SHORT PAPER A MOBILE WSN SINK NODE USING UNMANNED AERIAL VEHICLES: DESIGN AND EXPERIMENT A Mobile WSN Sink Node Using Unmanned Aerial Vehicles: Design And Experiment http://dx.doi.org/10.3991/ijim.v10i3.5808 Hui-Ru Cao, Zhi Yang and Ye-Qian Li Sun Yat-sen University, Guangzhou, China Abstract—In this paper, we construct a new type of mobile wireless data sinking platform for data collection based on unmanned aerial vehicle (UAV) technology, which aims to address the increasing demand for wireless sensor network (WSN) distribution in different monitoring areas and en- large the coverage for various application scenarios. A wireless environmental monitoring system is firstly studied, and then wireless communication capacity and data collec- tion experiments are performed. The communication ca- pacity test results show that when the RF modules operate with a transmission power above 1 dBm and a communica- tion distance below 100 m, the UAV wireless sink node can maintain a high quality communication data link. Addi- tionally, an outdoor data collection experiment is per- formed using this UAV platform within a mountainous area. In this outdoor experiment, the data analysis results show that the validity rate of the environmental data that is obtained from the WSN cluster head node on the ground is higher than 92%, and most of the missing data results from WSN communication failures. This experiment proves the feasibility of introducing UAV as a sink node in a clustering WSN. The overall contributions of this paper can provide guidance for building a UAV cooperative WSN system in future. Index Terms—Mobile Wireless Sensor Networks, Un- manned aerial vehicles, Sink platform, Data analysis I. INTRODUCTION Due to advantages of having no wiring, a flexible to- pology and a low cost, WSN has been applied to a wide range of fields including environmental monitoring, intel- ligent transportation, agricultural situation acquisition, and animal tracking and positioning [1-3]. The working WSN nodes are usually stationary as this type of node is restricted by its working environment, and the node mon- itoring range is also largely limited. As WSN is being continuously developed and mobile robots are being added to the network, vehicle engineering technologies such as wireless mobile nodes can further extend the WSN monitoring scope and improve the flexibility of the wireless network layout [4-7]. Relating research to combine Mobile Wireless Sensor Networks with practical requirements has been done both in China and in other countries. Young-Duk Kim et al. [8] have built an environment monitoring system with a mo- bile WSN based on wheeled robots and a TinyOS plat- form, and have proposed a multi-channel Medium Access Control protocol. Kwong et al. [9] have built a mobile WSN platform for monitoring livestock movement char- acteristics of free-range animals and have analyzed the optimization of wireless data transmission performance. The above studies have improved the performance and application scope of this domain, through adding mobile nodes to WSN. However, there are disadvantages that prevent ideal movement tracks from being built as the moving range is small and there is a poor environmental adaptability for mobile nodes. Unmanned Aerial Vehicle (UAV) technology is an ef- fective, high-speed platform that has been widely used in national defense security, agriculture, forestry and animal husbandry, logistics transportation, and other fields [10-11]. Introduction of this technology into wireless net- works is a new hot spot in current MWSN studies. Imad Jawhar et al. [12] have proposed UAV platforms including the linear WSN data framework, which includes an un- manned aerial vehicle, a relay node, a perception and gathering node, and a multi-hop network for data trans- mission. The have also performed simulation tests to verify the feasibility of this framework. Dac-Tu Ho et al. [13] have used the particle swarm algorithm to optimize the UAV WSN track point, based on the hierarchical network protocol Low Energy Adaptive Clustering Hierarchy (LEACH). The network parameters such as the life cycle and flying time are studied in a simulation test. Most of the current existing research on WSN UAV is based mainly on path planning and design of the frame- work, rather than research into combining UAV with wireless networks. In this paper, the ZigBee protocol stack is used based on the characteristics of MWSN and UAV technology, to build a new type of mobile wireless plat- form that can be used for a wide range of monitoring ap- plications and improve the network coverage area. Based on the convergence platform, the UAV platform can be used for actual WSN environment monitoring data net- work tests to verify the feasibility of the UAV gathering platform to provide a basis for UAV wireless networks that can be widely used in future. The structure