revisedv2

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/339938413
Wireless technologies for smart agricultural monitoring using internet of
things devices with energy harvesting capabilities
Article in Computers and Electronics in Agriculture · May 2020
DOI: 10.1016/j.compag.2020.105338
CITATIONS
READS
6
173
2 authors, including:
P. Spachos
University of Guelph
114 PUBLICATIONS 847 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
B. Debowski - MEng Project View project
LeftBrain-RightBrain Deep Hierarchical Model for NLP/NLU View project
All content following this page was uploaded by P. Spachos on 17 March 2020.
The user has requested enhancement of the downloaded file.
Wireless Technologies for Smart Agricultural Monitoring using Internet of
Things Devices with Energy Harvesting Capabilities
Sebastian Sadowskia , Petros Spachosa,∗
a School
of Engineering, University of Guelph, Guelph, Ontario, N1G 2W1, Canada
Abstract
Technological advances in the Internet of Things (IoT) have paved the way for wireless technologies to be used in
new areas. Agricultural monitoring is an example where IoT can help to increase productivity, efficiency, and output
yield. However, powering these devices is a concern as batteries are often required due to devices being located where
electricity is not readily available. In this paper, an experimental comparison is performed between IoT devices with
energy harvesting capabilities that use three wireless technologies: IEEE 802.11g (WiFi 2.4 GHz), IEEE 802.15.4
(Zigbee), and Long Range Wireless Area Network (LoRaWAN), for agricultural monitoring. Four experiments were
conducted to examine the performance of each technology under different environmental conditions. According to
the results, LoRaWAN is the optimal wireless technology to be used in an agricultural monitoring system, when the
power consumption and the network lifetime are a priority.
Keywords: Wireless Technologies, ZigBee, LoRa, WiFi, Internet of Things, Energy Harvesting, Smart Agriculture,
Monitoring, Solar power
1
1. Introduction
22
23
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
In the era of the Internet of Things (IoT), everyday
objects are equipped with microcontrollers and communication devices that can work together to help us transform our world for the better [1, 2]. A promising area
where IoT devices can alleviate many issues and provide promising solutions is in agriculture. When IoT
devices and Wireless Sensor Networks (WSN) are used
in agriculture, advanced farming techniques can be applied which is known as Precision Agriculture (PA) [3].
PA allows for a greater amount of control in the growing
of crops and the raising of livestock. By using technology to monitor crops, the efficiency can be increased
and costs can be reduced since more precise remedies
can be applied to crops [4]. IoT devices can be used
in monitoring systems consisting of nodes that interact
with the environment using sensors to gather real-time
information and transmit it to a control room for further
processing.
However, in every monitoring system [5], power consumption is often a top concern in order for the system
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
∗ Corresponding
author
Email address: [email protected] (Petros Spachos)
Preprint submitted to Computers and Electronics in Agriculture
45
46
to function properly. If a sensor node stops transmitting, data regarding its monitoring area would be missing and the system would no longer have accurate information. In PA applications with WSN, batteries are
used to power the sensor nodes while outside. This is a
major issue, since at some point either the battery needs
to be replaced or if possible, recharged. Due to the node
being outdoors, rechargeable batteries and energy harvesting devices can be used. Solar power is a promising
approach since it is readily available and it can be easily
harvested to allow for the sensor node to function for a
longer period of time.
At the same time, to optimize power consumption,
the power requirements of each component of the node
should be carefully examined. While IoT devices have
been of great value to society in the automation of tasks,
they often consume a large amount of energy [6, 7]. Energy consumption comes from various processes such
as sensing systems, application operating systems, and
the communication radio [8]. In order to improve energy efficiency, each of the individual processes needs
to be optimized [9]. When it comes to IoT devices,
processes such as the operating and sensing systems are
often based on the application requirements, making it
difficult to reduce their energy requirements.
March 4, 2020
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
Optimization to the device power consumption can
be achieved through an efficient communication radio.
Reducing the energy requirements of wireless communication can help to optimize the battery life of nodes.
In agricultural applications, IEEE 802.11 (WiFi) [10],
IEEE 802.15.4 (Zigbee) [11], and Long Range Wireless
Area Network (LoRaWAN) [12], are among the commonly used wireless technologies. Each technology has
unique characteristics. The proper selection of the wireless technology, can extend the lifetime of the nodes and
eventually extend the lifetime of the network.
