Development and Testing of a Two-UAV Communication Relay ... - MDPI

sensors Article

Development and Testing of a Two-UAV Communication Relay System Boyang Li, Yifan Jiang, Jingxuan Sun, Lingfeng Cai and Chih-Yung Wen * Department of Mechanical Engineering, The Hong Kong Polytechnic University, Hong Kong, China; [email protected] (B.L.); [email protected] (Y.J.); [email protected] (J.S.); [email protected] (L.C.) * Correspondence: [email protected]; Tel.: +852-2766-6644 Academic Editors: Felipe Gonzalez Toro and Antonios Tsourdos Received: 22 July 2016; Accepted: 10 October 2016; Published: 13 October 2016

Abstract: In the development of beyond-line-of-sight (BLOS) Unmanned Aerial Vehicle (UAV) systems, communication between the UAVs and the ground control station (GCS) is of critical importance. The commonly used economical wireless modules are restricted by the short communication range and are easily blocked by obstacles. The use of a communication relay system provides a practical way to solve these problems, improving the performance of UAV communication in BLOS and cross-obstacle operations. In this study, a communication relay system, in which a quadrotor was used to relay radio communication for another quadrotor was developed and tested. First, the UAVs used as the airborne platform were constructed, and the hardware for the communication relay system was selected and built up. Second, a set of software programs and protocol for autonomous mission control, communication relay control, and ground control were developed. Finally, the system was fully integrated into the airborne platform and tested both indoor and in-flight. The Received Signal Strength Indication (RSSI) and noise value in two typical application scenarios were recorded. The test results demonstrated the ability of this system to extend the communication range and build communication over obstacles. This system also shows the feasibility to coordinate multiple UAVs’ communication with the same relay structure. Keywords: two-UAV system; communication relay; mission control software; Ground Control Station

1. Introduction In recent years, as a result of decreased weights and sizes, reduced costs, and increased functionalities of different sensors and components, Unmanned Aerial Vehicles (UAVs) are becoming increasingly popular in a wide range of domestic applications (photography, surveillance, environment monitoring, search-and-rescue, etc.) [1–3]. As the market develops and the variety of applications expands, more requirements are now being imposed on extending the flight range and increasing the adaptability to the complex missions of UAV systems. In missions such are surveillance and search-and-rescue, UAV systems are usually required to operate in complicated environments while still being able to maintain uninterrupted communication with the ground control station for the purpose of reporting, monitoring, and control. For example, in the mountainous or high-density urban regions, the communication between UAVs and the ground control station can be easily interrupted, causing the ground control station to lose real-time data feedback from the UAVs, leading to mission failure. In addition, if the mission requires long flight range, the communication might also be interrupted by the distance. Commercially available yet economical UAV communication solutions, however, have a number of limitations that potentially hinder their widespread application. First, the communication range of major civil UAV communication solutions is comparatively short. Telemetry modules, such as 3DR Sensors 2016, 16, 1696; doi:10.3390/s16101696

www.mdpi.com/journal/sensors

Sensors Sensors 2016, 2016, 16, 16, 1696 1696

2 of 21 20

range of major civil UAV communication solutions is comparatively short. Telemetry modules, such Sik2, XBee, andXBee, other and Wi-Fiother modules, are common options for civil UAVs,for have communication as 3DR Sik2, Wi-Fiwhich modules, which are common options civil UAVs, have ranges that are limited to a few kilometers. Second, it is difficult to establish a stable data link when communication ranges that are limited to a few kilometers. Second, it is difficult to establish a stable obstacles trees, tallsuch buildings, mountains separate the UAVs and GSC. data link such whenasobstacles as trees,ortall buildings, or mountains separate the UAVs and GSC. In situations discussed above, it can be difficult to complete the mission with with only only one one UAV UAV In situations discussed above, it can be difficult to complete the mission deployed, while keeping cost and system complexity requirements met [4–7]. To harness deployed, while keeping cost and system complexity requirements met [4–7]. To harness the the advantages of UAVs in these situations and minimize the drawbacks at the same time, an alternative advantages of UAVs in these situations and minimize the drawbacks at the same time, an alternative solution solution is is to to deploy deploy aa multi-UAV multi-UAV system, system, which which could could utilize utilize the the inter-connectivity inter-connectivity among among multiple multiple UAVs every UAV UAV and the ground ground control control station. station. UAVs to to maintain maintain uninterrupted uninterrupted communication communication between between every and the In this study, a UAV communication relay solution was developed, which uses relay and routing In this study, a UAV communication relay solution was developed, which uses relay and routing to to extend extend the the communication communication range range and and bypass bypass obstacles obstacles at at low low cost. cost. The The objective objective was was to to develop develop and and test ability to to relay relay radio radio communication. communication. test aa proof-of-concept proof-of-concept two-UAV two-UAV system system that that demonstrated demonstrated the the ability This system only consisted of two UAVs, but its design could accommodate the addition This system only consisted of two UAVs, but its design could accommodate the addition of of more more UAVs. The system uses one UAV as communication relay point and enables the other UAV to UAVs. The system uses one UAV as communication relay point and enables the other UAV to operate operate in in areas areas where where direct direct communication communication with with ground ground control control cannot cannot be be established. established. This This configuration configuration allows communication to be established across obstacles or over a distance exceeding allows communication to be established across obstacles or over a distance exceeding the the range range of of the the onboard onboard radio radio transceiver, transceiver, both both of of which which will will be be demonstrated demonstrated in in field field tests tests later. later. The The typical typical application application scenario scenario of of this this system system is is shown shown in in Figure Figure 11 schematically. schematically.

Figure 1. Application scenarios of UAV communication relay relay system. system. UAV communication

The paper is organized as follows. Section 2 introduced the related works. Section 3 introduces The paper is organized as follows. Section 2 introduced the related works. Section 3 introduces the hardware configurations of the multiple UAVs communication relay system. Section 4 discusses the hardware configurations of the multiple UAVs communication relay system. Section 4 discusses the software aspects of the system. In Section 5, both the indoor and outdoor tests methods and results the software aspects of the system. In Section 5, both the indoor and outdoor tests methods and results are presented. A conclusion of the work is given in Section 6. are presented. A conclusion of the work is given in Section 6. 2. Related 2. RelatedWorks Works Numerous researches researches have have been been performed performed in in UAV UAV communication communication relay relay scheme. scheme. In In the the Numerous conceptualstudy studyof ofUAV UAVbased based military communication relays, Pinkney et[8] al. described [8] described a multiconceptual military communication relays, Pinkney et al. a multi-year year airborne communication range extension program plan which was primarily used to support airborne communication range extension program plan which was primarily used to support the the Battlefield Information Transmission System. main objective of program the program to provide Battlefield Information Transmission System. TheThe main objective of the waswas to provide the the beyond line of sight (BLOS) communications capabilities without the help of scarce satellite beyond line of sight (BLOS) communications capabilities without the help of scarce satellite resources. resources. range of can 50for to UAVs 100 miles for UAVs of interest. et al. Their rangeTheir of coverage cancoverage vary from 50vary to 100from miles of interest. Rangel et al. [9]Rangel developed a [9] developed a multipurpose aerial platform using a UAV for tactical surveillance of maritime multipurpose aerial platform using a UAV for tactical surveillance of maritime vessels. They concluded vessels. They concluded relay that the communication relay a UAV system which is a low-cost that the communication using a UAV system is a using low-cost application could application provide air which could provide air dominance with satisfied communication range with lost In dominance with satisfied communication range with lost cost resources. In Orfanus cost et al.resources. [10], the use Orfanus et al. [10], theparadigm use of thetoself-organizing tonetworks design efficient UAVmilitary relay networks to of the self-organizing design efficientparadigm UAV relay to support operations support military operations was proposed. was proposed. Abundant previous thethe algorithms forfor vehicle position assignment and Abundant previous studies studieshave havefocused focusedonon algorithms vehicle position assignment network configuration. Choi et al. [11] developed guidance laws to optimize the position of a Microand network configuration. Choi et al. [11] developed guidance laws to optimize the position of a UAV which relayed communication and video signals when the other UAV is out of radio contact range with the base. Zhan et al. [12] investigated a communication system in which UAVs were used

Sensors 2016, 16, 1696

3 of 21

Micro-UAV which relayed communication and video signals when the other UAV is out of radio contact range with the base. Zhan et al. [12] investigated a communication system in which UAVs were used as relays between ground terminals and a network base station. Their work mainly focused on developing the algorithm for optimizing the performance of the relay system. In [13,14], Cetin and coworkers proposed a novel dynamic approach to the relay chain concept to maintain communication between vehicles in the long-range communication relay infrastructure. Their path planning method was based on the artificial potential field. Many other researchers used simulation methods to propose and test the relay systems. Hauert et al. [15] conducted 2D simulations of the design on a swarm system which does not make use of positioning information or stigmergy for application in aerial communication relay. Jiang and Swindlehurst [16] simulated the dynamic UAV relay positioning for multiple mobile ground nodes and multi-antenna UAVs. In [17], a relay network composed of UAVs keeping the Wireless Sensor Network connected was proposed. They also used cooperative multiple input multiple output (MIMO) techniques to support communication. Ben Moussa et al. [18] provided a simulation-based comparison between two strategies for an autonomous mobile aerial node communication relay between two mobile BLOS ground nodes. Lee et al. [19] proposed an airborne relay-based regional positioning system to address pseudolite system limitations. They conducted simulations and demonstrated that the proposed system guarantees a higher accuracy than an airborne-based pseudolite system. For experimental studies, most of the previous works were based on IEEE 802.11b (Wi-Fi) standard. Brown et al. [20,21] implemented a wireless mobile ad hoc network with radio nodes mounted at fixed sites, ground vehicles, and UAVs. One of their scenarios was to extend small UAVs’ operational scope and range. They measured detailed data on network throughput, delay, range, and connectivity under different operating regimes and showed that the UAV has longer range and better connectivity with the ground nodes. Jimenez-Pacheco et al. [22] also built a mobile ad hoc network (MANET) of light UAVs. The system has implemented a line of sight dynamic routing in the network. Morgenthaler et al. [23] described the implementation and characterization of a mobile ad hoc network for light flying robots. Their UAVNet prototype was able to autonomously interconnect two end systems by the airborne relay. Asadpour et al. [24,25] conducted simulation and experimental study of image relay transmission for UAVs. Yanmaz et al. [26] proposed a simple antenna extension to 802.11 devices for aerial UAV nodes. They tested the performance of their system and showed that 12 Mbps communication can be achieved at around 300 m. Yanmaz et al. [27] also did a follow up work on a two-UAV network using 802.11a. In this work, they showed that high throughout can be achieved using two-hop communications. Johansen et al. [28] described the field experiment with a fixed wing UAV acted as a wireless relay for Underwater Vehicle. They only tested the download data rates. Notably, although the Wi-Fi network is convenient to build multiple access points, its communication range is highly restricted and a large number of nodes are needed for long range applications. One of the advantages of a Wi-Fi network is its high throughput. However, for UAV telemetry communication, the distance is a more important requirement than the transmission speed. For experimental works without the use of Wi-Fi network, Guo et al. [29] analyzed the use of small UAVs for 3G cellular network relay. They tested the upload and download throughputs and the ping time in both urban and rural environments. Deng et al. [30] developed the UAVs and the communication relay system for power line inspection. Their system consisted of multi-platform UAVs and multi-model communication system for image/video transmission. However, they did not test the communication range or quality of their system. The only one work used 433/915 MHz as the communication frequency was Willink et al. [31] who measured the low-altitude air-to-ground channel at 915 MHz. They mainly focused on studying the spatial diversity to and from the airborne node and also measured time delay and received power. For all above-mentioned works, The UAVs in their systems did not exchange messages with the GCS by the relayed communication, which means the UAVs were used as carriers to realize communication relay for other systems rather than acting as the active parts in the relay systems. Therefore, the applications would be constrained if UAVs lost connection with the GCS. The highlights

