Received 10 September 2022, accepted 16 September 2022, date of publication 27 September 2022,

date of current version 30 September 2022.

Digital Object Identifier 10.1109/ACCESS.2022.3208109

Soil Moisture Monitoring Through UAS-Assisted

Internet of Things LoRaWAN Wireless
Underground Sensors
FAHIM FERDOUS HOSSAIN 1 , RUSS MESSENGER 1 , GEORGE L. CAPTAIN1 ,
SABIT EKIN 1 , (Senior Member, IEEE), JAMEY D. JACOB2 , SALEH TAGHVAEIAN3 ,
AND JOHN F. O’HARA 1 , (Senior Member, IEEE)
1 School of Electrical and Computer Engineering, Oklahoma State University, Stillwater, OK 74078, USA
2 School of Mechanical and Aerospace Engineering, Oklahoma State University, Stillwater, OK 74078, USA
3 Biological Systems Engineering, University of Nebraska-Lincoln, Lincoln, NE 68588, USA

Corresponding author: John F. O’Hara ([email protected])

This work was supported in part by the United States Geologic Survey (USGS) under Award G16AP00077, in part by the National Science
Foundation (NSF-EPSCoR) under Award OIA-1946093, and in part by the OSU Unmanned Systems Research Institute.

1 ABSTRACT With growing usage and demand for the global freshwater supply, there is an increasing need
2 for technologies that facilitate water conservation and environmental stewardship in irrigated agriculture.
3 Toward this end, recent demonstrations of Internet of Things (IoT) sensors have revealed the value
4 of wireless soil moisture content sensors. However, existing wireless solutions often employ above-ground
5 wireless communication modules that physically interfere with routine farming operations. Underground
6 wireless network solutions are severely challenged by the large radio-frequency (RF) propagation loss
7 through soil. This paper presents a method that overcomes both problems by employing completely buried
8 underground IoT sensors and communication modules with UAS (uncrewed aircraft system) mounted
9 LoRaWAN gateways. The UAS mounted LoRaWAN gateway eliminates the need for any in-field base
10 stations and also allows the LoRa enabled sensors to transmit data over short distances with very low energy.
11 Field tests were carried out using this approach to serve as a proof of concept. The RSSI (received signal
12 strength indicator) demonstrates that the proposed solution has good communication link margin and a
13 significantly larger communication range than is necessary for reliable operation. Moreover, this solution
14 is easy to build, scalable, cost-effective, and can be implemented in a highly power efficient fashion.

15 INDEX TERMS Internet of Things (IoT), long range radio (LoRa), scientific irrigation scheduling, soil
16 moisture monitoring, uncrewed aircraft system (UAS).

17 I. INTRODUCTION biodiversity in soil [1]. On the other hand, over-irrigation can 26

18 The growing worldwide population presents an increasing be equally destructive, causing salinization [2] of the land and 27

19 impetus to properly manage the sources and strains on the pollution [3] of freshwater sources when chemigated water 28

20 global freshwater supply. Agriculture is the biggest consumer is used for irrigation. The objective then becomes to always 29

21 of global freshwater supply while about half of the world pop- provide the optimal level of irrigation, just enough to maintain 30

22 ulation experiences water scarcity [1]. Nevertheless, main- required soil moisture level. This is referred to as scientific 31

23 taining some minimum required soil moisture is important irrigation scheduling [4] which falls under the umbrella of 32

24 in various biophysical processes like plant growth, germina- precision agriculture, and soil moisture monitoring is an 33

25 tion of seeds, nutrient cycling and sustenance of the natural important element in meeting this objective [5], [6]. 34

There are different ways of sensing and monitoring soil 35

The associate editor coordinating the review of this manuscript and moisture levels in agricultural fields. Generally, crop fields 36

approving it for publication was Ghufran Ahmed . have large areas, therefore implementing a wireless solution 37

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
VOLUME 10, 2022 102107
 F. F. Hossain et al.: Soil Moisture Monitoring Through UAS-Assisted IoT LoRaWAN Wireless Underground Sensors

38 for monitoring soil moisture is preferable to avoid the cost and agricultural applications, LoRa is the best option [17]. LoRa 94

39 complexity of laying and maintaining long lengths of cable. uses chirped spread-spectrum modulation, with 433 MHz, 95

40 In the wireless scheme, multiple sensors can be deployed in a 868 MHz, or 915 MHz as the carrier frequency – depending 96

41 field to set up a wireless sensor network that can convey soil on region – with a maximum data rate of 50 kbps [18], 97

42 moisture data from various parts of the field to the network which is more than adequate for soil moisture monitoring 98

43 server. Several wireless technologies have been demonstrated and occasional reporting. The relatively low carrier frequen- 99

44 in this regard for precision agriculture, including GPRS (Gen- cies translate to manageable free-space path loss and some 100

45 eral Packet Radio Service), GSM (Global System for Mobile immunity to line-of-sight issues. This generally gives LoRa 101

46 Communication), Wi-Fi, Bluetooth, and Zigbee [7]. In a a coverage radius of between 1 to 10 km above ground. In a 102

47 similar application, forest fire monitoring systems based on work published by Renzone et al. [19], LoRaWAN sensors 103

48 Zigbee wireless sensors have also been demonstrated [8]. were buried underground with different soil types and depths 104

49 The research challenges of wireless underground sensor between 10 and 50 cm. In their experiment, they measured 105

50 networks have been reviewed in [9]. If all sensor nodes and the corresponding path and packet loss using a receiver with 106

51 communication modules can be placed completely under- its antenna placed on the ground 15 m horizontally away 107

52 ground, then devices can avoid interfering with farm machin- from the buried sensor. Additionally, in [19], the RSSI was 108

53 ery, a significant attraction. The main challenge of this measured over a 20 day time period to measure the impacts 109

54 scheme is the high RF (radio frequency) path loss, which of soil compaction. 110

55 causes low data rate for a given power, or high power con- Another research avenue of environmental and agricultural 111

56 sumption for a given data rate compared to above-ground monitoring involves uncrewed aircraft systems (UASs) [20]. 112

57 communication [9]. Data cannot be transported over long The applications of UAS in the field of irrigation management 113

58 distance in an underground sensor network without excessive have been comprehensively reviewed in [21]. UAS-based, 114

59 power use, so it must be re-routed relatively quickly to an low-altitude remote sensing technologies and wireless sen- 115

60 above-ground stationary or mobile base stations or gateways sor networks have been studied for precision weed manage- 116

61 [9]. This challenge is compensated by setting up a dense array ment [22] and for preventing frost in fragmented vineyards 117

62 of above-ground base stations, which is neither cost-effective by monitoring temperature and humidity [23]. Soil moisture 118

63 nor practical. As a less traditional wireless solution, magnetic content monitoring using UAS-based hyperspectral imagery 119

64 induction was also investigated for underground communica- has also been demonstrated [24]. Ground sensor communi- 120