At the same time, PA can be greatly benefitted from
the use of proper wireless technology with rechargeable
batteries and IoT devices with energy harvesting capabilities. New technologies and IoT devices have revolutionized the way farmers are able to interact and monitor their growth. By combining traditional approaches,
such as energy harvesting techniques, with low cost,
low power and inch scale IoT devices, PA can be performed. A promising solution toward achieving PA is
the use of an IoT system with energy harvesting capabilities. The IoT uses small, low power embedded electronics that transmit data across a network. Often, when
sensor nodes are configured and placed outdoors in a
field, a power source is required. When electricity is
not available, batteries must be used. Due to the need to
replace batteries once depleted, rechargeable batteries
are an optimal alternative.
Towards such a system, there is a need for detailed
reports on the energy requirements. Real experimental
results on the energy requirements of different technologies are needed for the proper selection of the different
components and technologies. This work tries to fill this
gap. In this paper, through extensive experimentation,
a comparison between the power consumption of three
wireless technologies: IEEE 802.11g (WiFi 2.4 GHz),
Zigbee, and LoRaWAN is performed. Narrowband-IoT
(NB-IoT) is another popular technology for IoT application however, its associated cost makes it less popular than the previous three technologies, while SigFox, although it is similar to LoRaWAN is not widely
available. The technologies were selected based on
the prevalence and popularity in agricultural applications [13]. Their energy requirements are compared
when they are used at an agricultural monitoring system
using IoT devices with solar energy harvesting capabilities. Three systems were created each performing similar tasks while using different wireless technologies.
Non-line-of-sight exists between the sensor and the base
station. Four experiments were conducted and the energy requirement of the system was modeled, while an
accurate estimation of the lifetime is also provided. The
108
experimental results can be used for the selection of the
proper wireless technology and IoT devices for agricultural monitoring following application requirements.
The rest of this paper is organized as follows: Section 2 reviews the related work on wireless technologies
in agricultural applications. In Section 3, the system architecture is presented, followed by Section 4, with a
description of the experimental procedure. The experimental results and a discussion are presented in Section 5. Finally, Section 6 concludes this work.
109
2. Related Work
99
100
101
102
103
104
105
106
107
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
2
Over the years, applications using IoT devices and
WSN in PA have increased in popularity [14, 15, 16,
17]. By using battery-powered sensor nodes combined
with traditional farming practices, an increase in output efficiency and a reduction in costs can be achieved.
While some researchers have focused on implementing
energy harvesting to extend sensor node lifetime, others have modified the sensor nodes to use less energy
through its standard operating procedure [18].
In [19], a survey was performed studying the lifetime
of WSN and the energy saved with different types of
network topologies. According to the results, there are
many problems and issues when selecting a topology
for extending the network lifetime. In order for the network lifetime to be extended trade-offs must occur and
other parameters are required to be sacrificed. It was
suggested that energy-efficient articles be developed to
optimize energy supply.
In [20] and [21], systems were proposed for agricultural monitoring. In [20], the designed system used
WiFi-based IoT devices to monitor nitrate concentrations in groundwater. In the design of the system, WiFi
was selected for communicating the information due to
its low cost, high throughput, and ease of integrating
with web-based services. In [21], a WSN for irrigation
control was developed using Zigbee for wireless communication. Zigbee was selected due to its low cost and
off-the-shelf components to reduce the hardware complexity. In order to reduce the power consumption the
transmit power was configured to be 0 dBm. In both of
the systems presented, the power consumption was not
a major concern, with a greater focus being placed on
the total cost and ease of integrating with the system.
In [22], a study was performed on sensor nodes with
solar energy harvesting capabilities. An energy management policy is used in order to produce the optimal throughput and minimizes the mean delay in the
network. Another method of optimizing power consumption can be seen in [23]. A circuit was created
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
for wireless sensor nodes where energy from a solar
panel could be transferred to a rechargeable battery even
in poor weather conditions. The sampling rate of the
nodes can also affect the energy consumed and the supplied power. In [24], a method is presented to decide
the sampling rate of sensor nodes to manage its energy
efficiently. Simulations demonstrated the efficiency of
the proposed algorithm compared to other algorithms.