Sensors 2016, 16, 1696

4 of 21

Sensors 1696 4 of 20 of our 2016, work16,are the building of an interface and protocol with the flight controller and direct monitoring and control of the UAVs with relayed communication. This study provides a viable and economic and economic way, with detailed hardware and software methods described, to construct a prototype way, with detailed hardware and software methods described, to construct a prototype of the UAV of the UAV communication relay system. communication relay system.

3. Hardware 3. HardwarePlatform Platform In this two-UAV system is introduced. A In this section, section, the thehardware hardwareconstruction constructionofofthe theminiature miniature two-UAV system is introduced. miniature UAV relay system with the ability to relay radio communications was built. The hardware A miniature UAV relay system with the ability to relay radio communications was built. The hardware mainly consisted communication nodes: mainly consisted of of three three communication nodes: aa GCS; GCS; aa UAV UAV that that relays relays radio radio communication, communication, which was named “Mom”; and a second UAV named “Son”, whose communication with the the GCS GCS which was named “Mom”; and a second UAV named “Son”, whose communication with was relayed relayed through through the the UAV UAV “Mom”. “Mom”. The The schematic schematic of of the the major major hardware hardware platform platform for for the the “Mom “Mom was and Son” relay system is illustrated in Figure 2 and the detailed communication subsystem will be and Son” relay system is illustrated in Figure 2 and the detailed communication subsystem will be introduced in Section 3.3 later. introduced in Section 3.3 later.

Figure 2. Whole architecture of the communication relay system.

3.1. System Architecture The relay system can be conceptualized as the following major functional blocks, with the Figure 3. 3. The corresponding hardware architecture shown in Figure The figure mainly describes the main UAV “Mom” “Mom” while while all all other other UAVs UAVshave havethe thesame samehardware hardwarearchitecture. architecture. components in UAV

•

 •

 •  •

 •  •

Aircraft: This This isisthe thebody body UAV, usually consisting of structural components, power Aircraft: of of thethe UAV, usually consisting of structural components, power system, system, andInbattery. the test system,were the constructed aircraft were constructed usingavailable commercially and battery. the test In system, the aircraft using commercially frame available frame kits. Both of the UAVs, i.e., “Mom” and “Son”, were selected as quadrotors for kits. Both of the UAVs, i.e., “Mom” and “Son”, were selected as quadrotors for easily launch and easily launch and recovery ability during the development period. Communication relay recovery ability during the development period. Communication relay systems, however, are not systems,tohowever, areand notcan limited to quadrotors anddesired can be airborne integrated into any desired airborne limited quadrotors be integrated into any platform such as fixed wing platform such as fixed wing UAV for long endurance and distance relay service. UAV for long endurance and distance relay service. Flight Controller: This includes main flight control boards as well as the necessary sensors, that Flight Controller: This includes main flight control boards as well as the necessary sensors, together provide attitude and position control for the UAVs. In this test system, 3DR Pixhawk that together provide attitude and position control for the UAVs. In this test system, 3DR Pixhawk kit [32] was selected as the main flight controller. The controller has an internal IMU, barometer, kit [32] was selected as the main flight controller. The controller has an internal IMU, barometer, external GPS, and a compass for attitude and position measure. It also includes safety switch, external GPS, and a compass for attitude and position measure. It also includes safety switch, buzzer, LEDs, and different kinds of interfaces combined to provide flight control. buzzer, LEDs, and different kinds of interfaces combined to provide flight control. Communication Module: This functional block directly handles wireless radio Communication Module: This functional block directly handles wireless radio communications, communications, and consists of radio transceivers and antennas. The paired modules will and consists of radio transceivers and antennas. The paired modules will connect automatically connect automatically and transmit data instructed by the mission controller. and transmit data instructed by the mission controller. Mission Controller: This functional block is a microcomputer running Linux operation system Mission Controller: This functional block is a microcomputer running Linux operation system and mainly responsible for the following functions: (1) conducting communication flow control, and mainly responsible for the following functions: (1) conducting communication flow control, message routing, and message error checking; (2) monitoring UAV flight conditions and message routing, and message error checking; (2) monitoring UAV flight conditions and reporting reporting to GCS; and (3) translating mission commands from GCS to low-level instructions, to GCS; and (3) translating mission commands from GCS to low-level instructions, which are which are directly supplied to the flight controller. directly supplied to the flight controller. Ground Control Station: This functional block refers to a Windows-based laptop running the Ground Control Station: This functional a Windows-based laptop GCS software to send commands to all theblock UAVrefers nodesto and monitor the status of allrunning UAVs. the GCS software to sendController: commandsThe to all the UAV nodes and monitor the of allcontroller UAVs. via Emergency Remote remote controller is connected to status the flight Emergency Remote Controller: The remote controller is connected to the flight controller 2.4 GHz wireless communication. In normal operation, all the commands and status can be set via 2.4 GHz wireless communication. In normal operation, all the commands and status can by the GSC point. This device has switch to directly takeover the control of UAVs in case be of emergency situation in the debugging process.

Sensors 2016, 16, 1696

5 of 21

set by the GSC point. This device has switch to directly takeover the control of UAVs in case of emergency situation in the debugging process. Sensors 2016, 16, 1696

5 of 20

Figure 3. 3. System architecture. Figure System hardware hardware architecture.

3.2. Communication Communication Subsystem Subsystem 3.2. To construct construct communication subsystem, a ofnumber of used commonly used wireless To the the communication subsystem, a number commonly wireless communication communication technologies were considered as the possible candidate: IEEE 802.11n Wi-Fi, 3G/4G-LTE technologies were considered as the possible candidate: IEEE 802.11n Wi-Fi, 3G/4G-LTE mobile mobile network, and 433 MHz/915 MHz radio. Wi-Fi can provide a reliable, high-speed connection, network, and 433 MHz/915 MHz radio. Wi-Fi can provide a reliable, high-speed connection, but but has a comparatively small range. Typical Wi-Fi routers haveananeffective effectiverange rangeof offewer fewer than than 100 100 m, has a comparatively small range. Typical Wi-Fi routers have m, rendering them unsuitable for long-range outdoor missions. The applicability of the 3G/4G-LTE rendering them unsuitable for long-range outdoor missions. The applicability of the 3G/4G-LTE mobile network network relies relies heavily heavily on on the the strength strength of of mobile mobile phone phone signal, signal, which which varies varies greatly in different mobile greatly in different locations, and and is is not not always always available available in in remote remote areas areas where where surveillance surveillance and and rescue rescue missions missions are are locations, likely to take place. A 433 MHz/915 MHz radio can be operated without a license in many parts of likely to take place. A 433 MHz/915 MHz radio can be operated without a license in many parts of the world, and can be implemented with easily accessible, low-cost transceivers (for example, 3DR the world, and can be implemented with easily accessible, low-cost transceivers (for example, 3DR Radio Telemetry as discussed below). A range of a few kilometers could be achieved. The radio can Radio Telemetry as discussed below). A range of a few kilometers could be achieved. The radio can also be deployed at any desirable site without the need for supporting infrastructure. These features also be deployed at any desirable site without the need for supporting infrastructure. These features give 433 MHz/915 MHz radio advantages in various outdoor missions. The frequency options, 433 MHz give 433 MHz/915 MHz radio advantages in various outdoor missions. The frequency options, or 915 MHz, can be selected to ensure compliance with local radio regulations. 433 MHz or 915 MHz, can be selected to ensure compliance with local radio regulations. Based on the discussion above, to establish communication between the UAVs and GCS and to Based on the discussion above, to establish communication between the UAVs and GCS and facilitate the communication relay system, 3DR 915 MHz Radio Telemetry (3DR Radio) [33] was to facilitate the communication relay system, 3DR 915 MHz Radio Telemetry (3DR Radio) [33] was selected as our radio transceiver hardware. 3DR Radio, the built-in processor of which runs openselected as our radio transceiver hardware. 3DR Radio, the built-in processor of which runs open-source source SiK firmware [34], has the following useful features that enable communication relay: SiK firmware [34], has the following useful features that enable communication relay:  Built-in time division multiplexing (TDM) support. • Built-in time divisionhopping multiplexing support. Supports frequency spread(TDM) spectrum (FHSS) among a configurable number of channels • Supports frequency hopping spread spectrum (FHSS) among a configurable of channels (configured to 50 channels in this study). Its frequency hopping sequence is number user-configurable. (configured to 50 channels in this study). Its frequency hopping sequence is user-configurable. Figure 4 illustrates the radio communication relay setup. Four 3DR Radios were used, numbered T1-1,Figure T1-2, 4T2-1, and T2-2. Table 1 gives the details of theFour configuration installation of the illustrates the radio communication relay setup. 3DR Radiosand were used, numbered telemetries. T1-1 and and T2-2. T1-2 carried GCS and UAV exclusively, T1-1, T1-2, T2-1, Table 1communication gives the detailsbetween of the configuration and“Mom” installation of the whereas T2-1 andand T2-2 carried communication between and UAV exclusively, “Son”. telemetries. T1-1 T1-2 carried communication betweenUAV GCS“Mom” and UAV “Mom” whereas T2-1 and T2-2 carried communication between UAV “Mom” and UAV “Son”.

