Project topics and materials
Get Complete Project Material File(s) Now! »
Use of UAVs to Establish a Network for Smart Agriculture
UAVs have gained a lot of attention in recent years. Their use nowadays is not limited to defense or military applications rather, civilian applications have taken advantage of the advance scored in the defense sector. One can find very successful applications in such areas as forest, ocean, environment, weather monitoring, topography, rescue, safety and farming, etc. Lately, suggestions on the use of UAVs have included airplane inspection. The success of UAVs is due to their versatility. They can be very small, carry a customizable payload, and may not necessarily require takeoff or landing strips. Furthermore, they are becoming very affordable to the point where a group of them can be used as a swarm in a coordinated structure to take on a variety of participatory tasks or serve for redundancy and backup. UAV command and control, condition and capability in terms of self-awareness, situational awareness, self-organization, reconfiguration, and adaptation are well-established concepts. It is worth mentioning that today; a simple UAV in the markets can exhibit many of these capabilities with added features including controllability through wearable computing devices such as smart-phones. UAV command and control interfaces along with trajectory planning options are also available. One can select a group of UAVs on a computer screen, designate a mission, specify the payload, assign a path or a destination, and launch a real-time scenario. In [13], UAVs are used in order to minimize the excessive usage of pesticides and fertilizers in agricultural areas. The process of applying the chemicals is controlled by means of the feedback obtained from the WSN planted in the crop field. The UAV trajectory is adjusted on base of feedback from the sensors. In [14], UAVs are used to serve as a relay network to eliminate the disconnection of parts of a WSN and guarantee the delivery of data. In [15], cooperative Multiple Input Multiple Output (MIMO) techniques are used to support communication among static sensors in a sparse WSN and a relay network composed of UAVs in order to keep the WSN connected. In [16], a customizable virtual environment to display conditions and capabilities of unmanned vehicles is disclosed. The virtual environment is highly customizable that can be used to generate mission scenarios and edit with the help of planning and strategy based on real-time data gathering and processing. UAV technology has given a high-technology makeover to agriculture industry. UAV can be utilized in every phase of crop cycle like:
A. Soil and field analysis: UAV can produce precise 3-D maps for soil analysis that is useful in seeding and planting patterns making, irrigation and nitrogen-level management.
B. Plantation: UAV plantation systems can decrease planting costs by 85 percent. These systems shoot pods with seeds and plant nutrients into the soil, providing the plant all the nutrients necessary to sustain its life.
C. Crop spraying: UAV can spray pesticide to selected areas even selected plant that can save resources, environment and wildlife all together.
D. Irrigation: UAV can identify water deficit areas by using hyperspectral, multispectral and thermal sensors by calculating vegetation index and heat signature. This information is very useful to get quality along quantity of crop and to save water resources.
E. Health assessment: Healthy and sick plants can be identified using visible and near-infrared light inspection. Sensors carried by UAV can differentiate between healthy and bacterial or fungal infected plant as they reflect different amounts of visual and NIR light.