65 tion [10]. This technique leverages the lower material losses cation with UAS-based small cell networks was developed 121

66 of non-propagating magnetic fields for communication, but in [25]. The ground sensors have low energy and low cov- 122

67 also suffers a very short range due to the more rapid roll-off erage area while the UAS-SC (UAS-Small Cell) acts as a 123

68 of near-field power density. Underground magnetic induction mobile transceiver that collects the data from the ground sen- 124

69 waveguide systems were subsequently investigated, which sors by flying over them [25]. In this way a UAS can be used 125

70 proved that a large range can be achieved by employing a as a mobile gateway for low power IoT sensors. But in [25], 126

71 large number of relay coils in short intervals between the above-ground communication modules were employed along 127

72 transmitter and receiver [11]. While feasible, this dramati- with solar cells for powering the modules and sensors. These 128

73 cally increases the complexity and cost of this solution. pose a significant practical hindrance for farm machinery 129

74 More possible agricultural solutions have emerged under operations. 130

75 the umbrella of the Internet of Things (IoT) - a vast network In this paper, we present a soil moisture monitoring system 131

76 of relatively simple devices such as sensors and cameras, which employs completely buried IoT sensors with integrated 132

77 connected to the internet via (predominantly) wireless com- communication modules, no in-field base stations or gate- 133

78 munication networks [12]. For example, [13] employed an ways, and a UAS-based mobile gateway for retrieving the soil 134

79 IoT-based wireless sensor network to develop a Greenhouse moisture data from the buried sensors. The buried sensors 135

80 Irrigation Management System to monitor various agricul- and the gateway aboard the UAS use LoRa-based wireless 136

81 tural parameters including soil moisture. It is significant to communication. A similar work [26] has been published by 137

82 this work that a variant of IoT called the Internet of Under- Cariou et al.; however, there are some important differences 138

83 ground Things (IoUT) has developed with applications in between our work and [26]. The operating frequency in our 139

84 agriculture wherein interconnected things are partially or work is 916 MHz whereas in [26], the operating frequency 140

85 completely placed underground [14]. is 868 MHz. Our sensor node was buried at a depth of 141

86 LoRa is a wireless radio technology and LoRaWAN is a 0.3 meters whereas the sensor node in [26] was buried at a 142

87 wide area network protocol that incorporates the LoRa tech- depth of 0.15 meters. Generally, sensor nodes buried deeper 143

88 nology [15]. In IoT applications, the LoRaWAN (Low Range into the ground can prove more beneficial as they are more 144

89 Wide Area Network), or just LoRa, protocol has emerged as likely to be cleared from farming machinery operations. 145

90 a very promising technology since it allows long range, low In our work, from experimental data and path loss analysis, 146

91 data rate wireless communication utilizing very low power we have demonstrated that the relative antenna orientation 147

92 LoRa enabled sensors [16]. After analyzing many different of the buried sensor and the above ground gateway has little 148

93 technologies, some researchers have concluded that for smart significance in communicating data packets - which in [26] 149

102108 VOLUME 10, 2022

 F. F. Hossain et al.: Soil Moisture Monitoring Through UAS-Assisted IoT LoRaWAN Wireless Underground Sensors

FIGURE 1. Block diagram for the mobile LoRaWAN gateway hardware (arrows indicate direction of data flow).

FIGURE 2. Block diagram for buried LoRa enabled sensor hardware setup (arrows indicate direction of data
flow).

150 has not been presented. We have also presented our analysis performance, and estimates of real-world limitations in terms 172

151 of percentage packet loss with varying distances which in [26] of battery life and operational ranges. 173

152 has been mentioned to be done in future work. In [26], it has

153 been shown that the RSSI (received signal strength indicator)
154 values after a particular horizontal distance is higher at a II. MATERIALS AND METHODS 174

155 higher altitude - the impact of the soil on the propagation of A. DEVELOPMENT of LoRaWAN GATEWAY and LoRa 175

156 the signal has been attributed as the reason for this - whereas ENABLED SENSOR 176

157 we have elaborated on the mechanisms of signal propaga- 1) MOBILE LoRaWAN GATEWAY HARDWARE 177

158 tion from soil to air from an electromagnetic perspective The LoRaWAN gateway used to collect data from the LoRa 178

159 and also using an electromagnetic simulation. In addition, enabled sensor was primarily based on a Raspberry Pi 4. 179

160 unlike in [26], we have presented a plausible estimation of A LoRa module was added to function as the communica- 180

161 the battery lifetime of the LoRa enabled sensor node. The tion module along with a dipole antenna. When mounted 181

162 advantages of our proposed solution are threefold: First, since on a UAS, the Raspberry Pi was also accompanied by a 182

163 all the end node devices are buried completely underground, GPS/GLONASS (Global Positioning System/ Global Nav- 183

164 they do not interfere with normal farming operations; Sec- igation Satellite System) module. The reason for using a 184

165 ond, since the UAS moves from one sensor to another to GPS/GLONASS module will be apparent in a later section. 185

166 collect data, the sensor nodes can transmit very little energy, The Raspberry Pi contained the programs necessary to con- 186

167 which makes them highly energy efficient and reduces the trol the LoRa module to collect data packets sent from the 187

168 need for costly or frequent battery maintenance; Third, this LoRa enabled sensor and it also acted as storage for the 188

169 solution is relatively simple and inexpensive to implement collected data. The Raspbian operating system was used to 189

170 and scale. Based on experimental measurements, we quan- govern operations. A block diagram of the LoRaWAN gate- 190

171 titatively present the relevant channel qualities, network way is shown in Fig. 1. 191

VOLUME 10, 2022 102109

 F. F. Hossain et al.: Soil Moisture Monitoring Through UAS-Assisted IoT LoRaWAN Wireless Underground Sensors

192 The LoRa radio was a bonnet-style RFM95W radio from

193 Adafruit and plugged into the general purpose input/output
194 (GPIO) pins of the Raspberry Pi. Depending on environmen-
195 tal obstructions, antenna type, frequency and power output,
196 the range can approach approximately 2 km. For our pur-
197 pose, a half-wave dipole antenna having a center frequency
198 of 916 MHz was employed. It had a torus-shaped antenna
199 pattern with 6 dBi of gain, and its bandwidth covered the
200 entire ISM 900 MHz band. The 120 mm (4.72 inch) long
201 antenna is rugged and damage-resistant and occupies little
202 space.