Sensors 2016, 16, 1696

6 of 21

Sensors 2016, 16, 1696

6 of 20

Sensors 2016, 16, 1696

6 of 20

Figure4.4.Radio Radio communication communication relay Figure relaysetup. setup. Table 1. Telemetry Installation and Settings.

Table 1. Telemetry Installation and Settings. 3DR Radio Installed at Frequency Hopping Setting 3DR Radio at Frequency Hopping Setting T1-1 Figure Installed hopping sequence 1 4. GCS Radio communication relay setup. T1-2 UAV “Mom” hopping sequence 1 T1-1 GCS hopping sequence 1 UAV “Mom” hopping sequence 2 1. Telemetry Installation and Settings. T1-2 T2-1 Table UAV “Mom” hopping sequence 1 T2-2 UAV “Son” hopping sequence 2 T2-13DR Radio Installed UAV “Mom” hoppingSetting sequence 2 at Frequency Hopping T2-2 UAV “Son” hopping sequence 2 T1-1 GCS hopping sequence 1 A combination of TDM and FHSS techniques was used to ensure that the presence of multiple T1-2 UAV “Mom” hopping sequence 1 radio telemetries did not lead to radio-frequency interference. By default, telemetries Tn-1 and Tn-2 T2-1 UAV “Mom” sequence A combination of TDM and FHSS techniques washopping used to ensure2that the presence of multiple (where n = 1, 2…) formedT2-2 one radioUAV link “Son” between each another. On power-up, Tn-1 and Tn-2 first hopping sequence 2 radiosearched telemetries did not lead tosignals radio-frequency interference. By default, telemetries Tn-1 and for synchronization from each another; these signals were used to synchronize the Tn-2 (where n = 1, 2 . . . ) formed one radio link between each another. On power-up, Tn-1 and Tn-2 internalAclocks of bothoftelemetries. Oncetechniques the clocks were synchronized, the the telemetries started a first combination TDM and FHSS was used to ensure that presencethen of multiple searched for synchronization signals from each another; these were usedTn-1 to synchronize TDM process, in which and Tn-2 alternately transmitted data over radio, and listened radio telemetries did notTn-1 lead to radio-frequency interference. Bysignals default, telemetries and Tn-2 to the internal clocks of both telemetries. Once the clocks were synchronized, the telemetries then started a incoming data when not transmitting. The alternation of a transmission window occurred at afirst fixed (where n = 1, 2…) formed one radio link between each another. On power-up, Tn-1 and Tn-2 frequency, and the time between consecutive alternations was referred to as the turnover time. The searchedinfor synchronization signals from each transmitted another; thesedata signals were usedand to synchronize TDM process, which Tn-1 and Tn-2 alternately over radio, listened tothe incoming selection of the turnover time depends on the time required to transmit messages. On one hand, internal clocks of both telemetries. Once the clocks were synchronized, the telemetries then started athe data when not transmitting. The alternation of a transmission window occurred at a fixed frequency, TDM process, in which Tn-1 and Tn-2 alternately transmitted data over radio, and listened to turnover time should be long enough to allow an adequate number of messages to be transmitted in of and the time between consecutive alternations was referred to as the turnover time. The selection data when not transmitting. Theother alternation a transmission window occurred a fixed a incoming single transmission window. On the hand, of the turnover time should not beattoo long, the turnover time depends on the time required to transmit messages. On one hand, the turnover frequency, and the time between consecutive referred as the turnover The otherwise, the other telemetry would have to alternations wait longerwas before it cantotransmit its owntime. messages, time should beoflong enough time to allow an on adequate number oftransmit messages to be transmitted in selection the turnover depends the time required to messages. On one hand, thea single causing delays in information update, and compromising the synchronization accuracy between transmission window. On the other hand, the turnover time should not be too long, otherwise, the turnover time bethe long enoughtime to allow anbe adequate number messages to be transmitted in telemetries. The should value of turnover could tuned based onof the above-mentioned principles. otherIntelemetry wouldthe have to waittime longer it100 canms, transmit ownshould messages, causing in athis single transmission window. On the before other the turnover time be too long, research, turnover was set tohand, which its ensured stablenot operation ofdelays the otherwise, the other telemetry would have to wait longer before it can transmit its own messages, information update, and compromising the synchronization accuracy between telemetries. The value communication relay function. Values ranging from 90 ms to 140 ms have been tested and could causing delays incould information update, and on compromising the synchronization accuracy of theprovide turnover time beprocess tuned the above-mentioned principles. Inbetween this research, similar results. This isbased illustrated in Figures 5 and 6. telemetries. The value of the turnover time could be tuned based on the above-mentioned principles. the turnover time was set to 100 ms, which ensured stable operation of the communication relay In this research, the turnover time was set to 100 ms, which ensured stable operation of the function. Values ranging 90 ms to 140 ms have been and could similar communication relay from function. Values ranging from 90 ms tested to 140 ms have beenprovide tested and could results. This process illustrated inThis Figures 5 and 6. provideissimilar results. process is illustrated in Figures 5 and 6.

Figure 5. Time division multiplexing (TDM) working principle.

Figure 5. Time division multiplexing (TDM) working principle.

Figure 5. Time division multiplexing (TDM) working principle.

Sensors 2016, 16, 1696 Sensors 2016, 16, 1696 Sensors 2016, 16, 1696

7 of 21 7 of 20

7 of 20

Figure 6. TDM TDM process chart. Figure process flow flow Figure 6. 6. TDM process flow chart. chart.

In addition to the TDM process, the transmission of telemetries Tn-1 and Tn-2 were frequencyIn addition additiontotothethe TDM process, the transmission of telemetries were TDM process, the transmission of telemetries Tn-1 andTn-1 Tn-2 and were Tn-2 frequencyhopped. The telemetries transmitted in one of the 50 sub-channels that were obtained by dividing the frequency-hopped. telemetries transmitted in sub-channels oneofofthe thealternation 50that sub-channels that by were obtained hopped. The telemetries transmitted one ofoccurrence the 50 were dividing the 915 MHz band intoThe 50 equal portions.in On the of aobtained transmission window, by dividing the 915 MHz band into 50 equal portions. On the occurrence of the alternation of 915both MHz band into 50 equal portions. On the occurrence of the alternation of a transmission window, Tn-1 and Tn-2 simultaneously hopped to a new sub-channel, which was determined by a a transmission window, both sequence Tn-1 and shared Tn-2 simultaneously hopped new sub-channel, which wasa both Tn-1 channel and Tn-2 simultaneously hopped to a new was determined by random hopping between Tn-1sub-channel, and Tn-2.toAsa which shown in Table 1, T1-1 and determined by a random channel hopping sequence shared between Tn-1 and Tn-2. As shown random channel hopping sequence shared between Tn-1 and Tn-2. As shown in Table 1, T1-1 and T1-2 were configured to use a different hopping sequence than T2-1 and T2-2. Therefore, although in Table 1, T1-1 and T1-2 were configured to use a different hopping sequence than T2-1 and T1-2 were configured to use a different hopping sequence than T2-1 and T2-2. Therefore, although multiple devices might be transmitting radio waves at the same time, the sub-channels in which they T2-2. Therefore, although multiple devices might be transmitting waves at the same time, multiple devices might be transmitting radioinwaves same time, radio the sub-channels in which they are transmitting are different, as illustrated Figureat 7.the the in which theyasare transmitting are different, as illustrated in Figure 7. are sub-channels transmitting are different, illustrated in Figure 7.

Figure 7. Frequency hopping of telemetries.

3.3. Mission Control SubsystemFigure 7. Frequency hopping of telemetries. The control of key onboard functions: communication relay, flight status monitoring, and 3.3. Mission Control Subsystem Subsystem 3.3. Mission Control mission control, is centrally managed by the mission control subsystem, which UAV “Mom” and The control control key onboard functions: communication relay, flight statusin monitoring, and “Son” both carry onboard. The onboard mission controlrelay, subsystem is mainly charge the The ofofkey onboard functions: communication flight status monitoring, andofmission mission is centrally by thecontrol mission control subsystem, UAV and following functions: control, iscontrol, centrally managed managed by the mission subsystem, which UAVwhich “Mom” and“Mom” “Son” both “Son” both carry onboard. The onboard mission control subsystem is mainly in charge of the carry The onboard mission control is mainly in charge of the following functions:  onboard. Executing radio communication frame subsystem processing and relay control. following functions:  Decoding commands contained in frames, and translating the commands into low-level • Executing radio communication frame processing and relay control. instructions to the UAV’s flight controller.  Executing radio communication frame processing and relay control. • Decoding commands contained in frames, and translating the commands into low-level  Monitoring UAV flight status, andinreporting the GCS on a regular  Decoding commands contained frames,toand translating the basis. commands into low-level instructions to the UAV’s flight controller.  instructions Handling ad hocUAV’s functions thatcontroller. specific missions necessitate, for example, image processing, to the flight • Monitoring flight status, and reporting the GCS on a regular basis. dynamics UAV flight smart target to identification,  Monitoring UAV route flightplanning, status, and reporting to the GCS onetc. a regular basis. • Handling ad hoc functions that specific missions necessitate, for example, image processing,  Handling ad hoc functions that specific missions necessitate, for example, image processing, dynamics flight route planning, smart target identification, etc. dynamics flight route planning, smart target identification, etc.

Sensors 2016, 16, 1696

8 of 21

Sensors 2016, 16, 1696

8 of 20

The above-mentioned UAVs to to be be equipped equippedwith withananon-board on-board computer The above-mentionedfunctions functions require require the the UAVs computer Sensors 2016, 16, 1696 8 of 20 with adequate processing power. In this research, Raspberry Pi 2 Module B (Rpi 2B) was chosen with adequate processing power. In this research, Raspberry Pi 2 Module B (Rpi 2B) was chosenasasthe onboard computer for both UAVs because of its small size,size, strong processing power, and the onboard computer for both UAVs because ofrelatively its relatively small strong processing power, The above-mentioned functions require the UAVs to be equipped with an on-board computer abundant IO port. Figure 8 shows the on-board computer module, with Rpi 2B mounted on top. and abundant IOprocessing port. Figure 8 shows theresearch, on-boardRaspberry computerPi module, with Rpi 2B on top. with adequate power. In this 2 Module B (Rpi 2B)mounted was chosen as the onboard computer for both UAVs because of its relatively small size, strong processing power, and abundant IO port. Figure 8 shows the on-board computer module, with Rpi 2B mounted on top.

