2020 International Conference on Computer Engineering and Application (ICCEA)

Design and Application of Intelligent Agriculture

Service System With LoRa-based on Wireless Sensor
Network

Yang Yang1
ShengLi College China University of Petroleum.
No. 271, North Second Road.
Dongying City, Shandong Province, China.

Abstract—In view of the current intelligent agriculture's

collection and processing of production site data and the control transmission, can realize four modes of fixed point, power
requirements of related equipment, ZigBee, WiFi, GPRS saving, transparent transmission and monitoring. Finally, the
traditional wireless sensing technology has short trans-mission transmitted data is processed and saved, and finally the
distance and signal interference in greenhouse environment intelligent farm monitoring model based on LORA technology
monitor-ing systems such as large-scale farms. shortcomings such is designed and completed. In the laboratory, the detection of
as complex net-working, This work proposes a low-power farm soil and model environmental parameters based on various
environment detection system based on LoRa wireless technology. sensors is completed, and lighting control, irrigation control,
This model can use various sensors in the laboratory to complete cooling control, and human-machine con-trol are realized.
the detection of soil and model environmental param-eters, and After the experimental model was established, the feasibility of
implement lighting control and irrigation control, cooling control the technology was verified, and the system was applied to the
and man-machine control. The designed system sleep current is dragon fruit greenhouse and data collection and analysis were
as low as 0.3μA, the average working current, has the advantages
completed.
of long transmission distance and low power consumption. After
verifying the feasibility of the technology through an
experimental model, the system was applied to a spe-cific II. THE DESIGN OF AGRICULTURE SERVICE SYSTEM
greenhouse and data collection and analysis were completed.
A. Structure of system
Keywords-component; LORA; Internet Agriculture; Wireless The overall design of the system is shown in figure1. It is
Sensor Network divided into three lay-ers. The user accesses the platform
control device to form the sensing layer composed of the
I. INTRODUCTION application layer, the network layer, and various sensors. There
At present, based on the rapid rise of Internet of Things are 4 kinds of sensors used in the sensing layer to detect carbon
technology, intelligent agriculture is being used [1]. Compared dioxide concen-tration, soil moisture, air humidity, temperature,
with traditional Agriculture, to ensure the real-time and and light intensity. The commu-nication module used in the
effectiveness of data, the collection, transmission and network layer is AS32TTL100 based on LoRa protocol. The
processing of parameters of crop growth environment require AS32TTL100 is used to establish a connection with the
various sensor devices and effective wireless transmission. gateway. The control-ler uses the STM32F103ZE
Technology and many farms or greenhouses in remote areas do microcontroller to establish the sensor and the micro-controller
not cover cellular data, let alone 4G [2-4]. The traditional wire- to obtain the sensor data and then store it in the memory and
less sensor technology for ZigBee, WiFi, and GPRS has a short send it. The STM32 will collect the data. The incoming data
transmission dis-tance and is vulnerable to greenhouse can be uploaded to the server over the network and can receive
environment monitoring systems such as large-scale farms. commands from the server.The application layer includes a
Insufficient external interference, the use of LORA technology user access platform and a control device composed of a PC
to build a private Internet of Things is very suitable for this end and a mobile phone client. The user access platform can
scenario [5-7]. monitor the growth envi-ronment information of the crops in
the greenhouse and process and save the system data, and the
In this paper, an intelligent agricultrue system based on control device node includes a fan node and a roller blind.
LORA technology is designed. Firstly, a system model is built Control nodes, lighting nodes, irrigation nodes, etc., perform
in the laboratory environment. Various sensors such as soil corresponding ac-tions according to instructions, such as
temperature and humidity sensor, light intensity sensor, and measuring, supplementing lighting, re-plenishing moisture,
carbon dioxide concentration are used to collect the parameters ventilation, and the like. At the same time, the PC interface and
in the crop growth environment. LORA is adopted. Wireless the mobile phone software can be designed to take the network
spread spectrum communication technology for data selection date and take out the historical data from the server,