The system presented in [25] used a WSN in a cotton
field to monitor soil moisture with automatic drip irrigation. Sensor nodes were developed to function using
battery power while relay nodes functioned using solar
power. A routing protocol was used in order to route
data and increase power savings. Experimental results
were conducted over a six month period and demonstrated that the system could function for a long period
of time while collecting sensor information.
There have been many works that have used renewable energy sources for agricultural applications to extend network lifetime. In [26], due to the unpredictability associated with weather conditions, solar energy harvesting was combined with wireless charging in order
to allow for nodes to function for longer periods of
time. By combining the advantages of both solar energy harvesting and wireless charging it was found that
a significant increase in network performance could be
achieved.
In this work, we expand on the papers described
above and our previous work [27] and design a wireless IoT system for agricultural monitoring with energy
harvesting capabilities. There is a lack of research performed on using different types of wireless technologies
for agricultural applications. In search of a system design, a comparison is performed between three wireless
technologies: WiFi, Zigbee, and LoRaWAN to determine which technology consumes the lowest amount of
power and provides the longest network lifetime. Three
systems, each using one of the wireless technologies to
communicate, are developed and tested. The experimental results can fill the gap and provide real experimental values on the energy requirements of each technology.
215
they just forward the data towards the destination. The
third type of node, the destination node, has only a communication unit that receives all the data and stores them
in a server for further processing. In this work, we focus
on measuring the energy requirements of three wireless
technologies to be used as the communication unit of
each node.
Each monitoring node has a number of hardware
components. It contains an Arduino Uno, a power converter, one rechargeable battery, a solar panel, a soil
moisture sensor, and a communication unit. The monitoring nodes forward the sensor data to the destination
node using the wireless technology that we measure every time. For WiFi, a CC3000 WiFi Shield was used, for
Zigbee, a Series 2 XBees with a 2 mW Wire Antennas
was used, and for LoRaWAN, a Dragino LoRa Shield,
was used.
216
3.1. Components
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
191
3. System Architecture
239
240
192
193
194
195
196
197
198
The proposed system has three types of nodes. The
monitoring nodes have sensors to collect information,
they form the packets and then forward them to the relay nodes. The monitoring node architecture is shown
in Fig. 1. The relay nodes receive and transmit the data
packets. They have a similar architecture with the monitoring nodes, however, they do not have any sensors and
241
242
243
244
245
246
247
3
• Solar Panel: To provide energy harvesting capabilities to the node, a Star Solar D165X165
monocrystalline solar panel was used [28]. Being
only 170 × 170 × 2 mm, the solar panel is capable of providing a 6.0 V output at a peak of 3.65 W
when full sunlight is present. The small size makes
it suitable for placement in a field where it would
have minimal interference to any of the growing
plants surrounding it while still providing a significant energy output. The solar panel is shown in
Fig. 2a.
• Grand-Pro Lithium Polymer Battery: A Grand-Pro
3.7 V 6600 mAh Lithium Polymer (LiPo) battery
was used [29]. The battery is shown in Fig. 2b.
• Power Converter: A developed power converter
was used to supply power to the wireless node [30].
When connected to the battery, the power converter
was designed to provide a constant 5 V power
output while the charge on the battery was above
3.4 V. If the charge dropped below 3.4 V, the power
converter would cease to function and would wait
until the battery was sufficiently charged before
supplying power again. This safety feature allowed
the battery to maintain a voltage level preventing
it from over-discharging and damaging the battery
cells. In addition to recharge the battery, the converter was also capable of interfacing with an energy harvesting device. The energy harvesting device could then be used to supply power to the node
and recharge the battery. If it was no longer providing power to the battery, it could then supply power
Figure 1: Monitoring node architecture used for experimentation.
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
to the system. The power converter is shown in
Fig. 2c.
• Microcontroller: In order to connect all the hardware components together, an Arduino Uno Rev3
microcontroller was chosen [31]. It was selected
based on its low power consumption and ease of
development in configuring all the components together. It is based on the ATmega328P, which contains six analog-to-digital converts that can be used
to easily connect and read data from analog sensors. The Arduino Uno is shown in Fig. 2d. Other
microcontrollers, such as the Arduino MKR series can also be used. The energy requirements of
each component were measured separately, hence,
if the microcontroller is changed, the overall energy requirements can be easily estimated. At the
same time, it worths to mention that, during the
measurement of the energy requirements of each
wireless technology, the microcontroller energy requirements were subtracted from the total energy
consumption of the system.