is organized as follows: Section 2 presents the platform design of an unmanned aerial vehicle. Section 3 presents the experimental materials and methods and tests are given in Section 4. Finally, Section 5 concludes the paper and suggests further work. II. PLATFORM DESIGN OF UNMANNED AERIAL VEHICLE As shown in Fig. 1, this paper proposes a UAV WSN gathering platform that consists of the following compo- nents: an unmanned aerial vehicle body (including a con- troller), a ZigBee module, a GPS module, an image acqui- sition module, a sensor module and a power supply. The flight of the unmanned aerial vehicle module is the main consideration as it carries the overall platform. The wire- 64 http://www.i-jim.org SHORT PAPER A MOBILE WSN SINK NODE USING UNMANNED AERIAL VEHICLES: DESIGN AND EXPERIMENT less module gathers the ground WSN data, and the data is stored in the storage unit. The GPS coordinates within the GPS module acquire the UAV location information. The image acquisition module collects the images during UAV flight. The sensor module acquires the status pa- rameters of UAV during the flight process. III. EXPERIMENTAL METHODS AND MATERIALS A. Experimental materials and equipment The platform built for this research adopts a single screw after push type fixed-wing carrier as shown in Fig. 2. The fuselage materials are foamed polypropylene (EPP) with a wingspan of 172 cm, a captain of 112 cm, and a 40 cm tail plane. The communication module components are: a CC2530 wireless module from TI Company with a carrier frequency of 2.4 GHz, a 3 dBi omni-directional sucker antenna and a Heaven and Earth fly remote control with eight channels. The power system uses the New West Company 2217 - KV1100 40A brushless motor, a brushless electronic governor, a 12 V power dc power supply, and 7g double propellers with a length of 254 mm. The other special test equipment in- cludes an APM2.5 CJMCU - 109 current sensor, an ANTAI faith APS3005D manostat, an ANTAI ADS1112CAL 7-inch screen oscilloscope with a 1G sampling rate, a 110 m ~ 200 m digital oscilloscope, a tripod, a tablet, a tape measure and a load. B. Experimental design A communication network experiment was performed to obtain the static and dynamic communication reliabil- ity of the constructed gathering platform. The parameters were as shown in table 1. The height was below 150 m, the step length was 5m; the transmission power between nodes was 4.5 dBm and 1 dBm, the communication rate was 250 KBPS, the height of the ground node was 1 m and the node omni-directional antenna gain was 3 dBi. Using the communication height as the test factor, 1000 data points were sent in a row for each communication height, and repeated three times and the average taken. The received signal strength index (RSSI) and the aver- age packet loss rate (PLR) of the wireless communication nodes was thus obtained between the gathering platform, when receiving the signal to measure the communication reliability of the platform. After completion of the static experiment, the feasibil- ity and the platform communication reliability during flight was obtained by performing a UAV wireless node network experiment. A cluster of WSNs were formed in a ground fruit farm, and the UAV platform was used three times to wirelessly collect different volumes of data. The success rate parameters were then obtained for the net- work communication platform in order to dynamically evaluate the reliability performance. IV. PRACTICAL COMMUNICATION NETWORK TESTS ON UAV PLATFORM OUTDOOR A. Communication test Fig. 3 is obtained by software analysis, and shows the relationship and fitting curve between different commu- nication distances, RSSI and PLR. Fig. 3 shows that the transmission power of the UAV wireless collection platform is 4.5 dBm and within the communication distance range of 150 m, the node can U A V GPS module ZigBee module Wind sensor Current sensor Power Camera Figure 1. Structural diagram of UAV platform Figure 2. The key geometric parameters of the UAV Figure 3. RSSI PLR in different communication distance TABLE I. UAV WSN PLATFORM PARAMETERS Impact factors Height /m Transmission power / dBm Communication rate / kbps Values 0~150 (/5m) 1, 4.5 250 establish effective communication with the ground when the RSSI is greater than -90 dBm and the PLR is less than 20%. The transmission power is 1 dBm when its corre- sponding RSSI and PLR are less than -95 dBm and 30% respectively. The test results show that the UAV WSN collection platform within the scope of the communication distance is 120 m as long as a transmission power is cho- sen that has greater than 1 dBm link quality in order to meet requirements. In order to ensure reliable communica- tion, the node flight height of all nodes should be chosen to be below 100 m to ensure a high quality of communica- tion. B. Network test Upon completion of the communications