203 2) LoRa ENABLED SENSOR HARDWARE

204 The LoRa enabled sensor or end node that was buried during
205 the experiment also had a Raspberry Pi 4 as its controller,
206 although a much simpler controller could be employed.
207 A block diagram of the LoRa enabled sensor is shown in
208 Fig. 2. Just like the mobile LoRaWAN gateway, an RFM95W FIGURE 3. Data being recorded in preliminary experiment.
209 LoRa radio module was plugged to the LoRa enabled sensor’s
210 Raspberry Pi 4 to enable sending data packets to the mobile
211 LoRaWAN gateway. The radio can output between +5 dBm unperturbed sidewall of the hole at a depth of approximately 246

212 and +23 dBm power, which is configurable from software. 30.5 cm. The hole was filled with soil so that the LoRa 247

213 A half-wave dipole antenna (center frequency 916 MHz) was enabled sensor and broadcasting antenna were under 30.5 cm 248

214 connected to the RFM95W module. The radio was configured (1 foot) of soil. The LoRa enabled sensor was powered from 249

215 to operate with a 125 kHz bandwidth. A spreading factor a car battery, whose 12 VDC supply was converted into 5 VDC 250

216 (SF) of 7 and a coding rate (CD) of 5 were chosen for the supply by a common USB charging pod. The DC power 251

217 RFM95W LoRa module - the same values were chosen for wires were run through the soil from the battery and volt- 252

218 the gateway LoRa module. A capacitive soil-moisture sensor age converter above the ground to the LoRa enabled sensor 253

219 manufactured by DFRobot (model SEN0193) was used to underground. 254

220 measure the moisture of the soil. The portable LoRaWAN gateway was fixed at the end of a 255

221 The capacitive soil-moisture sensor produces an analog wooden pole with the antenna having the same orientation 256

222 voltage between 0 to 3.3 volts depending on the moisture (co-polarized) as the LoRa enabled sensor’s antenna. The 257

223 content of the soil. The analog output of the sensor was LoRaWAN gateway was powered by a small, portable bat- 258

224 converted to digital bits via an analog to digital converter tery. Data sent from the buried sensor were received by the 259

225 (ADC-Pi, PIS-1352) that was then fed into the Raspberry LoRaWAN gateway and the received signal strength indica- 260

226 Pi. The ADC-Pi is a 17-bit ADC with eight channels for tors (RSSI) were logged as a function of several LoRaWAN 261

227 inputs. It can be configured to work at different data rates gateway positions relative to the end-node, both at different 262

228 such as 3.75 Sa/s (samples per second) at 17-bits resolution, distances and heights. Numerous distances were selected in 263

229 15 Sa/s at 15-bits, 60 Sa/s at 13-bits, or 240 Sa/s at 11-bits. both the south and west directions, all at several heights. 264

230 The ADC-Pi has two MCP3424 Microchip converters each In addition, to get an estimate of the maximum coverage 265

231 having four analog inputs. The ADC-Pi communicates with area of the buried LoRa enabled sensor, single-point mea- 266

232 the host Raspberry Pi via I2 C port and is powered by the GPIO surements were collected when the LoRaWAN gateway was 267

233 port. The data rate, resolution, and gain of the ADC-Pi can moved as far as possible to the north, south, east, and west 268

234 be selected or changed within software. For our purposes, directions until the data packets were no longer received. 269

235 we configured the ADC-Pi to operate at a sample rate of All the same measurements were repeated after changing 270

236 15 samples per second. the orientation of the LoRaWAN gateway antenna to east- 271

west, which put the transmit and receive antennas in a 272

237 B. PRELIMINARY MEASUREMENTS cross-polarized relative orientation. Fig. 3 shows an image 273

238 A plastic, water-tight enclosure was used to house the entire of the research team collecting data during the preliminary 274

239 LoRa enabled sensor except for the soil moisture sensor, experiment. 275

240 which was wired externally through a sealed gland so that

241 it could be inserted into the soil. The antenna inside the C. UAS-BASED MEASUREMENTS 276

242 enclosure was aligned on its linear axis in a north-south and After the preliminary experiment, the LoRaWAN gateway 277

243 horizontal orientation. An approximately 35 cm deep hole was mounted on a UAS. The UAS employed was the DJI 278

244 was dug and the LoRa enabled sensor was placed therein Phantom 4 Pro. This UAS has a maximum battery lifetime 279

245 with the soil moisture sensor inserted horizontally into the of approximately 30 minutes, a top speed of 72 kph, and a 280

102110 VOLUME 10, 2022

 F. F. Hossain et al.: Soil Moisture Monitoring Through UAS-Assisted IoT LoRaWAN Wireless Underground Sensors

FIGURE 4. LoRaWAN gateway mounted on UAS.

281 weight of 1,388 g. The UAS has an onboard GPS/GLONASS

282 module with an accuracy of ±0.5 m vertically and ±1.5 m
283 horizontally. As before, the LoRaWAN gateway consisted of
284 a Raspberry Pi, a portable battery, a separate GPS/GLONASS
285 module, and the antenna mounted to the front side of the
286 UAS. The UAS hardware setup is shown in Fig. 4. The
287 two GPS/GLONASS modules were used to synchronize
288 data between the UAS and LoRaWAN gateway. The UAS
289 provided flight logs with accurate position data and times-
290 tamps. The LoRaWAN gateway was programmed to synchro-
291 nize the time with GPS time. Once synchronized with GPS
292 time, the data received from the buried sensor’s transmitter
293 was timestamped with the LoRaWAN gateway’s time. When
FIGURE 5. Spiral flight path of the UAS viewed via Google Earth. Data
294 processing the data, the LoRaWAN gateway timestamps were collection at each flight level began at the center of the spiral and
295 matched to the UAS timestamps so that RSSI data from the proceeded outward.
296 LoRaWAN gateway could be accurately matched with corre-
297 sponding position data from the UAS’s flight log. The flight enabled sensor antenna orientation while measuring the RSSI 317

298 path of this system was uploaded to the UAS via Litchi [27] data. The maximum usable distance – beyond which packet 318

299 and resembled a grid of points in an outward spiral. This grid loss became 100% – was found to be greater than 36.5 m 319

300 was approximately 46 m × 46 m (150 ft × 150 ft) and the (120 feet). 320

301 UAS was flown at 0.91 m (3 ft) intervals from 0.91 m to Four heatmaps are shown in Fig. 6. Fig. 6(a) shows the 321

302 3.66 m (12 ft) height. All UAS-based measurements were RSSI values at different heights from ground and at different 322

303 performed with the LoRaWAN gateway antenna in co- and distances toward the west from the LoRa enabled sensor 323

304 cross-polarized orientations. Litchi allows the user to upload burial position. Fig. 6(b) shows the RSSI values at differ- 324

305 a CSV file of points to define a flight path. A Python script ent heights from ground and at different distances towards 325

306 was written to generate the coordinates at each height and the south from the LoRa enabled sensor burial position. 326

307 save them to individual files. A Google-Earth view of the Fig. 6(a) and Fig. 6(b) show that, at observation points not 327

308 flight path of the UAS is shown in Fig. 5. directly above the buried sensor, the RSSI value increases 328

with increasing height. It is noted that the observation window 329

309 III. RESULTS AND DISCUSSION is within 2 m height from ground surface and 5 m horizontal 330

310 A. PRELIMINARY RESULTS distance from the burial position of the sensor. However, 331

311 The received signal strength (RSSI) indicator data in dBm when the LoRaWAN gateway antenna is directly above the 332

312 units were recorded during the preliminary measurement LoRa enabled sensor antenna, the RSSI values decrease with 333

313 exercise. First the LoRaWAN gateway antenna orientation increase in height. At a particular height, the RSSI values 334

314 was co-polarized with the LoRa enabled sensor antenna mostly decrease with increasing horizontal distance from the 335

315 while recording the RSSI data. Next, the LoRaWAN gate- burial position of the LoRa enabled sensor. The mechanisms 336

316 way antenna orientation was cross-polarized to the LoRa behind this behavior are discussed in detail in the Comparison 337

VOLUME 10, 2022 102111

 F. F. Hossain et al.: Soil Moisture Monitoring Through UAS-Assisted IoT LoRaWAN Wireless Underground Sensors

FIGURE 6. (a) and (b) RSSI in dBm when both LoRaWAN gateway and LoRa enabled sensor antenna are co-polarized.
(c) and (d) RSSI in dBm when both LoRaWAN gateway and LoRa enabled sensor antenna are cross-polarized.