Figure 8. Onboard computer module. Figure 8. Onboard computer module. 8. Onboardexchange computer module. The onboard computer needsFigure to frequently data with and acquire information from The onboard computer needs to frequently exchange data and acquire information from other onboard devices: radio telemetries, the flight controller, the with GPS module, and various onboard The onboard computer needs to frequently exchange data with and acquire information from other onboard devices: computer radio telemetries, the flight the devices GPS module, and various onboard sensors. The onboard communicates with controller, the peripheral through USB ports. Rpi 2B other onboard devices: radio telemetries, the flight controller, the GPS module, and various onboard sensors. The onboard computer communicates with the peripheral devices through USB ports. Rpi readily supports USB protocol and has four USB ports available. Most peripheral devices, however,2B sensors. The onboard computer communicates with the peripheral devices through USB ports. Rpi 2B readily USB protocol andmethod has fournamed USB ports available. Most peripheral devices, however, utilizesupports a low-level communication Universal Asynchronous Receiver and Transmitter readily supports USB protocol and has four USB ports available. Most peripheral devices, however, utilize a low-level communication method named Receiver and Transmitter (UART). Therefore, a protocol conversion moduleUniversal based on Asynchronous CP2102 microchip was wired between utilize a low-level communication method named Universal Asynchronous Receiver and Transmitter (UART). Therefore, a protocol conversion module based on CP2102 microchip was wired between the USB ports of Rpi 2B and the peripheral devices, which handles the conversion between the USB (UART). Therefore, a protocol conversion module based on CP2102 microchip was wired between protocol and the UART protocol. Figure 9 shows the hardware connection diagrams of the onboard the USB ports of Rpi 2B2B and the peripheral handlesthe theconversion conversion between the USB ports of Rpi and the peripheraldevices, devices, which which handles between the the USBUSB computer module with other subsystems. The Rpi 2B of “Son” was connected to two USB devices protocol and the UART protocol. Figure 9 shows the hardware connection diagrams of the onboard protocol and the UART protocol. Figure 9 shows the hardware connection diagrams of the onboard which are T2-2 and theother flight controller. The Rpi 2B2B of of “Mom” connected to USB devices computer module with subsystems. The Rpi 2B of“Son” “Son”was was connected to two USB devices computer module with other subsystems. The Rpi was connected to three two USB devices including T1-2, T2-1, and the flight controller. which are are T2-2T2-2 andand thethe flight controller. “Mom”was wasconnected connected three USB devices which flight controller.The TheRpi Rpi 2B 2B of “Mom” to to three USB devices including T1-2, T2-1, and flightcontroller. controller. including T1-2, T2-1, and thethe flight

(a) (a)

(b) (b)

Figure 9. Hardware connection diagram of (a) UAV “Son” and (b) UAV “Mom”. Figure 9. Hardware connection diagram of (a) UAV “Son” and (b) UAV “Mom”. Figure 9. Hardware connection diagram of (a) UAV “Son” and (b) UAV “Mom”.

The onboard computer module is powered by the UAV’s main battery and is switched on when The onboard computer module is powered by the UAV’s main battery and is switched on when theThe UAV is powered up. After power-up, the onboard computer module then automatically executes onboard computer module is powered by thecomputer UAV’s main battery is switched on when the UAV is powered up. After power-up, the onboard module then and automatically executes a self-initialization program, which configures necessary system parameters and starts up a set of the UAV is powered up. After power-up, the onboard computer thenand automatically executes a self-initialization program, which configures necessary systemmodule parameters starts up a set of

Sensors Sensors 2016, 2016, 16, 16, 1696 1696

99 of of 21 20

software programs. The programs then run on the onboard computer module and execute functions a self-initialization program, which configures necessary system parameters and starts up a set of such as communication relay control and mission control. Details about software structure and software programs. The programs then run on the onboard computer module and execute functions realization of the communication relay subsystem will be discussed in Section 4. such as communication relay control and mission control. Details about software structure and realization of the communication relay subsystem will be discussed in Section 4. 4. Software Development 4. Software Development The software developed in this project includes mission control software running on the mission control and developed Mini GCS running on theincludes GCS laptop. Thecontrol missionsoftware control running softwareon is in of Theboard software in this project mission thecharge mission analyzing the target node as well as the command contents of each message it receives. If the ID of control board and Mini GCS running on the GCS laptop. The mission control software is in charge of the messages represents nodes, the software bypass this message the analyzing the target node as other well asUAV the command contents of eachwill message it receives. If the IDtoof the corresponding targets. If the ID of the message corresponds to thethis receiving UAV itself, the software messages represents other UAV nodes, the software will bypass message to the corresponding will further resolve the command data in the message and send it to the flight controller. The Mini targets. If the ID of the message corresponds to the receiving UAV itself, the software will further GCS was developed better theand multiple UAV system. condition of Mini each UAV will resolve the commandtodata incoordinate the message send it to the flightThe controller. The GCS was be displayed on the screen including the flight mode, the GPS position, the attitude etc. The Mini developed to better coordinate the multiple UAV system. The condition of each UAV will be displayed GCS was also including used to send to each UAV node bythe theattitude user friendly Graphic User was Interface on the screen thecommands flight mode, the GPS position, etc. The Mini GCS also (GUI). used to send commands to each UAV node by the user friendly Graphic User Interface (GUI).

4.1. Data for Communication Communication Relay Relay 4.1. Data Framing Framing for To facilitate facilitate the the communication communication relay, relay, the the messages messages and and data data to to be be exchanged To exchanged between between the the UAVs UAVs and GCS were grouped into frames, the format of which is shown in Figure 10. Each communication and GCS were grouped into frames, the format of which is shown in Figure 10. Each communication node (“Mom”, (“Mom”, “Son”, “Son”, and and GCS) GCS) was was given node given aa unique unique ID. ID. The The header header of of the the frame frame contained contained two two fields, fields, “Source ID” and “Target ID”. “Source ID” indicated the node from which the frame was transmitted, “Source ID” and “Target ID”. “Source ID” indicated the node from which the frame was transmitted, and “Target “Target ID” ID” indicated indicated the the target target receiver receiver of of this this frame. frame. This and This mechanism mechanism enabled enabled the the identification identification of the the sender sender and and receiver receiver of of aa data data frame, frame, which which was was important important to to the the realization realization of offrame framerouting. routing. of

Figure 10. Format of frame.

4.2. Mission Control Software The mission control software, which runs on the UAV’s on-board computer, consists of the the software’s software’s block block diagram diagram is is shown shown in in Figure Figure 11. 11. following seven modules, and the

Figure 11. Block diagram of mission control software.

1. 1. 2. 2.

Flight over Flight Controller Controller Interface: Interface: manages manages direct direct communication communication with with Pixhawk Pixhawk flight flight controller controller over USB USB port. port. UAV and reports reports them them UAV Status Status Monitor: Monitor: monitors monitors flight flight and and mission mission parameters parameters of of the the UAV, UAV, and regularly to the GCS. regularly to the GCS.

Sensors 2016, 16, 1696

3. 4. 5. 6.

10 of 21

Command Interface: translates user commands into low-level instructions to the flight controller; records and reports command execution status. Communication System Interface: manages low-level message traffic between mission control subsystem and ration telemetries; executes message framing, frame reception, and error checking. Communication Relay Control: controls the relay of radio communication. Startup Control: bootstraps the whole mission control software system during UAV power-up process.

4.3. Software Realization of Relay Process The presence of a relay implies that the relaying node (UAV “Mom” in this case) needs to handle a frame routing. When processing received frames, the UAVs made use of the “Source ID” and “Target ID” fields to filter and/or route the frames in the following manner.

•

•

UAV “Son” is the terminal receiver, which only receives frames over the relay from “Mom.” Therefore, “Son” only saves for further processing frames whose “Target ID” is equal to the ID of “Son” itself. UAV “Mom” functions as a communication relay node, but it also receives control commands from GCS. Therefore, “Mom” will save for further processing frames whose “Target ID” is equal to the ID of “Mom” itself, and will re-transmit frames targeted at “Son” over Radio T2-1.

The above-mentioned frame routing logic is elaborated in the following pseudocode (Algorithms 1 and 2) and flow chart in Figure 12. Algorithm 1: Frame Processing Logic, UAV Son For (true) do: if (frame receive from T2-2) do: if (frame_target_id == id_son) do: save frame for local processing. else do: discard frame. end end end Algorithm 2: Frame Processing Logic, UAV Mom For (true) do: if (frame receive from T1-2) do: if (frame_target_id == id_mom) do: save frame for local processing. else if (frame_target_id == id_son) do: send frame out from T2-1. end else if (frame received from T2-1) do: if (frame_target_id == id_mom) do: save frame for local processing. else if (frame_target_id == id_GCS) do: send frame out from T1-1. end end end

Sensors 2016, 16, 1696

11 of 21

Sensors 2016, 16, 1696

11 of 20

The The last last field field of of the the frame frame contained contained aa four-byte four-byte cyclic cyclic redundancy redundancy checking checking (CRC) (CRC) code, code, which which was computed using the IEEE 802.3 32-bit CRC polynomial. This CRC code was important because it was computed using the IEEE 802.3 32-bit CRC polynomial. This CRC code was important because helps verify data frame it helps verify data frameintegrity integrityand andprotect protectthe thesystem systemfrom fromthe theinfluence influenceof oferror-corrupted error-corrupted data data or messages. Error-corrupted data or messages could be dangerous, as they might be mistakenly or messages. Error-corrupted data or messages could be dangerous, as they might be mistakenly understood unpredictable understood by by the the UAV’s UAV’s mission mission control control subsystem subsystem as as wrong wrong commands, commands, resulting resulting in in unpredictable and and potentially potentially dangerous dangerous behavior. behavior.