978-1-7281-5904-1/20/$31.00 ©2020 IEEE 712

DOI 10.1109/ICCEA50009.2020.00155

Authorized licensed use limited to: National Taipei Univ. of Technology. Downloaded on August 06,2021 at 04:27:25 UTC from IEEE Xplore. Restrictions apply.
and then draw the data form the window interface for people to
make an intuitive judgment. People can also intervene
artificially by sending commands to from the receiving end, the
receiving end receives the command and then controls the
operation of the pro-duction device. At the same time, when
planting different plants, different plant growth databases are
needed, and different plants are controlled by software to
control different plants.

B. Selection of Wireless Communication Technology

There are many kinds of IoT application technologies,
which can be divided into two types according to the
transmission distance. One is short-distance commu-nication
technology, including Zigbee, Wi-Fi, Bluetooth, etc. The other
is long-distance wireless communication technology, including
wireless LAN (Wireless LAN). Wide-Area Net-works
(WWAN) and Low-Power Wide-Area Net-work (LPWAN),
low-power WAN technology is a revolutionary Internet of
Things access technology in recent years in smart homes. The
fields of intelligent agri-culture and other fields are becoming
increasingly widely used. Through the expe-rience of
intelligent control of greenhouse farms at home and abroad, the Fig. 1. Intelligent Agriculture Service Platform
appli-cation of sensors such as temperature, humidity, carbon
dioxide, and light inten-sity has important significance for
C. Design of sensor nodes
improving agricultural production and reducing water
consumption. If the growth environment of crops is guaranteed The selection of sensor modules is mainly for plant growth
to be the most range of good products, then the output of environment (car-bon dioxide concentration, soil moisture, air
agricultural products will be significantly improved, and the humidity, temperature, light intensi-ty), CCS811 for carbon
quality will be greatly improved, thus achieving efficient dioxide detection, HDC1080 for temperature and humidi-ty,
agricultural production. YL69 for soil moisture, and GY302 for light intensity.
Essential element of plant growth [11-13].
AS32- TTL-100 module power 100 mW (20 dBm),
receiving sensitivity -130 dBm, antenna gain 5 dBi, frequency The carbon dioxide concentration using the sensor for
410-441 MHz, providing a total of 32 chan-nels, configurable CCS811 is based on the Cambridge CMOS sensor's unique
address 65536, transmission distance ideally 3000 m With TTL micro-hot plate technology that enables gas with a highly
serial port input and output, baud rate 1200-115200 bps, there reliable solution for sensors, very fast cycle times and
are four commu-nication modes, namely general mode, wake significantly reduced average power consumption with
mode, power saving mode and sleep mode[8-9]. According to conventional MOX gas sensors [14]. The sensor is highly
the SX1278 modulation chip produced by the LoRaWAN sensitive to carbon dioxide and has good selectivity for fast
protocol developed by the LoRa Alliance, a low-speed and response. It has a long-term service life and reliable stability.
long-distance technolo-gy is applied. Physical layer or wireless The principle is to utilize the galvanic effect of the reaction of
modulation is used to establish long-distance communication metal oxides with organic compounds, and then generate
links. Many traditional wireless systems use frequency shift digital data by amplification conversion according to the mag-
keying (FSK) modulation as the physical layer because it is a nitude of the generated current. The sensor has a built-in
very efficient modulation that achieves low power consumption microprocessor and is accompanied by a program. When the
[10]. The MAC layer protocol is mainly implemented by power is turned on, the command is sent through the IIC bus,
program control, which realizes multi-channel communi-cation, the BootLoader program is started on the sensor, the sensor is
channel management and handover between communications, switched from sleep to the measurement state, and then the
adaptive transmission rate, timing, transmission and reception, register address is sequentially transmitted, and then the sensor
node access check, and data encryption, etc. LoRaWAN can data is taken out.
also eliminate Incompatibility between hardware, such as the The air temperature and humidity sensor uses HDC1080, a
emergence of multiple LoRa devices in the same system, digital hu-midity sensor with an integrated temperature sensor
LoRaWAN plays a key role. Linear spread spectrum has been that provides an excellent temperature sensor. Measurement
used in military and space communications for decades because accuracy at very low power. A key feature of HDC1080 is its
it can achieve long communication distances and robustness to low power consumption, which makes the device suitable for
interference. bat-tery or power harvesting applications. In these applications,
HDC1080 spends most of its time in sleep. Mode: With a
typical 100nA current consumption in sleep mode, the average
current consumption is minimal. Its low consumption in
measurement mode minimizes any self-heating. An integrated