338 with Numerical Simulation section. In short, however, the and Fig. 7(b). Here, it is observed that for a particular height, 368

339 radiation pattern of the buried dipole near the air-soil interface the RSSI values decrease with increasing distance from the 369

340 appears to be the principal factor in this observed behavior. LoRa enabled sensor’s position. But, at a particular distance 370

341 Interestingly, these observations hold true for the measure- from the LoRa enabled sensor’s position, the RSSI values 371

342 ments when the antennas were cross-polarized to each other - generally increase with increase in height. However, directly 372

343 the data are shown in Fig. 6(c) and Fig. 6(d). Surprisingly, above the LoRa enabled sensor, the increase in height accom- 373

344 stronger RSSIs were frequently observed in a cross-polarized panied reducing RSSI values. These results are congruent 374

345 configuration. This suggests some significant mechanism of with the results of our preliminary measurement campaign. 375

346 polarization rotation or randomization by wave scattering.

347 Wave scattering does not occur significantly above ground,
348 so it further suggests that polarization randomization occurs C. PACKET LOSS ANALYSIS 376

349 from multiple scattering events either within the soil or with The measurements were then analyzed in terms of packet 377

350 the electronics surrounding the buried sensor antenna (or loss. The LoRaWAN gateway encountered a best case packet 378

351 both). In this case, the electronics packaging, soil composi- loss of 12% when the antennas were co-polarized and 10% 379

352 tion, soil compaction, and soil saturation would also likely when the antennas were cross-polarized. Packet loss can also 380

353 affect the polarization state of the wave exiting the soil. be quantified for different intervals of distances. Fig. 8(a) 381

354 Therefore, very careful measurements with tightly controlled and Fig. 8(b) show the percentage packet loss computed 382

355 soil parameters and system design would be required to iso- for various intervals of distances between the UAS-mounted 383

356 late this mechanism, which is beyond the scope of this work. LoRaWAN gateway and the buried LoRa enabled sensor. 384

Fig. 8(a) indicates that when the antennas were co-polarized, 385

percentage packet loss was no more than 15% for distances 386

357 B. DATA TAKEN FROM DRONE FLIGHT smaller than 61 m and after this distance the percentage 387

358 When the LoRaWAN gateway was mounted on the UAS, packet loss increases dramatically. From Fig. 8(b), it can be 388

359 the RSSI data were taken at heights of 0.91, 1.83, seen that when the antennas were cross-polarized, percentage 389

360 2.74, and 3.66 meters (3, 6, 9, and 12 feet) over an area packet loss was no more than 17% for distances smaller than 390

361 of approximately 2,090 m2 (22,500 square feet). This data 82 m. The horizontal distance is contributing the most to 391

362 logging experiment was carried out two times. During the the packet loss, compared to the height component, since 392

363 first time, the LoRaWAN gateway antenna and the LoRa the UAS never reached a height greater than 12 ft (3.66 m) 393

364 enabled sensor antenna were co-linearly polarized and during from the ground. Summarily, the LoRaWAN gateway and the 394

365 the second time, the antennas were cross-polarized. buried device could communicate without experiencing more 395

366 The LoRa enabled sensor’s longitude and latitude were than 17% packet loss at distances ≤61 m from each other. 396

367 normalized to zero to present the measured data in Fig. 7(a) This is more than adequate for the intended purpose. 397

102112 VOLUME 10, 2022

 F. F. Hossain et al.: Soil Moisture Monitoring Through UAS-Assisted IoT LoRaWAN Wireless Underground Sensors

FIGURE 7. 3D heatmaps of RSSI in dBm when (a) both LoRaWAN gateway and LoRa enabled sensor antennas were co-polarized and
(b) when the antennas were cross-polarized.

FIGURE 8. (a) Percentage packet loss in different intervals of distance between UAS and buried sensor when antennas
were in co-polarized orientation and (b) cross-polarized orientation.

398 D. BATTERY LIFETIME ANALYSIS transmission is set to 4.7 s (typical values), then over the 418

399 Another important question is the battery lifetime of the course of the five minute beacon mode, the total transmis- 419

400 buried LoRa enabled sensor. It could clearly be made sion time will be approximately 8.5 s. The current draw for 420

401 more power efficient if a low-power microprocessor was each component device was obtained from their respective 421

402 employed instead of a Raspberry Pi. The Texas Instruments data sheets. Using these currents, the radio’s delivered power 422

403 IC MSP430AFE2×3 is one example alternative. In principle, (+20 dBm) to the antenna, and supply voltage (3.3 V), the 423

404 the microprocessor could normally be in low-power standby average power draw by the LoRa enabled sensor can be 424

405 mode and only wake 24 times a day. Each awakening would estimated for both its active and dormant period. Under these 425

406 last for about one second to collect data from the soil sensor conditions and assuming a battery total capacity of approxi- 426

407 and store it in memory. The one exception would be when it is mately 9,000 J and a 1.5 V to 3.3 V converter circuit efficiency 427

408 also tasked with transmitting data to the LoRaWAN gateway. of 80%, the LoRa enabled sensor should be able to operate 428

409 The soil sensor itself can also be powered for only one second, without battery maintenance for about 3.5 years. It is noted 429

410 every time the microprocessor wakes up. In addition, the UAS that the duration the sensor node stays awake every time 430

411 can collect data from any one LoRa enabled sensor only once it wakes up and the number of times the sensor wakes up 431

412 a day, at a scheduled time. In this scenario, each LoRa enabled in a day can vary depending on the application, which will 432

413 sensor is awake for a total of less than 1 min per day. impact the battery lifetime. Moreover, the sensor nodes may 433

414 For a more conservative estimate of battery life, we can be required to wake up at different times on different days to 434

415 assume the radio RFM95W is active in beacon mode during send their stored data to the UAS-mounted gateway. As an 435

416 a five minute window every day for transmitting data. If the example, inclement weather may prohibit UAS flight on a 436

417 transmit time is set to 0.13 s and the idle time between each particular day. Data storage with on-board sensor memory 437

VOLUME 10, 2022 102113

 F. F. Hossain et al.: Soil Moisture Monitoring Through UAS-Assisted IoT LoRaWAN Wireless Underground Sensors

FIGURE 9. Measured and modeled path loss for the cases where the antennas are (a) co-polarized, (b) cross-polarized.