(a)

(b)

Figure “Son” and and (b) (b) UAV UAV“Mom”. “Mom”. Figure 12. 12. Flow Flow chart chart of of mission mission control control software software for for (a) (a) UAV UAV “Son”

4.4. Mini GCS 4.4. Mini GCS A ground control software, named “Mini GCS”, was developed in-house. Mini GCS is an A ground control software, named “Mini GCS”, was developed in-house. Mini GCS is an integrated UAV control interface that is compatible with Windows 7, 8, and 10 systems. The functions integrated UAV control interface that is compatible with Windows 7, 8, and 10 systems. The functions of Mini GCS include handling data exchanges with UAVs, displaying information about the UAVs, of Mini GCS include handling data exchanges with UAVs, displaying information about the UAVs, and and providing an interface for users to send commands to the UAVs. It allows users to monitor the providing an interface for users to send commands to the UAVs. It allows users to monitor the status of status of both UAVs, and to control the mission of each UAV from a single, integrated interface, both UAVs, and to control the mission of each UAV from a single, integrated interface, which makes it which makes it a crucial communication node in the communication relay network. Mini GCS was a crucial communication node in the communication relay network. Mini GCS was implemented using implemented using C#, and the graphic user interface (GUI) was constructed using Microsoft C#, and the graphic user interface (GUI) was constructed using Microsoft Windows .NET framework. Windows .NET framework. Because the main focus of this work is to demonstrate a signal relay Because the main focus of this work is to demonstrate a signal relay system for the UAV applications, system for the UAV applications, details of this Mini GCS software are not emphasized in the details of this Mini GCS software are not emphasized in the manuscript and the source code is included manuscript and the source code is included in the section of Supplementary Materials for reference. in the section of Supplementary Materials for reference. Figure 13 shows a screenshot of Mini GCS. A number of features are available. Figure 13 shows a screenshot of Mini GCS. A number of features are available.  UAV Status Display: displays to users a selected set of important information about each UAV, • UAV Status Display: displays to users a selected setmode, of important including communication status, GPS status, flight altitude,information speed etc. about each UAV, including communication status, GPS flight mode, altitude, speed  UAV Command Line: a command linestatus, interface that allows users to sendetc. specific commands to • UAV Command Line: a command line interface that allows users to send any UAV nodes in the network such as takeoff, land, return to launch, etc.specific commands to any UAV nodesDisplay: in the network such takeoff, land, return launch,of etc.  UAV Position displays theasreal-time position andtoheading both UAVs in a map of • UAV Position Display: displays the real-time position and heading of both UAVs in a map of the the mission area. mission area.

Sensors 2016, 16, 1696 Sensors 2016, 16, 1696

12 of 21 12 of 20

Figure 13. 13. Graphic Graphic user user interface interface (GUI) (GUI) of of Mini Mini GCS. GCS. Figure

5. Performance 5. PerformanceTests Tests To verify verify the the performance performance of of the the system, system, aa series series of of indoor indoor and and outdoor To outdoor tests tests were were carried carried out. out. First, the data loss rate and bit error rate (BER) were measured in the laboratory environment to First, the data loss rate and bit error rate (BER) were measured in the laboratory environment to choose the suitable data transmission rate for the relay system. Then, for outdoor tests, Received choose the suitable data transmission rate for the relay system. Then, for outdoor tests, Received Signal Strength Strength Indicator Indicator (RSSI), (RSSI), aa measurement measurement of of the the power power present present in in aa received received radio radio signal, signal, was was Signal measured. The value scales scales as as 1.9 1.9× offset, which which means means the the measured. The RSSI RSSI value × the the dBm dBm signal signal strength, strength, plus plus an an offset, RSSI is proportional to the dBm [35]. We tested both UAVs in two typical application scenerios in RSSI is proportional to the dBm [35]. We tested both UAVs in two typical application scenerios in Hong Kong: Kong: one Hong one was was to to cross cross obstacles obstacles and and the the other other was was to to extend extend the thecommunication communication range. range. Finally, Finally, a BLOS demonstration was conducted to show the rundown and results of a whole application a BLOS demonstration was conducted to show the rundown and results of a whole application process process in Notably, Taiwan.when Notably, when the transmission of module the telemetry is set to the in Taiwan. the transmission power of the power telemetry is set tomodule the maximum value, maximum is 20 dBm (100 mW), theachieved typical range standard configuration which is 20value, dBm which (100 mW), the typical range usingachieved standardusing configuration and antenna and antenna is around 500 m. Due to the stringent flight regulation and the area constraint of the is around 500 m. Due to the stringent flight regulation and the area constraint of the test sitetest in site in Hong Kong, it is almost impossible for the authors to test the beyond-line-of-sight flight Hong Kong, it is almost impossible for the authors to test the beyond-line-of-sight flight and and the the maximum power which would exceed the allowed flight Therefore, to demonstrate maximum power relayrelay which would exceed the allowed flight area.area. Therefore, to demonstrate the the relay function in Hong Kong, the authors have adjusted the transmission power of the radio to relay function in Hong Kong, the authors have adjusted the transmission power of the radio to its its minimum value (1/10 the max.)sosothat thatthe theradios radioswill willdisconnect disconnectat at about about 100 100 m. m. In minimum value (1/10 ofof the max.) In this this way, way, the relay methods can be tested in a relatively small area, otherwise, the telemetry modules will the relay methods can be tested in a relatively small area, otherwise, the telemetry modules will always always be connected. In missions requiring long distance communication, the transmission power be connected. In missions requiring long distance communication, the transmission power can be can beturned easily turned to a maximum to get best communication easily to a maximum value tovalue get the bestthe communication range. range. Tests 5.1. Indoor Tests To investigate investigatethe thereliability reliabilityofof our radio communication relay, the data loss and rate the andBER thewere BER To our radio communication relay, the data loss rate were first tested indoors. Verifying the data loss rate and BER in an indoor environment under first tested indoors. Verifying the data loss rate and BER in an indoor environment under different different gave us knowledge about bandwidth the usable bandwidth of and the helped system us and helpedthe us data ratesdata gaverates us knowledge about the usable of the system estimate estimate the maximum rate that be used in The the flight tests. The GCStowas configured maximum data rate that data can be used in can the flight tests. GCS was configured transfer data atto a transfer data at a fixed data rate, and the statistics of the received data at UAV “Son” were measured. fixed data rate, and the statistics of the received data at UAV “Son” were measured. Two aspects of Twodata aspects of the data transferthe were thewhich data loss which isof the percentage data that the transfer were studied: datastudied: loss rate, is therate, percentage data that was of lost during was lost during transmission (data that was transmitted from GCS but was not received by UAV transmission (data that was transmitted from GCS but was not received by UAV “Son”); and the BER, “Son”);was andthe thepercentage BER, which of data that was by “Son” buttest. did not pass which of was datathe thatpercentage was received by “Son” but received did not pass the CRC the CRC Teststest. were conducted at four different data rates: 4 kilobits per second (kbps), 8 kbps, 12.8 kbps, Tests were conducted at four different data rates: kilobits perof second (kbps), kbps, 12.8 kbps, and 16 kbps. Figure 14 shows the setup of the test. The 4percentages data loss and 8BER with respect and 16data kbps. Figure 14 shows the setup to the rate are shown in Figure 15. of the test. The percentages of data loss and BER with respect to the data rate are shown in Figure 15.

Sensors 2016, 16, 1696

13 of 21

Sensors Sensors 2016, 2016, 16, 16, 1696 1696

13 13 of of 20 20

Figure 14. relay testing setup. Figure 14. 14. Radio Radio communication relay testing testing setup. setup. Figure Radio communication communication relay

(a) (a)

(b) (b)

Figure 15. Percentage of data loss and rate with respect to data rate. Figure 15. (a) (a) and (b) (b) bit bit error error rate rate with with respect respect to to data data rate. rate. Figure 15. (a) Percentage Percentage of of data data loss loss and (b) bit error

The The tests tests indicated indicated that that when when the the data data rate rate was was equal equal to to or or higher higher than than 44 kbps, kbps, the the data data lost lost The tests indicated that when the data rate was equal to or higher than 4 kbps, the data lost percentage percentage rose rose from from 3% 3% to to 9% 9% and and the the data data error error rate rate rose rose from from 0.2% 0.2% to to 0.5%, 0.5%, approximately. approximately. A A percentage rose from 3% in to both 9% and the data error rate rose chance from 0.2% to 0.5%, approximately. higher data rate resulted more data loss and a higher of data transmission errors. higher data rate resulted in both more data loss and a higher chance of data transmission errors. A higheral. data estimated rate resulted the in both more data loss and amission higher chance of isdata transmission Andre Andre et et al. [7] [7] estimated that that the throughput throughput of of typical typical UAV UAV mission commands commands is less less than than 10 10 kbps. kbps. errors. Andre et al. [7] estimated that the throughput of typical UAV mission commands is less In In our our tests, tests, the the current current used used data data rate rate for for aa two-UAV two-UAV system system was was about about 88 kbps. kbps. Then, Then, for for than 10 kbps. In our tests, the current usedthis data rate for a two-UAV systemawas about 8ofkbps. Then, communication relay applications with system and configuration, data rate 8 kbps communication relay applications with this system and configuration, a data rate of 8 kbps is is for communication relay applications with this system and configuration, a data rate of 8 kbps is recommended recommended to to ensure ensure an an acceptable acceptable compromised compromised data data lose lose and and error error rate. rate. recommended to ensure an acceptable compromised data lose and error rate. 5.2. 5.2. Cross-Obstacle Cross-Obstacle Test Test 5.2. Cross-Obstacle Test The The outdoor outdoor flight flight tests tests were were first first conducted conducted in in the the International International Model Model Aviation Aviation Center Center of of the the The outdoor flight tests were first conducted in the International Model Aviation Center of the Hong Kong Model Engineering Club (22°24′58.1″ N 114°02′35.4″ E). Figure 16 shows the Google map Hong Kong Model Engineering Club (22°24′58.1″ N 114°02′35.4″ E). Figure 16 shows the Google map Hong Kong Model Engineering Club (22◦ 240 58.100 N 114◦ 020 35.400 E). Figure 16 shows the Google map of of the the test test site site and and final final positions positions of of the the UAVs UAVs in in the the test. test. In In the the simulated simulated cross-obstacle cross-obstacle of the test site and final positions of the UAVs in the test. In the simulated cross-obstacle communication communication communication relay relay test, test, the the UAV UAV “Son” “Son” was was designed designed to to conduct conduct aa mission mission that that takes takes off off from from the the relay test, the UAV “Son” was designed to conduct a mission that takes off from the “Home Position”, “Home Position”, flies over a part of the forest with tall trees, and then lands on the road across the “Home Position”, flies over a part of the forest with tall trees, and then lands on the road across the flies over a part of the forest with tall trees, and then lands on the road across the forest (UAV “Son” forest forest (UAV (UAV “Son” “Son” in in Figure). Figure). The The tall tall trees trees would would block block the the communication communication signal signal from from GCS GCS to to UAV UAV in Figure). The tall trees would block the communication signal from GCS to UAV “Son”. Then the “Son”. Then the UAV “Mom” took off and loitered at the UAV “Mom” position, which was above “Son”. Then the UAV “Mom” took off and loitered at the UAV “Mom” position, which was above UAV “Mom” took off and loitered at the UAV “Mom” position, which was above the forest. The UAV the the forest. forest. The The UAV UAV “Son” “Son” would would recover recover the the communication communication with with home home GSC GSC with with the the help help of of relay relay “Son” would recover the communication with home GSC with the help of relay point UAV “Mom”. point point UAV UAV “Mom”. “Mom”. The The relative relative positions positions of of all all the the communication communication nodes nodes can can be be more more clearly clearly The relative positions of all the communication nodes can be more clearly conceived in the schematic conceived conceived in in the the schematic schematic shown shown in in Figure Figure 17. 17. shown in Figure 17.