713

Authorized licensed use limited to: National Taipei Univ. of Technology. Downloaded on August 06,2021 at 04:27:25 UTC from IEEE Xplore. Restrictions apply.
sensor produced by the change of humidity in the air to stage directly. This is the instruction sent to the wireless
produce different conductivity and the change of the thermistor module through the serial port according to the fixed time, and
due to the temperature change resistance. The sensor is then waits for the data returned from the measuring end. After
internally provided with a register to store data, and the the comput-er receives the data, it first The data is saved in the
working state of the sensor can also be changed by changing buffer to determine whether the data is normal. If it is abnormal,
the register value through the IIC. The bus communicates with the data needs to be discarded. If it is nor-mally put into the file
the control chip. and imported into the file, the file is numbered according to the
date of the day plus the device number, so that the historical
The soil moisture sensor is made of YL69 and consists of a data can be viewed later.
stainless steel probe and a waterproof probe. It can be used in
soil and dam for a long time. It can car-ry out fixed-point Click the database view button to pop up the following
monitoring and on-line measurement of surface and deep soil. dialog box. After select-ing the date, click OK to display it
Used in conjunction with data loggers, it can be used as a tool directly in the display area of the main inter-face. By
for moisture point monitoring or mobile measurement to performing multiple measurements every day, the data is scaled
measure soil volumetric water content, primarily for soil through the pixel points on the interface. It is convenient for
moisture detection and agricultural irrigation and forestry pro- people to intuitively feel the change of data to understand the
tection. The potentiometer adjusts and controls the state of plant growth. At the same time, you can click the
corresponding threshold. When the humidity is lower than the display button to view the image.
set value, DO outputs a high level. When the value is higher The mobile phone program is written in JAVA language,
than the set value, DO outputs a low level. This function is and the Android Studio software is used for interface and
often used for the humidity threshold control switch; the program kernel development. The mobile phone program
comparator uses The LM393, operates stably, and has a wide currently uses Bluetooth to simulate network communication.
operating voltage to ensure the stability of the work. Here, the mobile phone needs to be connected with the
The light intensity sensor uses GY302, GY-302 uses computer, which is equivalent to connecting with the server.
ROHM original BH1750FVI, the light intensity range The UI interface of the mobile phone is as shown in the
measures the output range from 0 to 65,535 lux, and the sensor figure below. There are three windows. The main function is
has a built-in 16-bit AD converter for direct digital output, equivalent to the mobile remote control. It is also portable.
omitting the complicated calculations.GY-302 is equipped with Sometimes people may need to observe the field. At this time,
photosensitive elements inside, which generate different the computer cannot be used to view data or control devices.
current values under different illuminations. According to the Operation, when the mobile phone has the need to exist, it
complicated experimental data, the current data is in one-to-one performs the same function as the manual control button on the
corre-spondence with the light intensity lux, and the data with
computer side.
the illuminance range of 0-65535 can be directly output.
Through the IIC protocol connected to the single-chip
microcomputer, the single-chip microcomputer can collect the III. RESULTS OF DRAGON FRUIT SHED
data of GY-302, the stronger the illumination, and then the
purpose of measuring the light intensity. A. Laboratory model test results
First, STM32 must first establish a connection with the
D. The design of Intelligent Agriculture Service Platform gateway through AS32TTL100, so that the data collected by
The program is developed by Visual Studio software of STM32 can be uploaded to the server through the network, so
Microsoft Corporation using Visual Basic language. The that it can have the process of intelligent analysis and manual
computer simulates the function of the server here, which is supervision, and can receive instructions from the server and
equivalent to the data sent from the measurement terminal, is execute according to the instructions. Corresponding actions
directly transferred to the computer through the gateway, and such as measuring, supple-menting the light, replenishing
saved, ignoring the server link. moisture, ventilation, etc. Then, the sensor is connected with
the STM32. The sensor is built into the sensor to AD
The program is always in the background, running state, conversion, register configuration, and data reading in different
can not be closed, the program sets the timer, and each time the environments. CCS881, GY302, HDC1080 communicate
timer reaches a fixed duration, it exe-cutes the serial port to through IIC, and STM32 establishes analog IIC.
send the command program, which is equivalent to send-ing to Communication, acquiring sensor data and then storing it in
the gateway, and the gateway sends it to the measuring end to memory and trans-mitting it. Secondly, through the high and
perform the corresponding action. Manual control can also be low of the STM32IO port, the relay switch is controlled to
performed. At this time, the automatic management link is control the operation of the production equipment. At the same
closed, and all actions need to be manually performed manually. time, the camera system is designed for image transmission, so
that peo-ple can directly observe plant growth through the
The data returned by the measuring end is first saved
software without taking a lot of manpower to conduct on-the-
according to the date and equipment number, and then the
spot observation. People can timely discover prob-lems such as
production status of the farm is analyzed based on the current
plant diseases and insect pests, and then decide whether to sup-
data, and the corresponding environment is changed. The
plement fertilizers and spray pesticides in time.
running process of the program can be divided into two parts.
After the software is opened, it will enter the automatic control