438 would usually permit this data to be collected later, but such and γ = 3.088. The high KdB values can be attributed to the 474

439 considerations remain important in a system-wide design. attenuation of the wave by the soil and the moisture in the soil. 475

440 In another application, the gateway could actively wake the In the ideal case, γ = 2, which indicates a path loss exponent 476

441 sensor nodes when they reach close proximity. Reference [28] of free space. The fact that our values of γ are greater than 477

442 discusses methods of waking up sensor nodes by a gateway two in both cases is a consequence of the transmitted signal 478

443 mounted on a UAS using radio frequency signals. Such vari- partially traveling through soil and also experiencing the 479

444 ations on the communication scheme could easily change effects of the soil-air boundary, as described further below. 480

445 battery life estimates by significant amounts and must be The γ value in the case of cross-polarized antennas is larger 481

446 considered for each particular application. than the co-polarized case which is expected. This means 482

that for a real-world application, keeping the transmitter and 483

447 E. PATH LOSS ANALYSIS receiver antenna in parallel will be statistically advantageous, 484

448 The loss of signal strength of a wave traveling between the although this benefit is surprisingly slight. In practice, the 485

449 transmitter and receiver is quantified by the path loss. At a polarization state of the antennas would have little significant 486

450 given distance d from the transmitter, the received signal effect. 487

451 power PR can be written as follows [29]:

452 PR = KPT (d0 /d)γ (1) F. COMPARISON WITH NUMERICAL SIMULATION 488

A numerical simulation was performed in COMSOL Mul- 489

453 where K is a dimensionless constant that is dependent on the tiphysics to obtain the radiated power pattern of the buried 490
454 transmission frequency and antenna characteristics, PT is the antenna and thereby better understand the mechanisms 491
455 transmitted power of the signal by the transmitter in watts, limiting performance. The soil properties assumed in these 492
456 d0 is a far-field (relative to the transmitter) reference distance simulations were comparable to those of the site of our pre- 493
457 in meters and γ is the dimensionless path loss exponent. K in liminary measurements. To get these properties, we entered 494
458 dB can be expressed as the coordinates of the site in the USDA Web Soil Survey 495

459 KdB = 10 log10 K (2) website [30] and obtained an estimate of the soil composition. 496

We also measured the volumetric water content of the soil on 497

460 The KdB value signifies the minimum loss at reference the site, which was found to be 41.4% at 12 inches depth into 498

461 distance. The path loss, PLdB may then be quantified as soil, obtained using a calibrated meter, Campbell Scientific 499
 
d Hydrosense II. Incidentally, this is quite wet and lossy soil, 500

462 PL dB = γ 10 log10 − KdB , (3) having just experienced rain in previous days. From these we 501
d0
obtained values of the complex relative√electric permittivity 502

463 where PL dB = 10 log10 (PT /PR ) is the path loss in decibels. as ˜r = 24.5 − j2.23 (where j = −1) by exploiting 503

464 In our case, d0 = 1.2192 m, which is the minimum measured the empirical equations in [31]. The complex relative mag- 504

465 distance between the UAS and the buried sensor. Separate netic permeability was assumed to be unity. The complex 505

466 path loss analyses were done for the cases when the antennas permittivity and permeability were assumed to be frequency 506

467 were co-polarized and cross-polarized. The measured and independent as our operating bandwidth was only 125 kHz. 507

468 modeled path losses versus distance are shown in Fig. 9(a) Consequently, the simulation was performed for only one 508

469 and Fig. 9(b). Using MATLAB’s curve fitting tool, the path frequency, i.e. 915 MHz. From the simulation, radiated power 509

470 loss versus distance lines were fitted in the measured data. was examined in two vertical planes (both perpendicular to 510

471 The curve fitting analysis shows that KdB = −71.54 dB and ground) - one plane is parallel to the antenna axis and contains 511

472 γ = 2.956 for the co-polarized antenna case. When the anten- the antenna axis center line while the other plane bisects the 512

473 nas were cross-polarized, the values were KdB = −67.02 dB antenna and is perpendicular to its axis. In both cases, the 513

102114 VOLUME 10, 2022

 F. F. Hossain et al.: Soil Moisture Monitoring Through UAS-Assisted IoT LoRaWAN Wireless Underground Sensors

FIGURE 10. (a) Antenna power pattern for the vertical plane parallel to the antenna. The solid black line at a height of zero meters is the
antenna. (b) Antenna power pattern for the vertical plane perpendicular to the antenna axis. The black dot at a height of zero meters is
the antenna.

FIGURE 11. (a) Antenna power pattern for the vertical plane parallel to the antenna. The solid black line at a height of zero meters is
the antenna. (b) Antenna power pattern for the vertical plane perpendicular to the antenna axis. The black dot at a height of zero
meters is the antenna.

514 antenna is oriented parallel to the soil-air interface. The power possible due to the close proximity of the antenna to the 537

515 patterns in the two planes are depicted in Fig. 10. In the interface. It is worth mentioning that the simulated power 538

516 figure, the antenna is at height of 0 m. It is surrounded by pattern inside the soil just next to the antenna is dominated 539

517 soil both above and below it for 0.3 m (1 ft), and at heights by near field components, which would normally die off 540

518 greater than 0.3 m the material is air. By looking at both and become insignificant in the far-field (air). The near field 541

519 patterns, it is apparent that the power is concentrated above components can act as energy storage (resonant) fields which 542

520 and surrounding the antenna. As the horizontal distance and leads to artificially inflated values of RSSI inside the soil next 543

521 height from the antenna increase, the power decreases rapidly. to the antenna using our simple power calculation method 544

522 This is particularly true in the soil where the high conduc- (P ∝ |E|2 ). 545

523 tivity induces strong absorption losses that lead to simu- In addition, Figs. 10 and 11 show that the power pattern 546

524 lated power values lower than −125 dBm within less than within the soil looks similar to that of a λ/2 dipole antenna. 547

525 1 m of propagation. This is shown more clearly in Fig. 11, On the other hand, it is reshaped in air by the discontinuity 548

526 which is the same data but with an increased color scale at the soil-air interface. Of particular note is that the critical 549

527 to clarify the behavior in the soil. There also appears to be angle interface produces strongly refracted waves that prop- 550

528 significant reflection losses at the soil-air boundary which are agate parallel to the interface on the air side (see Fig. 10(a)). 551

529 highly exacerbated near or beyond the critical angle formed This represents power flow in directions that would be 552

530 because of the smaller permittivity of air versus soil. Any forbidden from an unbounded dipole antenna (parallel to 553