Sensors 2016, 16, 1696 Sensors 2016, 16, 1696 Sensors 2016, 16, 1696

14 of 20 14 of 21 14 of 20

teststest testininHong HongKong. Kong. Figure 16.16. Final positions cross-obstacletests Figure Final positionsofofnodes nodesininthe theGoogle Google map map in the cross-obstacle

Figure 17. Schematic of cross-obstacle communication relay flight test. Figure 17. Schematic of cross-obstacle communication relay flight test.

RSSI and noise of both UAV “Son” and UAV “Mom” were recorded and shown in Figure 18. RSSI and noise ofstrength both UAV “Son” and UAV “Mom” were recorded and in Figure 18. RSSI can reflect of “Son” wireless signal the UAV “Mom” andshown UAV “Son”. After RSSI and noisethe of both UAV and UAVreceived “Mom” by were recorded and shown in Figure 18. RSSI RSSI can reflect the strength of wireless signal received by the UAV “Mom” and UAV “Son”. After we found only signal when RSSI valueby is the 20 scale 7 dBm) higher than the average canseveral reflect tests, the strength of that wireless received UAV(about “Mom” and UAV “Son”. After several several tests, we found that only whenkeep RSSIa value is 20 scale (about 7 dBm) higher than the average noise value can the wireless modules stable connection. Therefore, a dashed blue line is added tests, we found that only when RSSI value is 20 scale (about 7 dBm) higher than the average noise noise valuefigure can the wireless modules a stable Therefore, a started dashedatblue line iswhen added to each to indicator when the keep connection is connection. lost or unstable. Theatest 0 s,added value can the wireless modules keep a stable connection. Therefore, dashed blue time line is to to each figure to indicator when the connection is lost or unstable. Thewhole test started at time 0 s, when UAV “Son” took off and flew to the mission area. The duration of the test mission was about each figure to indicator when the connection is lost or unstable. The test started at time 0 s, when UAV 10 “Son” min. Since s, theflew UAV reached theThe areaduration behind the tallwhole trees so themission RSSI decreased UAV took120 off and to “Son” the mission area. of the test was about “Son” took off and flew to the mission area. The duration of the whole test mission was about 10 min. until the noise level the and area the connection the GCS wasRSSI lost.decreased At this 10 significantly min. Since 120 s, itthereached UAV “Son” reached behind thewith tall trees so the Since 120 s, the UAV “Son” reached the area behind the tall trees so the RSSI decreased significantly moment, the UAV and level flew to the the relay point. Before 120the s, UAV stayed significantly until it “Mom” reachedtook the off noise and connection with GCS“Mom” was lost. At at this until it reached the noise level and the connection with the GCS was lost. At this moment, the UAV home near the GSC so the RSSI remained at a high value. After take off, the RSSI of UAV “Mom” moment, the UAV “Mom” took off and flew to the relay point. Before 120 s, UAV “Mom” stayed at “Mom” took off and flew to the relay point. Before 120 s, UAV “Mom” stayed at home near the GSC so decreased butGSC still so remained much higherat than the noise, that the UAV kept home near the the RSSI remained a high value.which After means take off, the RSSI of“Mom” UAV “Mom” theconnecting RSSI remained at a high value. After take off, the RSSI of UAV “Mom” decreased but still remained GSC. During 180 tothan 420 s, thenoise, UAV which “Son” flew over thethe forest and“Mom” landed on decreased butwith still the remained muchthe higher the means that UAV kept much higher than the noise, which means that thesame UAVscale “Mom” kept connecting with the GSC. During the road. In the Figure 17a, the RSSI fell to the with noise value. In this period, the UAV connecting with the GSC. During the 180 to 420 s, the UAV “Son” flew over the forest and landed on the“Son” 180 to 420 s, the disconnected UAV “Son” flew the forest and landed on the road. In theflew Figure 17a,tothe RSSI withover GCS. the time UAV “Mom” close UAV the road. remained In the Figure 17a, the RSSI fell to After the same scale420s, withthe noise value. In this period, the UAV fell“Son” to theand same scale with noise value. In this period, the UAV “Son” remained disconnected with relayed the communication to help UAV “Son” reconnect with the GCS. Then the RSSI “Son” remained disconnected with GCS. After the time 420s, the UAV “Mom” flew close to UAV GCS. After the time 420s, the UAV “Mom” flew close to UAV “Son” and relayed the communication to

“Son” and relayed the communication to help UAV “Son” reconnect with the GCS. Then the RSSI

Sensors 2016, 16, 1696 Sensors 2016, 16, 1696 Sensors 2016, 16, 1696

15 of 21 15 of 20 15 of 20

help “Son”recovered. reconnectThe with the GCS. Then the RSSI flied valuetoofUAV “Son” recovered. The UAV “Mom” valueUAV of “Son” UAV “Mom” continually “Son” (away from the GCS) for value of “Son” recovered. The UAV “Mom” continually flied to UAV “Son” (away from the GCS) for continually flied to UAV “Son” (away from the GCS) for a while which caused further increase of RSSI a while which caused further increase of RSSI of “Son” and decrease of “Mon” at the end period. a while caused further increase RSSI of “Son” and decrease of “Mon” at the end period. of “Son”which and decrease of “Mon” at the of end period.

(a) (a)

(b) (b)

Figure 18. Received Signal Strength Indicator (RSSI), noise and critical curve of both (a) UAV “Son” Figure 18. Received Signal Strength and (b)18. UAV “Mom”. Figure Received Signal Strength Indicator Indicator (RSSI), (RSSI), noise noise and and critical critical curve curve of of both both (a) (a) UAV UAV “Son” “Son” and (b) UAV “Mom”. and (b) UAV “Mom”.

5.3. Extend Distance Test 5.3. Extend Distance Test 5.3. Extend Test In the Distance communication distance extension tests in Hong Kong, the UAV “Son” was firstly given In the communication distance extension tests in Hong Kong, the UAV “Son” was firstly given a mission to fly away as far as possible until the connection lost. After thefirstly lost position, In the communication distance extension tests in Hongwas Kong, the UAVrecording “Son” was given a awhich mission to fly 100 away as far as possible until the connection was lost. After recording the lost isto about “Son” until continued to fly away forlost. about 60 mrecording and stayed the position, position, mission fly awaym, asthe far UAV as possible the connection was After theatlost position, which is about 100 m, the UAV “Son” continued to fly away for about 60 m and stayed at the position, waiting the100 UAV relaycontinued the communication. “Mom” took off and which is for about m, “Mom” the UAVto“Son” to fly awayThereafter, for about 60the m UAV and stayed at the position, waiting for the UAV “Mom” to“Son” relay until the communication. Thereafter, theThe UAV “Mom” took of offUAV and flew to the direction of UAV the connection was rebuild. final positions waiting for the UAV “Mom” to relay the communication. Thereafter, the UAV “Mom” took off and flew to the direction of UAV “Son” 19. until the the connection The final positions of UAV nodes test is of shown Figure With help was of awas relayrebuild. node, 160 positions m total connection was flew to in thethis direction UAVin “Son” until the connection rebuild. The afinal of UAV nodes nodes in this test is shown in Figure 19. With the help of a relay node, a 160 m total connection was achieved. in this test is shown in Figure 19. With the help of a relay node, a 160 m total connection was achieved. achieved.

Figure Figure 19. 19. Final Final positions positions of of nodes nodes in in the the Google Google map map in in the the extended extendeddistance distancetest testin inHong HongKong. Kong. Figure 19. Final positions of nodes in the Google map in the extended distance test in Hong Kong.

Sensors 2016, 16, 1696 Sensors 2016, 16, 1696 Sensors 2016, 16, 1696

16 of 21 16 of 20 16 of 20

The RSSI, noise and critical curves during the full test are shown in Figure 20. As is similar with The RSSI, noise and critical curves during the full test are shown in Figure 20. As is similar with the cross obstacle test, the communication between UAV “Son” and GCS was first lost at about 180 s the obstacle test, the between UAV “Son”takeoff and GCS was first losttoatUAV about 180 s andcross finally reconnected atcommunication about 300 300 ss when when the the UAV UAV “Mom” “Mom” takeoff and and flew flew near near to UAV “Son”. “Son”. and finallyofreconnected about 300 when the UAV “Mom” takeoff and flewhome nearposition. to UAV “Son”. RSSI of of UAV UAVat“Mom” “Mom” wassdue due toits itstakeoff takeoff andflying flying away from The drop of RSSI was to and away from home position. The drop of RSSI of UAV “Mom” was due to its takeoff and flying away from home position.

(a) (a)

(b) (b)

Figure “Son” and and (b) (b) UAV UAV “Mom”. “Mom”. Figure 20. 20. RSSI, RSSI, noise, noise, and and critical critical curve curve of of both both (a) (a) UAV UAV “Son” Figure 20. RSSI, noise, and critical curve of both (a) UAV “Son” and (b) UAV “Mom”.

