage energy storage system, which can reduce leakage current and match the output of the solar cell of 4.4 V; a lithium battery with a capacity of 1120 mAh and a working voltage of 3.7 V is used as a backup energy source. [align=center] Figure 2 System physical diagram[/align] The voltage curve of the energy storage device composed of two 25 F capacitors connected in series under the influence of leakage current is shown in Figure 3. The measurement time range is 24 hours. In order to reflect the overall trend of the energy storage device, a piecewise linear fitting method is used to obtain the following leakage current variation law: (3) Where V0 is the initial voltage in V and t is the time in hours. [align=center] Figure 3 Leakage current curve[/align] As can be seen from the above, the influence of super leakage current must be considered within 24 hours, and the energy lost due to leakage current cannot be utilized by the system. The energy storage device composed of solar energy and supercapacitor must be charged by solar energy. The charging energy of the solar cell must match the capacitive energy of the supercapacitor. That is, under normal light conditions, the solar cell can fully charge the two-stage energy storage device. Figure 4 shows the voltage curve of the energy storage device during charging. The charging voltage curve of the energy storage device can be obtained by fitting the measured voltage value. The fitted equation is as follows: (4) Where A is the voltage when charging is stopped, that is, the voltage provided by the solar cell, which is 4.4. The two coefficients B and C represent the external conditions during charging, such as light. B and C will change with different external conditions. In this curve, B is 0.06348, C is 2.17868, and the fitting degree is 99.766%. The selection of solar cells that match the energy storage must be limited by the above results. [align=center] Figure 4 Charging curve[/align] The control part of the system uses a C8051F121 microcontroller [4], and uses on-chip AD to monitor the voltage of the supercapacitor to determine its energy state. The DS2760 can be used to protect the lithium battery and monitor the system power consumption and the energy state of the lithium battery. The power consumption in the node active mode is 48mA. According to the manual, the power consumption in the low power mode is 35uA. According to formula (2), if the node active cycle is 1%, the average power consumption is (48mA+99×35uA)÷100, which is 515uA. According to the energy size and state of each level of energy memory, the microprocessor controls the energy memory selection switch to control the solar cell to charge the energy memory and the energy memory to power the system. The control process is as follows: a: Start the system (battery powered). b: Charge the primary energy storage; c: Determine if the primary energy storage is fully charged; if fully charged, proceed to step d; if not fully charged, proceed to step b; d: The primary energy storage supplies power to the system and charges the secondary energy storage; e: Determine if the secondary energy storage is fully charged; if fully charged, proceed to step f; if not fully charged, proceed to step d; f: Charge the primary energy storage and determine if the battery needs additional energy; if so, proceed to step g; if not, proceed to step h; g: Fully charge the battery; h: The system operates normally according to the above steps. The simulation node operates at 1% duty cycle, meaning the node spends 99% of its time in low-power mode and 1% in normal mode. A power supply experiment is conducted. With the supercapacitor fully charged, the power supply time is measured. Based on multiple experiments, the energy in the single-stage energy storage device can power the system for 745 minutes. Online debugging is performed at 1% duty cycle; the system operates stably for two days, preliminarily proving the system's feasibility. 4 Conclusion A system structure for providing permanent energy to wireless network sensor nodes using environmental energy was proposed and initially realized. Its innovation lies in: cleverly utilizing energy mining technology to mine energy from the environment, enabling the node to have the ability to replenish energy, avoiding the one-way energy reduction process, and further combining energy management, energy transfer and energy storage technologies, thereby fundamentally solving the energy supply problem of the node. References [1] Joseph A. Paradiso, Thad Starner. "Energy Scavenging for Mobile and Wireless Electronics". Pervasive Computing, JANUARY–MARCH 2005. [2] Vibration Powering Wireless Sensors. Electronic Design Technology (EDN CHINA), 2006, 09. [3] Guo Peng, Zhao Zhan, Fang Zhen, Zhang Yuguo, Design of Layered Multiplexing WSNs Nodes and Their Software Platform, Microcomputer Information: Embedded and SOC On-Chip Systems, 2006, 11, 49-51 [4] Translated by Pan Zhuojin. C8051F120/1/2/3/4/5/6/7C8051F130/1/2/3 Series Mixed-Signal ISP FLASH Microcontroller Datasheet [M]. 2004, 12.