Clustering of Sensor Nodes for Efficient Data Collection
Clustering is one of the important means to prolong the network lifetime of wireless Sensor Networks (WSNs). Clustering mean dividing wireless sensor nodes into virtual groups or arranging them in hierarchical structure and electing one of them as a Cluster Head (CH) for each cluster. All Cluster Members (CM) should send their data to the corresponding CH which forwards the aggregated data to the Base Station (BS). A major challenge in WSN clustering is to select a suitable node as the CH. Advantages of cluster based WSN over flat network are energy efficiency, better network communication, minimized delay, efficient topology management and so forth. Wireless sensor networks are always highly resource constrained having limited power, storage, bandwidth, and computational capabilities. Therefore, in case of energy depletion in sensor nodes, it becomes inoperable and irreplaceable. Increasing network lifetime and sustainability are the key issues for the contemporary research areas in sensor domain. Normally, energy depletion is highly dominated by radio transmission and even more for far-distances. Clustering techniques increase sensor network lifetime by limiting the number and range of radio transmissions. In contrast, performance of the flat network degrades with the growth of network size because increasing the network size and control overhead are directly proportional to each other. We can classify WSN clustering techniques into four categories (see section 1.2.1): Static Sink Static Nodes (SSSN) routing where base station, CH and sensor nodes all assumed to be stationary; the second type is Mobile Sink Static Nodes (MSSN) data collection which is more of our interest because our project lies in this category. One of the famous MSSN technique is rendez-vous based clustering where some data collection centers are established to collect data by a moving vehicle considered as sink node. Another MSSN type of clustering is network assisted clustering where predefined network is scanned by some moving sinks to collect data. In both these techniques WSN assists moving sink node to navigate and collect data. There is also exists UAV assisted routing where UAV assist the network to farm clusters. A general overview of SSSN, MSSN and more specific UAV assisted data collection techniques are given in next section and more detail and technicalities are presented in clustering chapter number 3. The remaining two types of data collection; Static Sink and Mobile Nodes (SSMN) such as cellular system and Mobile Sink and Mobile Node (MSMN) like ad-hoc communication, are not considered in this thesis. As in both cases, sensor nodes are considered as moving and more focus is given to manage the clusters and data routes according to the predicted or monitored mobility pattern of sensor nodes. In our scenario, crop sensors are always static.
Mobile sink routing techniques
In static sink routing, some serious issues can not be prevented, such as: communication overhead one of the leading issue where sensor nodes have to relay whole data towards BS. The second issue is that the nodes near the BS become effected anyhow because these nodes are the only option to approch BS. In the presence of these issues, the performance of a static sink routing can not be enhanced much. The only solultion to overcome these issues is the use of mobile sink nodes. This type of sink nodes are capable to harvest data by visiting the sensor nodes, especially the use of UAV becomes a trend and emerging technology. The simplest example of data gathering with mobile sink is a direct contact. In direct contact, UAV has to visit individually all nodes in the network to collect data [74]. In this scheme, many ways can be adapted to optimize the network traversing such as: square Grid tessellation, triangle tessellation, Snake like traversal, Boundary traversal, Traveling salesman problem, etc. more details can be grabbed from [75] or [76]. As we have already mentioned, clustering is a better choice than the direct contact, therefore we will only focus on UAV supported clustering schemes. We can divide UAV supported clustering in two broad categories according to the UAV control:
1) WSN clustering with controlled sink path. There are three possibilities for controlled sink path a. Fixed/static: In fix/static path, the sink always follows the same path that is known to all ground sensors.
b. Controlled by WNS: In WSN controlled path, the sink has to follow the path instructed by ground sensors to collect the data from predefined CHs.
c. Random. In random path, the sink has to search the CHs in the field to harvest the data where CM nodes can help to find it.
Localization of Field Sensors by UAV
The main function to establish a sensor network is to collect and forward data to destination. It is very important to know about the location of the nodes to collect data in efficient way. This kind of
information can be obtained using localization techniques in WSNs. It is highly desirable to design lowcost, scalable, and efficient localization mechanisms for WSNs. There are many ways to find the location of a node like:
A. GPS Based and GPS Free: In Global Positioning System (GPS) based schemes; localization accuracy is very high but is very expensive as well in terms of cost and resources. Embedding GPS receiver into small size energy sensitive sensors, availability of GPS signal in very remote area, indoor environment and covered area are major problems of GPS based systems. In the scenario under consideration where sensor nodes are not changing its position, GPS receiver is an extra burden on it to monitor the same location every time. An energy efficient (311 Joule tracking and 389 Joule during acquisition) and highly sensitive (-165 dBm), PmodGPS module made by Digilent is shown in sensors (see section 2.6.1) is 756 Joule in this regard, size and power consumption of this GPS module is significant for this sensor.