test, multiple hops are set up within heterogeneous clusters of WSN within groves, to collect orchard parameters such as tem- perature and humidity, light intensity and wind speed. This wireless network includes a collection node and an ordi- nary node. As shown in Fig. 4, the entire experimental system includes a UAV wireless platform, a ground station and the WSN. The dotted line in the overall flight path iJIM ‒ Volume 10, Issue 3, 2016 65 SHORT PAPER A MOBILE WSN SINK NODE USING UNMANNED AERIAL VEHICLES: DESIGN AND EXPERIMENT shown in Fig. 4 indicates that the UAV gathers from the clearing platform, establishes communication within WSN over the WSN, and completes the whole network data collection. In this experiment, the flight height of UAV is less than 100 m, and the flight time is approxi- mately 30 minutes after reaching WSN UAV, where the airspace and the cluster head nodes establish communica- tion links, cluster the environmental monitoring data, retrieve the cache in the pseudo code to describe the communication process, such as the communication pro- cess adopted error retransmission and interference retreat mechanism in table 2. The altitude change curve is shown in Fig. 5. From Fig. 5 it can be seen that once the desired height is reached, the UAV flies in cruise state. The experi- mental data analysis shows that the height control of the variance is 7.3 and the mean square error is 2.7, indicat- ing that the UAV meets design requirements when the flight control system has a flying height error below 3 m and a relative altitude error of less than 3%. In the experiment, the fixed-wing UAV is controlled by the manipulators and the UAV climbs to 100 m after 96 s. At this point, the onboard flight control system (APM) 2.5 and the WSN data acquisition subsystem are launched, the UAV is flying along its designated route while constantly attempting to contact the ground WSN nodes at the same time. There is a high risk for fixed-wing UAV controlled landing. In this experiment, after the data acquisition is completed, the UAV is re- turned to the landing field by the flight control system of the flight control hand remote for closing the heading, height, the throttle adjustment function, and keeping the balance function, drones in the manipulators, flight con- trol using a coordinated control method for smooth land- ing to reduce the difficulty of control. Fig. 6 shows the UAV wireless converged network packet acquisition test platform. The acquisition success Fly path Node Sink node Land station Figure 4. Outdoor practical tests Figure 5. Flight level curves of UAV rate of the three packets is more than 92%, meeting appli- cation requirements. The actual number of packets is less than theoretical predicted which has been analyzed and is caused by packet loss due to WSN communication inter- ference, as the lost packets have no successful transmission from the WSN child nodes to the cluster head nodes. C. Environmental monitoring data analysis Regional data (temperature, humidity, light intensity, wind speed) is monitored by the UAV for six hours be- tween 13:30 to 19:30 and the results of the analysis are shown in Fig. 7 and 8. TABLE II. THE COMMUNICATION PROCESS OF UAV WITH GROUND WSN PLATFORM Receive Data From WSN() 1: 2: 3: 4: 5: 6: 7: 8: 9: 10: 11: Try broadcasting UAV beacon packets if beacon packets were ACKed for 3 then Start querying data from ground WSN sink node with RTS CTS and ACK strat- egy if CRC error then Re-transmit current packet if CRC error > 5 then Wait for 10s and retry else Prepare for next packet if no ACK from ground WSN node for 10 rounds then Restart from sending UAV beacons Figure 6. Data acquisition test Figure 7. Temperature and humidity change curves of the monitoring area 66 http://www.i-jim.org SHORT PAPER A MOBILE WSN SINK NODE USING UNMANNED AERIAL VEHICLES: DESIGN AND EXPERIMENT Figure 8. Monitoring area - light intensity and wind speed curve Fig. 7 shows the temperature changes, which displays a downward trend after first rising, and peaks appear at 15:00. At the same time, the humidity shows a change trend that is in contrast to the temperature change. Figure 8 shows that the light intensity curve has an overall downward trend, with an obvious correlation to the tem- perature changes. The overall change in the monitoring period wind speed curve is smooth with no obvious change. V. CONCLUSION AND DISCUSSION In areas without GPRS or 3G mobile communication network coverage, UAV can be distributed through dif- ferent locations of a network to monitor data. Therefore a UAV-WSN platform can effectively expand the monitor- ing area and its scope. This contribution of this paper is to build a wireless collection platform based on UAV. The platform was used to test tension, rotational speed and the wind field. It can be concluded that for data collection and analysis, this platform provides