714

Authorized licensed use limited to: National Taipei Univ. of Technology. Downloaded on August 06,2021 at 04:27:25 UTC from IEEE Xplore. Restrictions apply.
 The PC software first removes the historical data from the
server through the network selection date, and then draws the
data from the window interface for people to make an intuitive
judgment. People can also intervene artificially by sending
instructions to the receiving end, and the receiving end receives
the da-ta. The instructions then control the operation of the
production equipment. At the same time, when planting
different plants, different plant growth databases are needed,
and different plants are controlled by software to control
different plants. The purpose of the mobile terminal is
portability. When people observe the problem, if they find the
problem, they can directly send the command to the
measurement terminal through the mobile phone for control.
The function is the same as the PC, and it also interacts with
the server and can view historical data and control in real time.
Figure 4 shows the results of the mobile terminal inter-face test. Fig. 4. Air humidity test results for a certain da y

B. Dragon fruit greenhouse test results Equations

After laboratory tests prove that the system is feasible,
select a dragon fruit greenhouse for on-site testing, the main
test parameters are greenhouse tem-perature, air humidity, soil
moisture, greenhouse carbon dioxide concentration, and
greenhouse light intensity (to ensure the growth of dragon fruit,
the green-house in the afternoon Four points are the time to
turn on the lights). Figure 2 - Figure 6 test results are as
follows 4 conclusions

Fig. 5. Soil moisture test results for a certain day

Fig. 2. temperature test results for a certain day

Fig. 6. Photograph of the light intensity test for a certain day

It can be seen from the figure that the changes of these five
parameters are convenient for the greenhouse managers to
observe.