Table of contents :
List of Figures
List of Table
List of Acronyms
1. Introduction
1.1 Background
1.2 Thesis Contribution
1.2.1 Dynamic clustering
1.2.2 Dynamic cluster head selection
1.2.3 Virtual antenna for localization
1.3 Project Approval
1.4 Thesis Organization
2. State of the Art
2.1 Introduction
2.2 Scope of Agriculture in Saudi Arabia
2.3 Smart Farming
2.4 Deployment of Heterogeneous Sensor Nodes in Crop Field
2.4.1 Bug monitoring and control sensors
2.4.2 Crop health monitoring
2.4.3 Soil parameter monitoring
2.4.4 Monitoring crop health by aerial view using hyperspectral imaging
2.4.5 Issues and challenges in hyperspectral imaging
2.4.6 Categories of crop sensors
2.5 Use of UAVs to Establish a Network for Smart Agriculture
2.6 Clustering of Sensor Nodes for Efficient Data Collection
2.6.1 Static sink routing
2.6.2 Mobile sink routing techniques
2.7 Localization of Field Sensors by UAV
2.7.1 Lateration
2.7.2 Angulation
2.7.3 RSSI based localization technique
2.7.4 Time of arrival
2.7.5 Time difference of arrival
2.7.6 Angle of arrival
2.8 Uniform Linear Antenna for Localization
2.9 Virtual Antenna Array
2.9.1 MUSIC algorithm for localization
2.9.2 ESPRIT algorithm for localization
2.10 Project Worthiness
2.11 Major Challenges
2.12 Suggested Solutions
2.13 Conclusion
3. Clustering
3.1 Introduction
3.2 LEACH
3.3 HEED
3.4 Network Assisted Clustering
3.5 Proposed UAV Assisted Dynamic Clustering
3.5.1 UAV
3.5.2 Sensor node
3.6 Three-Layer Architecture of the Developed System
3.6.1 Discovery
3.6.2 Navigation
3.6.3 Communication
3.6.4 CH selection
3.6.5 Time synchronization
3.7 Developed System Algorithm
3.7.1 Development of the sensor node
3.7.2 UAV functionality
3.8 Link Budget
3.8.1 F1 433 MHz UHF signal
3.8.2 F2 2.4 GHz WiFi signal
3.9 OMNeT ++ Simulation
3.9.1 Sensor node simulation model
3.9.2 UAV simulation model
3.9.3 Simulation cases
3.9.4 Simulation results
3.10 MATLAB Simulation
3.11 Proof of Concept
3.12 Conclusion
4. Localization
4.1 Introduction
4.1.1 Uniform linear array
4.1.2 Multi sources
4.1.3 Multiple signal classification
4.2 Proposed System
4.3 Virtual Antenna Array
4.3.1 Geometrical variation
4.3.2 Example -1, long range localization
4.3.3 Example -2, short range localization
4.3.4 Sampling
4.3.5 Calibration
4.3.6 Rectification
4.3.7 Adjustment of incident angle
4 .4 Virtual Antenna Array in Case of UAV Non-linear Movement
4.5 Simulation Model and Analysis
4.6 Conclusion
5. Hardware Development
5.1 Introduction
5.2 UAV
5.2.1 UAV thrust calculation
5.2.2 UAV development
5.2.3 Firmware installation
5.2.4 Accelerometer calibration
5.2.5 Compass calibration
5.2.6 ESC calibration
5.2.7 Flight mode setting
5.2.8 Mission planning
5.2.9 Trouble shooting
5.3 UAV Connection with Developed Routing Protocol
5.4 UAV Flight Data Monitoring and Feed Back
5.5 Conclusion
6. Conclusion & Future Work
6.1 Conclusion
6.1.1 Concluding remarks
6.2 Future Work
6.2.1 Clustering
6.2.2 Localization
6.2.3 Mission optimization
6.2.4 Hardware development
Appendix A
A.1 LEACH
A.2 HEED
A.3 Network Assisted Data Collection
Appendix B
B.1 AGI STK MATLAB Integration
Appendix C
C.1 OMNet++ Installation
C.2 First Practical Example in OMNeT++
C.3 Extension of This Simulation
C.4 Developed URP
C.4.1 NED file
C.4.2 INI file
C.4.3 Server.cc file
C.4.4 Host.cc file
Appendix D
D.1 Data Collection Simulation
D.2 Localization Simulation
Appendix E
E.1 First Hands on Example with Arduino
Appendix F
Appendix G
Publications
Bibliography