necessary support for a wide range of future applications. At the same time, the platform communication network performance is tested, the static and dynamic reliability of the communication using the platform is verified. Finally, a practical applica- tion test is performed using an MWSN which illustrates the feasibility of the platform. Due to experiment’s environmental limit, this experi- ment did not test a wide range of network authentication and other environmental factors. Future research can be performed according to the specific application of WSN UAVs, and determine the UAV combination of WSN in flight and more precise mathematical models by con- structing different types of UAV and different network modes. ACKNOWLEDGEMENT This work is supported by the colleagues in the De- partment of Electronic Communication & Software En- gineering Nanfang College of Sun Yat-sen University. REFERENCES [1] J. Yick, B. Mukherjee, and D. Ghosal, “Wireless sensor network survey,” Computer Networks, vol. 52, issue 12, pp.2292-2330, 2008. http://dx.doi.org/10.1016/j.comnet.2008.04.002 [2] M. Chen, J. Wan, S González, X. Liao, and V. C. M. Leung, “A Survey of Recent Developments in Home M2M Networks,” IEEE Communications Surveys and Tutorials, vol.16, issue 1, pp. 98-114 2014. http://dx.doi.org/10.1109/SURV.2013.110113.00249 [3] J. Wan, C. Zou, K. Zhou, D. Lu, and D. Li, “IoT Sensing Frame- work with Inter-cloud Computing Capability in Vehicular Net- working,” Electronic Commerce Research, vol. 14, issue 3, pp. 389-416, 2014. http://dx.doi.org/10.1007/s10660-014-9147-2 [4] X. Li, D. Li, J. Wan, A.V. Vasilakos, C. Lai, and S. Wang, “A review of industrial wireless networks in the context of Industry 4.0,” DOI:10.1007/s11276-015-1133-7. http://dx.doi.org/10.1007/s11276-015-1133-7 [5] A.M. Hossein, A.S. Gaddafi, and A.A. Hanan, “A survey of wire- less sensor network approaches and their energy consumption for monitoring farm fields in precision agriculture,” Precision Agri- culture, vol. 16, issue 2, pp.216-238, 2015. [6] M. Chen, S. Gonzalez, and V.C.M. Leung, “Applications and design issues for mobile agents in wireless sensor networks,” IEEE Wireless Communications, vol. 14, issue 6, pp.20-26, 2007. http://dx.doi.org/10.1109/mwc.2007.4407223 [7] Y. Alayev, F. Chen, Y. Hou, M.P. Johnson, A. Bar-Noy, T. F. La Porta, and K.K. Leung, “ Throughput maximization in mobile WSN scheduling with power control and rate selection,” IEEE Transactions on Wireless Communications, vol. 13, issue 7, pp. 4066-4079, 2014. http://dx.doi.org/10.1109/TWC.2014.2315196 [8] Y.D. Kim, Y.M. Yang, W.S. Kang, and D.K. Kim, “On the design of beacon based wireless sensor network for agricultural emergen- cy monitoring systems,” Computer Standards & Interfaces, vol. 36, issue 2, pp. 288-99, 2011. [9] K.H. Kwong, T.T. Wu, H.G. Goh, K. Sasloglou, B. Stephen, I. Glover, C. Shen, W. Du, C. Michie, and I. Andonovic, “Imple- mentation of herd management systems with wireless sensor net- works,” IET Wireless Sensor Systems, vol. 1, issue 2, pp. 55-65, 2011. http://dx.doi.org/10.1049/iet-wss.2010.0057 [10] H. Adam, T. Michael, and S. Inna, “Multi-Camera tracking and mapping for unmanned Aerial vehicles in unstructured environ- ments,” Journal of Intelligent and Robotic Systems, vol. 78, issue 2, pp. 291-317, 2015. [11] B. Bollard-Breen, D. J. Brooks, and R. L. Matthew, “Application of an unmanned aerial vehicle in spatial mapping of terrestrial bi- ology and human disturbance in the McMurdo Dry Valleys, East Antarctica,” Polar Biology, vol.38, issue 4, pp.573-578, 2015. http://dx.doi.org/10.1007/s00300-014-1586-7 [12] I. Jawhar, N. Mohamed, and J. Al-Jaroodi, “A Framework for using unmanned aerial vehicles for data collection in linear wire- less sensor networks,” Journal of Intelligent & Robotic Systems, vol.74, issue 1/2, pp. 437-453, 2013. [13] H. Dac-Tu, G. E. Ingar, and P.B. Sujit, “Optimization of wireless aensor network and UAV data acquisition,” Journal of Intelligent and Robotic Systems, vol.78, issue 1, pp.159-179, 2015. AUTHORS Hui-Ru Cao is with Nanfang College of Sun Yat-sen University, Guangzhou, Guangdong, 510970, China (e-mail: xiaocao0924@163.com). Yang Zhi is now with Nanfang College of Sun Yat-sen University, Guangzhou, Guangdong, 510970, China (e-mail: issyz@mail.sysu.edu.cn). Ye-Qian Li is with Nanfang College of Sun Yat-sen University, Guangzhou, Guangdong, 510970, China (e-mail: liyq@mail.nfu.edu.cn). This work was supported in part by the Youth Innovation Project of Important Program for college of Guangdong province, China, in 2015 with the No. 2015KQNCX228 and Technology Innovation Project of Guangdong Province with the No. 2016-18. Submitted, 02 May 2016. Published as resubmitted by the authors on 10 June 2016. ! iJIM ‒ Volume 10, Issue 3, 2016 67 iJIM – Vol. 10, No. 3, 2016 A Mobile WSN Sink Node Using Unmanned Aerial Vehicles: Design And Experiment