5.4. Beyond-Line-of-Sight Tests 5.4. Beyond-Line-of-Sight Tests Tests 5.4. Beyond-Line-of-Sight To further further verify verify the the reliability reliability of of the the communication communication relay relay system, system, including including mini mini GCS, GCS, in in aa To To further verify the reliability of the communication relay system, including mini GCS, in a beyond-line-of-sight missions, outdoor flight tests were conducted in Xigang flight site, Tainan City, beyond-line-of-sight missions, outdoor flight tests were conducted in Xigang flight site, Tainan City, beyond-line-of-sight missions, outdoor flight tests were conducted in Xigang flight site, Tainan City, Taiwan (23 (23°6′21.5″ 120°12′36.0″ E). In this test, the transmission power of the telemetry modules ◦ 60 21.500 N, Taiwan N, 120°12′36.0″ 120◦ 120 36.000 E). E).In In this this test, test, the the transmission transmission power power of of the the telemetry modules Taiwan (23°6′21.5″ N, telemetry modules were turned turned to to the the maximum maximum value value to to get get the the longest longest communication communication range. range. The The Google Google map map of of the the were were turned to the maximum value to get the longest communication range. The Google map of the test site and final positions of the UAVs in the test are shown in Figure 21. test site site and and final test final positions positions of of the the UAVs UAVs in in the the test test are areshown shownin inFigure Figure21. 21.

Figure 21. Final positions of nodes in the Google map in the extended distance test in Taiwan. Figure Figure 21. 21. Final Final positions positions of of nodes nodes in in the the Google Google map map in in the the extended extended distance distance test test in in Taiwan. Taiwan.

1. 1. 2. 1. 2. 3. 2. 3.

The rundown of the typical flight mission is given below: The rundown of the typical flight mission is given below: The rundown of theand typical flight mission given below: All UAVs (“Mom” “Son” in this case)isare powered up. All UAVs (“Mom” and “Son” in this case) are powered up. Start GCS software “Mini GCS”. All UAVs software (“Mom” and “Son” in this case) are powered up. Start “Mini GCS”. Wait GCS for communication relay network to start. The startup process is automatic and normally Start GCS software “Mini GCS”. Wait for communication relay network to start. The startup process is automatic and normally takes less than three minutes. takes less than three minutes.

Sensors 2016, 16, 1696

17 of 21

3.

Wait for communication relay network to start. The startup process is automatic and normally takes less than three minutes. Sensors 2016, 16, 1696 17 of 20 4. Using “Mini GCS”, command “Son” to take off and start the execution of a preprogrammed whichGCS”, bringscommand “Son” to “Son” a destination radio communication between 4. mission, Using “Mini to take far off enough and startfor thethe execution of a preprogrammed “Son” and which “Mini brings GCS” to break. mission, “Son” to a destination far enough for the radio communication between 5. Wait until the communication between “Son” and “Mini GCS” breaks due to distance. “Son” and “Mini GCS” to break. 6. 5. Using GCS”, command “Mom” take off move intobreaks the proximity of the destination Wait “Mini until the communication betweento“Son” andand “Mini GCS” due to distance. 6. area Using “Mini GCS”, command “Mom” performs to take off aand move into the proximity of it themoves destination of “Son”. “Mom” automatically communication relay when toward area of “Son”. “Mom” automatically performs a communication relay when it moves toward “Son” and reconnects. “Son” and reconnects. The rundown of the test is shown as a series of pictures in Figure 22. A video clip demonstrating The rundown of the test is shown as a series of pictures in Figure 22. A video clip demonstrating the complete test procedure is given in the Supplementary Material. In the video demonstration, the complete test procedure is given in the Supplementary Material. In the video demonstration, UAV “Son” started a pre-defined mission. The mission is to fly away for about 600 m until the UAV UAV “Son” started a pre-defined mission. The mission is to fly away for about 600 m until the UAV was BLOS inin Figure “Son”and andGCS GCSwas waslost lost(indicated (indicated was BLOS Figure22c. 22c.When Whenthe theconnection connection between between “Son” byby thethe green “Connected” icon changing to the red “Disconnected”), “Mom” was deployed to a designated green “Connected” icon changing to the red “Disconnected”), “Mom” was deployed to a designated relaying position (about 400400 m away fromfrom homehome position). Once Once “Mom” reached the relaying position, relaying position (about m away position). “Mom” reached the relaying communication between “Son” and GCSand wasGCS re-established automatically and was again position, communication between “Son” was re-established automatically and not was lost not lost before the end of the mission. again before the end of the mission.

(a)

(b)

(c)

(d)

(e)

(f)

Figure 22. A rundown of flight test (a) Pre-takeoff setup of flight test; (b) UAV “Son” starts the Figure 22. A rundown of flight test (a) Pre-takeoff setup of flight test; (b) UAV “Son” starts the mission; mission; (c) “Son” is out of radio range and communication with it is lost; (d) UAV “Mom” is deployed (c) “Son” is out of radio range and communication with it is lost; (d) UAV “Mom” is deployed to the to the relay position; (e) Communication with “Son” is re-established after being relayed through relay position; (e) Communication with “Son” is re-established after being relayed through “Mom”; “Mom”; (f) “Mom” and “Son” return to launch position; the mission is completed. (f) “Mom” and “Son” return to launch position; the mission is completed.

Eight flight tests were conducted. In all of the tests, the system was able to re-establish communication between “Son” and “Mini GCS” through the relay of “Mom”. When “Mom” reached the designated relaying position, the communication connection between “Son” and “Mini GCS” was

Sensors 2016, 16, 1696

18 of 21

flight SensorsEight 2016, 16, 1696

tests were conducted. In all of the tests, the system was able to re-establish 18 of 20 communication between “Son” and “Mini GCS” through the relay of “Mom”. When “Mom” reached adequately stable for mission operations. The successful rate of the communication the designated relaying position, the communication connection between “Son” and re-establishment “Mini GCS” was is 100%. The tests that: (1) system rate is capable of maintainingre-establishment uninterrupted adequately stable fordemonstrated mission operations. Thethe successful of the communication communication between GCS and anthat: end node once the relaying nodeofismaintaining properly deployed; and (2) is 100%. The tests demonstrated (1) the system is capable uninterrupted connectivity between GCSGCS and and the end node can once be automatically when the relaying communication between an end node the relaying re-established node is properly deployed; and node reaches the properGCS relaying position, evencan if it been previously lost duewhen to an the inadequate (2) connectivity between and the end node behas automatically re-established relaying relay. node reaches the proper relaying position, even if it has been previously lost due to an inadequate relay. 6. Conclusions Conclusionsand andFuture FutureDevelopment Development In this attention waswas given to implementing a functional and reliable proof-of-concept thisstudy, study,major major attention given to implementing a functional and reliable proof-ofUAV communication relay system uses that commercially available hardware components combined concept UAV communication relaythat system uses commercially available hardware components with customized software. In the prototype system, two quadrotors, each mounted with a Pixhawk combined with customized software. In the prototype system, two quadrotors, each mounted with a flight controller and a Raspberry Pi 2 Model B microcomputer were used as the aerial platforms. Pixhawk flight controller and a Raspberry Pi 2 Model B microcomputer were used as the aerial The onboard relay controlrelay protocol and protocol program and wereprogram developed to realize UAV to communication platforms. The onboard control were developed realize UAV extension and cross-obstacle The GSC software wassoftware also developed facilitate communication extension andcommunication. cross-obstacle communication. The GSC was alsoto developed thefacilitate two-UAV operations. A series ofAindoor and outdoor were carried Theout. indoor to therelay two-UAV relay operations. series of indoor and tests outdoor tests wereout. carried The tests aimed determine the suitable transmission rate, while the outdoor indoor teststoaimed to determine thedata suitable data transmission rate, while tests the demonstrated outdoor tests reliable communication performance performance in two proposed application Hong Kong and demonstrated reliable communication in two proposed scenarios applicationinscenarios in Hong one BLOS simulation Taiwan. All the outdoor testsoutdoor showedtests the system’s capability of Kong and mission one BLOS mission in simulation in Taiwan. All the showed the system’s maintaining quality real-time communication through relays, through and automatically recovering from capability ofgood maintaining good quality real-time communication relays, and automatically communication recovering from losses. communication losses. communication relay relay through through aa single single relaying relaying node node (UAV (UAV The current work focused only on communication “Mom”). although feasible, feasible, will will be conducted conducted in the near “Mom”). Relays through multiple relaying nodes, although The current current configuration configuration of of the the system system allows allows the the system system to be extended by adding more future. The relaying nodes. A possible configuration is illustrated in Figure 23, in which multiple relaying nodes (“Mom 1” to “Mom N”) are connected connected in series, series, forming forming a bidirectional bidirectional relay relay system. system. The current design of the data frame permits, at most, 13 relaying nodes between “Son” and the GCS. When the format of the data frames was designed, four binary bits were assigned to represent the “ID” of the UAV. With among which which UAV UAV “Son”, “Son”, UAV. With four four bits bits to to use, use, at at most most 16 unique IDs could be represented, among broadcasting ID occupied occupied three places. places. Therefore, Therefore, there there were were 13 13 unique unique IDs IDs left, left, GCS, and a special broadcasting which could then be taken up by 13 multiple UAV “Moms”, theoretically. However, the maximum number of relaying nodes in practice may also be subjected to the constraint of the data rate. With If the the data data rate rate is kept kept unchanged, unchanged, more UAVs in the system, more information will be exchanged. If communication between UAVs and GCS might experience longer delays. UAVs and GCS might experience longer delays.

Figure nodes. Figure 23. 23. Extending Extending the the relay relay system system by by adding adding relaying relaying nodes. Supplementary Materials: The following is available online at https://www.youtube.com/watch?v=HxN0oafNmzw, Video S1: Multiple UAVs Telecommunication Demonstration. https://github.com/YifanJiangPolyU/UAVcommunication-relay, Source code 1: Mission control software source code. https://github.com/HKPolyUUAV/Mini_GCS, Source code 2: Mini-GSC source code. Acknowledgments: This work is sponsored by Innovation and Technology Commission, Hong Kong under Contract No. ITS/334/15FP. We also appreciate Hiu Fung Wan for his help in mechanical parts of this work.