531 plane wave incident upon the soil-air interface at an angle the antenna axis), which is possibly another benefit to this 554

532 greater than critical angle (about 10.7◦ in this case) suffers application. 555

533 complete reflection back into the soil. Nevertheless, there Simulation further permits a one-to-one comparison of 556

534 remains some signal coverage even just above the interface power flow results with the experimental results obtained 557

535 due to the strongly refracted waves. We point out that a in preliminary measurements. These corresponding power 558

536 strict plane wave interpretation of these phenomena is not patterns are compared in Fig. 12. 559

VOLUME 10, 2022 102115

 F. F. Hossain et al.: Soil Moisture Monitoring Through UAS-Assisted IoT LoRaWAN Wireless Underground Sensors

FIGURE 12. Comparison of (normalized) simulations and measured RSSI values from preliminary measurements. Only the case of
co-polarized antennas is shown. (a) Antenna power pattern from preliminary experimental data for the vertical plane parallel to
the antenna; (b) antenna power pattern from simulation for the vertical plane parallel to the antenna. (c) Antenna power pattern
from preliminary experimental data for the vertical plane perpendicular to the antenna; (d) antenna power pattern from
simulation for the vertical plane perpendicular to the antenna. All heights are measured from ground level and the antenna center
is at the horizontal zero position.

560 In Fig. 12, each simulated power pattern is normalized interface, complete reflection beyond critical angle, and large 587

561 to the maximum RSSI value obtained during preliminary soil losses. A perfect match between simulated and measured 588

562 measurements and then expressed in dBm. In both simulated power is certainly not expected. Some of the possible sources 589

563 cases (Fig. 12), the power values are calculated using only of error include the following. First, in the experiment the 590

564 the electric field component that is parallel to the antenna. buried antenna was not immediately surrounded by soil; 591

565 Looking at the power patterns in Fig. 12, it is seen that the rather it was inside an air-filled (albeit small) plastic enclo- 592

566 patterns are similar in the sense that all of them have high sure. Second, the buried antenna was placed next to other 593

567 power straight above the antenna and with the increase in electronics including a Raspberry Pi and ADC which may 594

568 height and horizontal distance from the antenna the observed have affected the radiation behavior. Finally, there may have 595

569 power decreases. been air pockets present in the soil above the buried enclosure 596

570 Similar comparisons can be made for the case of cross- containing the LoRa enabled sensor, an unavoidable result of 597

571 polarized antennas. But simulation results from COMSOL perturbing the soil during digging and burying. 598

572 for cross-polarized orientation of antennas were on average

573 45 dB lower than that of co-polarized antenna simulations, IV. CONCLUSION 599

574 which is to be expected due to the absence in simulation The concept and experimental verification of using a UAS 600

575 of any real-world perturbations to the wave polarization. mounted LoRaWAN gateway with completely buried LoRa 601

576 Since the cross-polarized simulation results would imply enabled sensors for soil monitoring is presented. Both 602

577 undetectable signal levels, in contrast to measurements, there manual and automated field measurements using a UAS 603

578 is no significant value to presenting this comparison. It is revealed highly reliable communications over large distances 604

579 notable, however, that the experimental results proved the (>36.5 meters), even with wet, lossy soil. Communication 605

580 system more robust than expectations derived from simula- was found to be largely insensitive to antenna polarizations 606

581 tion. In many cases, the measured RSSI was actually stronger as evidenced by total packet losses of no more than 12%, 607

582 than expected. regardless of antenna orientations. Having measured multiple 608

583 The similarity of the general trend in simulated power heights, these tests also prove that this concept is viable 609

584 patterns and experimental power patterns in Fig. 12 sug- even when the UAS needs to clear the heights of almost 610

585 gests that the experimental power patterns were the result of all kinds of crops. The simulations revealed that the signal 611

586 various mechanisms including wave reflections from soil-air crosses the soil-air boundary in a relatively narrow region 612

102116 VOLUME 10, 2022

 F. F. Hossain et al.: Soil Moisture Monitoring Through UAS-Assisted IoT LoRaWAN Wireless Underground Sensors

613 directly above the antenna, but this does not dramatically [9] I. F. Akyildiz and E. P. Stuntebeck, ‘‘Wireless underground sensor 672

614 reduce the viability of the approach, since waves can refract networks: Research challenges,’’ Ad Hoc Netw., vol. 4, no. 6, pp. 669–686, 673
Nov. 2006. 674
615 sharply to the horizontal at the soil/air interface. Practically, [10] N. Jack and K. Shenai, ‘‘Magnetic induction IC for wireless communica- 675
616 the narrow emission region is irrelevant when the LoRaWAN tion in RF-impenetrable media,’’ in Proc. IEEE Workshop Microelectron. 676

617 gateway can hover directly above the LoRa enabled sensors. Electron Devices (WMED), Apr. 2007, pp. 47–48. 677
[11] Z. Sun and I. F. Akyildiz, ‘‘Magnetic induction communications for 678
618 In addition, since LoRa enabled sensors can be buried at wireless underground sensor networks,’’ IEEE Trans. Antennas Propag., 679
619 least 0.3 meters (1 ft) underground, they are very unlikely to vol. 58, no. 7, pp. 2426–2435, Jul. 2010. 680

620 interfere with normal farming operations. Finally, the mea- [12] E. A. Abioye, M. S. Z. Abidin, M. S. A. Mahmud, S. Buyamin, 681
M. K. I. AbdRahman, A. O. Otuoze, M. S. A. Ramli, and O. D. Ijike, ‘‘IoT- 682
621 surements strongly suggest that LoRa enabled sensors can based monitoring and data-driven modelling of drip irrigation system for 683
622 be developed with battery lifetimes on the order of multiple mustard leaf cultivation experiment,’’ Inf. Process. Agricult., vol. 8, no. 2, 684

623 years, eliminating the need for regular maintenance. One pp. 270–283, Jun. 2021. 685
[13] A. N. Harun, M. R. M. Kassim, I. Mat, and S. S. Ramli, ‘‘Precision 686
624 disadvantage of this approach may be that a person may irrigation using wireless sensor network,’’ in Proc. Int. Conf. Smart Sensors 687
625 have to supervise the UAS flights. Although, if the UAS is Appl. (ICSSA), May 2015, pp. 71–75. 688

626 programmed to be autonomous, then this approach can be [14] M. C. Vuran, A. Salam, R. Wong, and S. Irmak, ‘‘Internet of underground 689
things: Sensing and communications on the field for precision agriculture,’’ 690
627 utilized with minimal human intervention. in Proc. IEEE 4th World Forum Internet Things (WF-IoT), Feb. 2018, 691
pp. 586–591. 692