IV. CONCLUSION
After laboratory testing and actual greenhouse data
collection, it is proved that the intelligent farm monitoring
Fig. 3. Test results of carbon dioxide concentration test for a certain day system based on LoRa wireless transmission technology is
feasible. Selecting a dragon fruit greenhouse for on-site testing
can realize greenhouse temperature, air humidity, soil moisture,
greenhouse carbon dioxide concentration, and shed. The
collection of internal light intensity, these data can help farm

715

Authorized licensed use limited to: National Taipei Univ. of Technology. Downloaded on August 06,2021 at 04:27:25 UTC from IEEE Xplore. Restrictions apply.
managers better understand the growing environment of crops, [7] Computers in Technical Systems. IEEE, pp. 11 85–1189, 2016.
with good real-time and effectiveness. [8] M. Saari, A. M. Baharudin, and S. Hyrynsalmi, “Survey ofprototyping
solutions utilizing Raspberry Pi,” in 2017 40th International Convention
on Information and Communication Technology, Electronics and
REFERENCES Microelectronics, MIPRO 2017 -Proceedings, 2017.
[1] Gardašević, Gordana, Mladen Veletić, Nebojša Maletić, Dragan [9] Tanumihardja, Wisena Aditya, and Edy Gunawan. "On the application
Vasiljević, Igor Radusi-nović, Slavica Tomović, and Milutin of IoT: Monitoring of troughs water level using WSN," in IEEE
Radonjić."The IoT Architectural Framework, Design Is-sues and Conference on Wireless Sensors, 2015, pp. 58-62
Application Domains." Wireless Personal Com munications 92, no. 1 [10] F. Orfei, C. Mezzetti, and F. Cottone, “Vibrations powered LoRa sensor:
(2017): 127-148 An electrome-chanical energy harvester working on a real bridge,” IEEE
[2] U. Raza, P. Kulkarni, and M. Sooriyabandara, “Low power wide area SENSORS, pp. 1-3, 2016.
networks: An over-view,” IEEE Communications Surveys Tutorials, vol. [11] M. J. Mudumbe and A. M. Abu-Mahfouz, “Smart water meter system
19, pp. 855–873, Secondquarter 2017. for user-centric consumption measurement,” in Proceeding of the IEEE
[3] “Introducing new strategy to unify iot services under smartthings, in International Conference on Industri-al Informatics,2015, pp. 993–998.
addition to the next-generation bixby 2.0, and advanced ar capabilities [12] C. P. Kruger, A. M. Abu-Mahfouz, and S. J. Isaac, “Modulo: A modular
on more galaxy smartphones.” Samsung Developer Conference, October sensor network node optimised for research and product
18, 2017. development,”in Proc of the IST-Africa Conference and Exhibition,
[4] Vodafone, “Benchmarks and best practices to help achieve the greatest 2013, pp. 1–9.
return from internet of things projeA. M. Baharudin, M. Saari, P. [13] P. R. Page, A. M. Abu-Mahfouz, and M. Mothetha, “Pressure
Sillberg, P. Rantanen, J. Soini, and management of water distribution systems via the remote real-time
[5] T. Kudora, “Low-energy algorithm for self-controlled wireless sensor control of variable speed pumps,” J. Water Resour. Plan. Manag., 5. vol.
nodes,” in The 4th International Conference on Wireless Networks and 143, no. 8, 2017.
Mobile Communications (WINCOM), At Fez, Morocco. IEEE, pp. 1–5, [14] Wixted, Andrew J., Peter Kinnaird, Hadi Larijani, Alan Tait, Ali
2016. Ahmadinia, and Niall Stra-chan. "Evaluation of LoRa and LoRaWAN
[6] M. Saari, A. M. Baharudin, P. Sillberg, P. Rantanen, and J. for wireless sensor networks." In SENSORS, 2016 IEEE, pp. 1-3.
Soini,“Embedded Linux con-trolled sensor network,” in 2016 39th IEEE,2016.
International Convention on Information and Commu -nication
Technology, Electronics and Microelectronics (MIPRO), no. CTS

716

Authorized licensed use limited to: National Taipei Univ. of Technology. Downloaded on August 06,2021 at 04:27:25 UTC from IEEE Xplore. Restrictions apply.