Sensors 2016, 16, 1696

19 of 21

Supplementary Materials: The following is available online at https://www.youtube.com/watch?v=HxN0oafNmzw, Video S1: Multiple UAVs Telecommunication Demonstration. https://github.com/YifanJiangPolyU/UAVcommunication-relay, Source code 1: Mission control software source code. https://github.com/HKPolyU-UAV/ Mini_GCS, Source code 2: Mini-GSC source code. Acknowledgments: This work is sponsored by Innovation and Technology Commission, Hong Kong under Contract No. ITS/334/15FP. We also appreciate Hiu Fung Wan for his help in mechanical parts of this work. Author Contributions: Y.J. and J.S. conceived the relay system. B.L. built up the hardware platform. Y.J. developed the protocol and onboard software. L.C. developed the G.C.S. software. B.L. and Y.J. performed the tests and wrote the paper. C.W. was in charge of the whole project management. Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations The following abbreviations are used in this manuscript: BLOS UAV GCS RSSI FHSS TDM UART CRC BER GUI

Beyond-Line-of-Sight Unmanned Aerial Vehicle Ground Control Station Received Signal Strength Indication Frequency-Hopping Spread Spectrum Time Division Multiplex Universal Asynchronous Receiver and Transmitter Cyclic Redundancy Checking Bit Error Rate Graphic User Interface

References 1. 2. 3. 4. 5.

6. 7.

8. 9.

10. 11.

12. 13.

Nonami, K. Prospect and recent research & development for civil use autonomous unmanned aircraft as UAV and MAV. J. Syst. Des. Dyn. 2007, 1, 120–128. Francis, M.S. Unmanned air systems: Challenge and opportunity. J. Aircr. 2012, 49, 1652–1665. [CrossRef] Colomina, I.; Molina, P. Unmanned aerial systems for photogrammetry and remote sensing: A review. ISPRS J. Photogramm. Remote Sens. 2014, 92, 79–97. [CrossRef] ˙ Sahingoz, O.K.; Temel, S¸ . Flying ad hoc networks (fanets): A survey. Ad Hoc Netw. 2013, 11, Bekmezci, I.; 1254–1270. [CrossRef] Arena, A.; Chowdhary, G.; Conner, J.; Gaeta, R.; Jacob, J.; Kidd, J. Development of an Unmanned Aircraft Systems (UAS) Option at the Graduate Level. In Proceedings of the 51st AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition, Grapevine, TX, USA, 7–10 January 2013. Gupta, L.; Jain, R.; Vaszkun, G. Survey of important issues in UAV communication networks. IEEE Commun. Surv. Tutor. 2016, 18, 1123–1152. [CrossRef] Andre, T.; Hummel, K.A.; Schoellig, A.P.; Yanmaz, E.; Asadpour, M.; Bettstetter, C.; Grippa, P.; Hellwagner, H.; Sand, S.; Zhang, S.W. Application-driven design of aerial communication networks. IEEE Commun. Mag. 2014, 52, 128–136. [CrossRef] Pinkney, F.J.; Hampel, D.; DiPierro, S. Unmanned Aerial Vehicle (UAV) Communications Relay. In Proceedings of the Milcom 96, McLean, VA, USA, 21–24 October 1996; Volumes 1–3, pp. 47–51. Rangel, R.K.; Kienitz, K.H.; de Oliveira, C.A.; Brandao, M.P. Development of a Multipurpose Tactical Surveillance System Using UAV’S. In Proceedings of the 2014 IEEE Aerospace Conference, Big Sky, MT, USA, 1–8 March 2014; pp. 1–9. Orfanus, D.; de Freitas, E.P.; Eliassen, F. Self-organization as a supporting paradigm for military UAV relay networks. IEEE Commun. Lett. 2016, 20, 804–807. [CrossRef] Choi, Y.; Pachter, M.; Jacques, D. Optimal Relay UAV Guidance—A New Differential Game. In Proceedings of the 50th IEEE Conference on Decision and Control and European Control Conference (CDC-ECC), Orlando, FL, USA, 12–15 December 2011; pp. 1024–1029. Zhan, P.C.; Yu, K.; Swindlehurst, A.L. Wireless relay communications with unmanned aerial vehicles: Performance and optimization. IEEE Trans. Aerosp. Electron. Syst. 2011, 47, 2068–2085. [CrossRef] Cetin, O.; Zagli, I. Continuous airborne communication relay approach using unmanned aerial vehicles. J. Intell. Robot. Syst. 2012, 65, 549–562. [CrossRef]

Sensors 2016, 16, 1696

14. 15. 16. 17.

18.

19. 20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30. 31. 32.

20 of 21

Cetin, O.; Zagli, I.; Yilmaz, G. Establishing obstacle and collision free communication relay for UAVs with artificial potential fields. J. Intell. Robot. Syst. 2013, 69, 361–372. [CrossRef] Hauert, S.; Zufferey, J.-C.; Floreano, D. Evolved swarming without positioning information: An application in aerial communication relay. Auton. Robot. 2009, 26, 21–32. [CrossRef] Jiang, F.; Swindlehurst, A.L. Dynamic UAV Relay Positioning for the Ground-to-Air Uplink. In Proceedings of the 2010 IEEE Globecom Workshops, Miami, FL, USA, 5–10 December 2010; pp. 1766–1770. Marinho, M.A.M.; de Freitas, E.P.; da Costa, J.P.C.L.; de Almeida, A.L.F.; de Sousa, R.T. Using Cooperative Mimo Techniques and UAV Relay Networks to Support Connectivity in Sparse Wireless Sensor Networks. In Proceedings of the 2013 International Conference on Computing, Management and Telecommunications (Commantel), Ho Chi Minh City, Vietnam, 21–24 January 2013; pp. 49–54. Ben Moussa, C.; Gagne, S.; Akhrif, O.; Gagnon, F. Aerial Mast vs. Aerial Bridge Autonomous UAV Relay: A Simulation-Based Comparison. In Proceedings of the International Conference on New Technologies, Dubai, United Arab Emirates, 30 March–2 April 2014; pp. 3118–3123. Lee, K.; Noh, H.; Lim, J. Airborne relay-based regional positioning system. Sensors 2015, 15, 12682–12699. [CrossRef] [PubMed] Brown, T.X.; Argrow, B.; Dixon, C.; Doshi, S.; Thekkekunnel, R.-G.; Henkel, D. Ad Hoc UAV Ground Network (Augnet). In Proceedings of the AIAA 3rd Unmanned Unlimited Technical Conference, Chicago, IL, USA, 20–23 September 2004; pp. 1–11. Brown, T.X.; Doshi, S.; Jadhav, S.; Himmelstein, J. Test Bed for a Wireless Network on Small UAVs. In Proceedings of the AIAA 3rd Unmanned Unlimited Technical Conference, Chicago, IL, USA, 20–23 September 2004; pp. 20–23. Jimenez-Pacheco, A.; Bouhired, D.; Gasser, Y.; Zufferey, J.C.; Floreano, D.; Rimoldi, B. Implementation of a Wireless Mesh Network of Ultra Light MAVs with Dynamic Routing. In Proceedings of the IEEE Globe Work, Anaheim, CA, USA, 3–7 December 2012; pp. 1591–1596. Morgenthaler, S.; Braun, T.; Zhao, Z.L.; Staub, T.; Anwander, M. Uavnet: A Mobile Wireless Mesh Network Using Unmanned Aerial Vehicles. In Proceedings of the IEEE Globe Work, Anaheim, CA, USA, 3–7 December 2012; pp. 1603–1608. Asadpour, M.; Giustiniano, D.; Hummel, K.A. From Ground to Aerial Communication: Dissecting WLAN 802.11N for the Drones. In Proceedings of the WiNTECH 0 13 8th ACM International Workshop on Wireless Network Testbeds, Experimental Evaluation & Characterization, Miami, FL, USA, 30 September–4 October 2013; pp. 25–32. Asadpour, M.; Giustiniano, D.; Hummel, K.A.; Heimlicher, S.; Egli, S. Now or Later?—Delaying Data Transfer in Time-Critical Aerial Communication. In Proceedings of the 2013 ACM International Conference on Emerging Networking Experiments and Technologies (Conext ‘13), New York, NY, USA, 9–12 December 2013; pp. 127–132. Yanmaz, E.; Kuschnig, R.; Bettstetter, C. Achieving Air-Ground Communications in 802.11 Networks with Three-Dimensional Aerial Mobility. In Proceedings of the IEEE Infocom Ser, Turin, Italy, 14–19 April 2013; pp. 120–124. Yanmaz, E.; Hayat, S.; Scherer, J.; Bettstetter, C. Experimental Performance Analysis of Two-Hop Aerial 802.11 Networks. In Proceedings of the 2014 IEEE Wireless Communications and Networking Conference (WCNC), Istanbul, Turkey, 6–9 April 2014; pp. 3118–3123. Johansen, T.A.; Zolich, A.; Hansen, T.; Sørense, A.J. Unmanned Aerial Vehicle as Communication Relay for Autonomous Underwater Vehicle—Field Tests. In Proceedings of the Globecom 2014 Workshop—Wireless Networking and Control for Unmanned Autonomous Vehicles, Austin, TX, USA, 8–12 December 2014. Guo, W.; Devine, C.; Wang, S. Performance Analysis of Micro Unmanned Airborne Communication Relays for Cellular Networks. In Proceedings of the 9th International Symposium on Communication Systems, Networks & Digital Signal Processing (CSNDSP), Manchester, UK, 23–25 July 2014; pp. 658–663. Deng, C.; Wang, S.; Huang, Z.; Tan, Z.; Liu, J. Unmanned aerial vehicles for power line inspection: A cooperative way in platforms and communications. J. Commun. 2014, 9, 687–692. [CrossRef] Willink, T.J.; Squires, C.C.; Colman, G.W.K.; Muccio, M.T. Measurement and characterization of low-altitude air-to-ground mimo channels. IEEE Trans. Veh. Technol. 2016, 65, 2637–2648. [CrossRef] 3DR Pixhawk. Available online: https://store.3dr.com/products/3dr-pixhawk (accessed on 30 June 2016).

Sensors 2016, 16, 1696

33. 34. 35.

21 of 21

Sik Telemetry Radio. Available online: http://ardupilot.org/copter/docs/common-sik-telemetry-radio. html (accessed on 30 June 3016). Dronecode/Sik. Available online: https://github.com/Dronecode/SiK (accessed on 30 June 2016). Sik Radio—Advanced Configuration. Available online: http://ardupilot.org/copter/docs/common3dr-radio-advanced-configuration-and-technical-information.html#common-3dr-radio-advancedconfiguration-and-technical-information (accessed on 15 August 2016). © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).