628 V. FUTURE WORK [15] A. J. Wixted, P. Kinnaird, H. Larijani, A. Tait, A. Ahmadinia, and N. Stra- 693
chan, ‘‘Evaluation of LoRa and LoRaWAN for wireless sensor networks,’’ 694
629 In this work, one underground sensor node has been demon- in Proc. IEEE SENSORS, Oct. 2016, pp. 1–3. 695
630 strated to work in conjunction with a UAS mounted gateway. [16] J. de Carvalho Silva, J. J. Rodrigues, A. M. Alberti, P. Solic, and 696

631 In future, the performance of this approach can be evaluated A. L. Aquino, ‘‘LoRaWAN—A low power WAN protocol for Internet of 697
Things: A review and opportunities,’’ in Proc. 2nd Int. Multidisciplinary 698
632 when multiple underground sensor nodes are deployed in a Conf. Comput. Energy Sci. (SpliTech), Jul. 2017, pp. 1–6. 699
633 field. To collect the data from the underground nodes effi- [17] S. Subashini, R. Venkateswari, and P. Mathiyalagan, ‘‘A study on 700

634 ciently, the UAS may have to fly in a path that is energy LoRaWAN for wireless sensor networks,’’ in Computing, Communication 701
and Signal Processing. Singapore: Springer, Sep. 2019, pp. 245–252. 702
635 efficient and fast - to this end, optimization algorithms can be [18] M. A. Ertürk, M. A. Aydin, M. T. Büyükakkaşlar, and H. Evirgen, 703
636 exploited to optimize the flight path of the UAS. In addition, ‘‘A survey on LoRaWAN architecture, protocol and technologies,’’ Future 704

637 the performance of this approach can be investigated by bury- Internet, vol. 11, no. 10, p. 216, Oct. 2019. 705
[19] G. Di Renzone, S. Parrino, G. Peruzzi, A. Pozzebon, and D. Bertoni, 706
638 ing the sensor nodes at different depths underground. Further ‘‘LoRaWAN underground to aboveground data transmission performances 707
639 refinements could also improve the simulations by including for different soil compositions,’’ IEEE Trans. Instrum. Meas., vol. 70, 708

640 inconsistencies in the soil and including the surrounding pp. 1–13, 2021. 709
[20] D. Gao, Q. Sun, B. Hu, and S. Zhang, ‘‘A framework for agricultural pest 710
641 electronics, e.g., the Raspberry Pi and its housing. and disease monitoring based on Internet-of-Things and unmanned aerial 711
vehicles,’’ Sensors, vol. 20, no. 5, p. 1487, 2020. 712
642 ACKNOWLEDGMENT [21] J. L. Chavez, A. F. Torres-Rua, W. E. Woldt, H. Zhang, C. C. Robertson, 713

643 The authors thank Taylor Mitchell, Allan Burba, and Andrew G. W. Marek, D. Wang, D. M. Heeren, S. Taghvaeian, and C. M. Neale, 714
‘‘A decade of unmanned aerial systems in irrigated agriculture in the 715
644 Cole for their assistance with the UAS-based measurement Western U.S.,’’ Appl. Eng. Agricult., vol. 36, no. 4, pp. 423–436, Jan. 2020. 716
645 campaigns. [22] Y. Huang, K. N. Reddy, R. S. Fletcher, and D. Pennington, ‘‘UAV low- 717
altitude remote sensing for precision weed management,’’ Weed Technol., 718
vol. 32, no. 1, pp. 2–6, Feb. 2018. 719
646 REFERENCES [23] J. Valente, D. Sanz, A. Barrientos, J. del Cerro, Á. Ribeiro, and C. Rossi, 720
647 [1] B. Kashyap and R. Kumar, ‘‘Sensing methodologies in agriculture for soil ‘‘An air-ground wireless sensor network for crop monitoring,’’ Sensors, 721
648 moisture and nutrient monitoring,’’ IEEE Access, vol. 9, pp. 14095–14121, vol. 11, no. 6, pp. 6088–6108, Jun. 2011. 722
649 2021. [24] X. Ge, J. Wang, J. Ding, X. Cao, Z. Zhang, J. Liu, and X. Li, ‘‘Com- 723
650 [2] D. Suarez, ‘‘Soil salinization and management options for sustainable crop bining UAV-based hyperspectral imagery and machine learning algo- 724
651 production,’’ in Handbook Plant Crop Stress, M. Pessarakli, Ed., 3rd ed. rithms for soil moisture content monitoring,’’ PeerJ, vol. 7, p. e6926, 725
652 Boca Raton, FL, USA: CRC Press, 2010, ch. 3, pp. 41–54. May 2019. 726
653 [3] M. Anderson, ‘‘Chemigation,’’ in Central Plains Irrigation Short Course [25] S. Duangsuwan, C. Teekapakvisit, and M. M. Maw, ‘‘Development of 727
654 and Exposition Proceedings. Fort Collins, CO, USA: Colorado State Univ. soil moisture monitoring by using IoT and UAV-SC for smart farming 728
655 Libraries, Feb. 1997, pp. 53–55. application,’’ Adv. Sci., Technol. Eng. Syst. J., vol. 5, no. 4, pp. 381–387, 729
656 [4] S. Taghvaeian, A. A. Andales, L. N. Allen, I. Kisekka, 2020. 730
657 S. A. O’Shaughnessy, D. O. Porter, R. Sui, S. Irmak, A. Fulton, and [26] C. Cariou, L. Moiroux-Arvis, F. Pinet, and J.-P. Chanet, ‘‘Data collection 731
658 J. Aguilar, ‘‘Irrigation scheduling for agriculture in the United States: from buried sensor nodes by means of an unmanned aerial vehicle,’’ 732
659 The progress made and the path forward,’’ Trans. ASABE, vol. 63, no. 5, Sensors, vol. 22, no. 15, p. 5926, Aug. 2022. 733
660 pp. 1603–1618, 2020. [27] Litchi. (Apr. 2022). Litchi for DJI Drones. [Online]. Available: 734
661 [5] T. Lakhankar, N. Krakauer, and R. Khanbilvardi, ‘‘Applications of https://flylitchi.com/ 735
662 microwave remote sensing of soil moisture for agricultural applications,’’ [28] J. Chen, Z. Dai, and Z. Chen, ‘‘Development of radio-frequency sensor 736
663 Int. J. Terraspace Sci. Eng., vol. 2, no. 1, pp. 81–91, Jan. 2009. wake-up with unmanned aerial vehicles as an aerial gateway,’’ Sensors, 737
664 [6] S. Rawal, ‘‘IoT based smart irrigation system,’’ Int. J. Comput. Appl., vol. 19, no. 5, p. 1047, Mar. 2019. 738
665 vol. 159, no. 8, pp. 7–11, Feb. 2017. [29] A. Goldsmith, ‘‘Path loss and shadowing,’’ in Wireless Commun., 1st ed. 739
666 [7] D. Thakur, Y. Kumar, A. Kumar, and P. K. Singh, ‘‘Applicability of New York, NY, USA: Cambridge Univ. Press, 2005, ch. 2, p. 46. 740
667 wireless sensor networks in precision agriculture: A review,’’ Wireless Pers. [30] United States Department of Agriculture. (2022). Web Soil Survey. 741
668 Commun., vol. 107, no. 1, pp. 471–512, Apr. 2019. [Online]. Available: https://websoilsurvey.sc.egov.usda.gov/ 742

669 [8] J. Zhang, W. Li, and N. Han, ‘‘Forest fire detection system based on a [31] N. R. Peplinski, F. T. Ulaby, and M. C. Dobson, ‘‘Dielectric properties 743

670 ZigBee wireless sensor network,’’ Frontiers Forestry China, vol. 3, no. 3, of soils in the 0.3–1.3-GHz range,’’ IEEE Trans. Geosci. Remote Sens., 744

671 pp. 369–374, 2008. vol. 33, no. 3, pp. 803–807, May 1995. 745

VOLUME 10, 2022 102117

 F. F. Hossain et al.: Soil Moisture Monitoring Through UAS-Assisted IoT LoRaWAN Wireless Underground Sensors

746 FAHIM FERDOUS HOSSAIN received the B.Sc. JAMEY D. JACOB received the B.S. degree 800
747 degree in electrical and electronic engineering in aerospace engineering from the University of 801
748 from the Bangladesh University of Engineering Oklahoma, in 1990, and the M.S. and Ph.D. 802
749 and Technology, Dhaka, Bangladesh, in 2016. degrees in mechanical engineering from the Uni- 803
750 He is currently pursuing the Ph.D. degree with versity of California at Berkeley, in 1992 and 804
751 the Ultrafast Terahertz and Optoelectronics Labo- 1995, respectively. He is currently the Director of 805
752 ratory, Oklahoma State University, Stillwater, OK, the OSU Unmanned Systems Research Institute 806
753 USA. He was a Lecturer at the Electrical and Elec- and the John Hendrix Chair and a Professor of 807
754 tronic Engineering Department, Uttara University, aerospace engineering with the School of Mechan- 808
755 Dhaka, from 2016 to 2017. His research interests ical and Aerospace Engineering, Oklahoma State 809
756 include wireless communication, artificial electromagnetic materials, and University. His current efforts are focused on advanced air mobility and their 810
757 terahertz measurement systems. enhanced operation in the national airspace for broader innovative applica- 811
tions. He is currently lead on the NASA University Leadership Initiative Pro- 812
gram WINDMAP to develop aviation weather solutions for advanced aerial 813
mobility applications, including drones and urban air taxis, and the Director 814

758 RUSS MESSENGER received the B.Sc. and M.S. for the Counter-UAS Center of excellence. He was a National Research 815

759 degrees in electrical engineering from Oklahoma Council Summer Faculty Fellow at the Air Force Research Laboratory. 816

760 State University (OSU), Stillwater, OK, USA, in He received the SAE Ralph Teetor Award, the Lockheed Martin Teaching 817

761 2019 and 2021, respectively, where he is currently Award, and the OSU Regents Distinguished Teaching and Research Awards, 818

762 pursuing the Ph.D. degree in electrical engineering among other mentoring accolades. He is a native Oklahoman and dedicates 819

763 under his advisor Dr. John F. O’Hara. He is cur- much of his efforts to STEM workforce development, tribal engagement, and 820

764 rently working as a Graduate Research Assistant increasing diversity in engineering and science. 821

765 with the Ultrafast Terahertz and Optoelectronic

766 Laboratory, School of Electrical and Computer
767 Engineering, OSU. His current research interests
768 include the Internet of Things, intelligent reflecting surfaces, terahertz band
769 wireless communication, and terahertz time domain spectroscopy. SALEH TAGHVAEIAN received the B.S. and 822
M.Sc. degrees in irrigation engineering from the 823
Ferdowsi University of Mashhad, Iran, in 2003 and 824
2006, respectively, and the Ph.D. degree in irri- 825

770 GEORGE L. CAPTAIN is currently pursuing gation engineering from Utah State University, 826

771 the bachelor’s degree in electrical engineering in 2011. He was a Postdoctoral Fellow at Colorado 827

772 and computer engineering with a minor in com- State University, from 2011 to 2013. He then 828

773 puter science with Oklahoma State University. joined the Biosystems and Agricultural Engi- 829

774 He expects to graduate in the Fall Semester of neering Department, Oklahoma State University, 830

775 2022. He is currently an Undergraduate Research as a State Water Resources Extension Specialist. 831

776 Assistant with the Ultrafast Terahertz Optoelec- In August 2022, he joined the Biological Systems Engineering Department, 832

777 tronics Laboratory. During the summer of 2021, University of Nebraska, where he is currently serving as an Associate Profes- 833

778 he received the OK EPSCoR Research Experience sor. His research interests include sensor technologies for optimizing on-farm 834

779 for undergraduates grant to work on soil monitor- water management, irrigation scheduling, and crop water stress monitoring. 835

780 ing assisted by UAV.

781 SABIT EKIN (Senior Member, IEEE) received the JOHN F. O’HARA (Senior Member, IEEE) 836
782 B.Sc. degree in electrical and electronics engineer- received the B.S. degree in electrical engineering 837
783 ing from Eskisehir Osmangazi University, Turkey, from the University of Michigan, in 1998, and 838
784 in 2006, the M.Sc. degree in electrical engi- the Ph.D. degree in electrical engineering from 839
785 neering from New Mexico Tech, Socorro, NM, Oklahoma State University, in 2003. He was the 840
786 USA, in 2008, and the Ph.D. degree in electri- Director of Central Intelligence Postdoctoral Fel- 841
787 cal and computer engineering from Texas A&M low at Los Alamos National Laboratory (LANL) 842
788 University, College Station, TX, USA, in 2012. until 2006. From 2006 to 2011, he was with the 843
789 He has four years of industrial experience as a Center for Integrated Nanotechnologies (LANL), 844
790 Senior Modem Systems Engineer at Qualcomm and worked on numerous metamaterial projects 845
791 Inc., where he has received numerous Qualstar Awards for his achieve- involving dynamic control over chirality, resonance frequency, polariza- 846
792 ments/contributions on cellular modem receiver design. He is currently an tion, and modulation of terahertz waves. In 2011, he founded a Consult- 847
793 Associate Professor of electrical and computer engineering at Oklahoma ing/Research Company, Wavetech, LLC. In 2017, he joined Oklahoma 848
794 State University (OSU). He is also the Founding Director of the OSU State University, where he is currently an Assistant Professor and the 849
795 Wireless Laboratory (OWL). His research interests include the design and Jack H. Graham Endowed Fellow of engineering. His research interests 850
796 analysis of wireless systems, including mmWave and terahertz communica- include metamaterials, terahertz communications, the Internet of Things 851
797 tions in both theoretical and practical point of views, visible light sensing, applications, and photonic sensing technologies. He has around 100 pub- 852
798 communications and applications, non-contact health monitoring, and the lications in journals and conference proceedings. 853
799 Internet of Things applications. 854

102118 VOLUME 10, 2022