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
269
270
271
272
273
274
275
• Grove Soil Moisture Sensor: To measure the moisture levels in the soil a Grove Soil Moisture Sensor
was used [32]. Soil moisture is a commonly measured parameter in agricultural monitoring allowing for a system design similar to what would be
used in a real-life application. The selected sensor
draws a significant amount of current and reduc-
297
298
299
300
301
302
303
4
ing the device lifetime when environmental conditions are being measured. This was not an issue
for our design since we focused on the energy consumption of the wireless technologies, however, it
is something that can be optimized in future implementations. The soil moisture sensor is shown in
Fig. 2e.
• CC3000 WiFi Shield: To connect the Arduino using WiFi a Sparkfun CC3000 WiFi Shield was
used [33]. WiFi is one of the most commonly used
wireless technologies, available in most devices,
used to connect to a Wireless Local Area Network
(WLAN) and the Internet. WiFi has a short transmission range in Line-of-Sight (LoS) only capable
of reaching up to 50 m distance. In addition, WiFi
has a very large power consumption which often
makes it a poor choice to use in wireless devices
outdoors that require a power supply. The CC3000
communicates using the IEEE 802.11g standard.
The WiFi shield is shown in Fig. 2f.
• Series 2 XBee with 2 mW Wire Antenna: To create a Zigbee network between the devices the Series 2 XBees with a 2 mW Wire Antennas were
used [34]. The Series 2 XBees are low power
radios which communicate on the Zigbee mesh
network. These devices are capable of creating
point-to-point or multi-point networks connecting
together hundreds of nodes. Devices using Zig-
(a) Solar Panel.
(e) Soil moisture sensor.
(b) LiPo Battery.
(c) Power Converter.
(f) WiFi (2.4Ghz).
(g) Series 2 XBee.
(d) Arduino Uno.
(h) Dragino Lora.
Figure 2: Hardware components used for experimentation.
304
305
306
307
308
309
bee have a transmission range up to 120 m in LoS,
which can provide many benefits when it is used in
an agricultural monitoring system such as reducing
the system costs and allowing for easy configuration of the devices. The Series 2 XBee is shown in
Fig. 2g.
325
326
327
328
329
330
331
332
333
334
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
• Dragino LoRa Shield: To communicate devices
using LoRaWAN the LoRa Shields for Arduino developed by Dragino were used [35]. LoRaWAN
is known for being a long-range technology communicating at a low frequency of 915 MHz (in
North America), and the signals produced have
large wavelengths, hence, they can travel further
distances. In LoS, LoRaWAN is capable of transmitting up to 15000 m, which makes it one of
the best technologies for agricultural monitoring.
Its large transmission range can greatly reduce the
number of nodes required and its low power consumption can keep nodes functioning for a longer
period of time compared to more commonly used
technologies. The LoRa Shield is shown in Fig. 2h.
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
5
3.2. System Characteristics
When comparing different types of wireless technologies, parameters such as the power consumption and the
transmission range are important in determining the optimal technology for agricultural monitoring. While the
power consumption of a wireless technology is important in determining the longevity of a node’s power supply, other parameters should be considered as well, in
the selection of the proper communication technology.
The transmission range is among the crucial parameters.
By using devices that transmit further, a smaller number of relay nodes are required in order for a transmission to reach the intended destination. Another common
parameter is the throughput. When there is a need for
a large amount of data to be transmitted, a technology
with a high throughput should be used. Other parameters often used are the cost and ease of implementation.
Having a low cost per-device can allow for a large number of nodes to be implemented in the network while
having a simple and easy implementation can allow for
the network to have a low set up time and make it easier
to debug if a problem occurs.
A summary of the parameters of the wireless technologies being compared in this paper can be seen in Table 1. For transmission range, LoRaWAN is the optimal
Table 1: Summary of wireless technologies and their characteristics.
Wireless
Technology
WiFi
(2.4GHz)
Zigbee
LoRaWAN
Throughput
Transmission Power ConRange
sumption
50 m
Moderate
Wide availability
High energy consumption
250 kbit/s
120 m
Easy to set up
Requires extra hardware
50 kbit/s
15000 m
Low
Extremely
Low
Wide range
Requires extra hardware
Parameter
Battery Current Supply
Arduino Uno (Max. Current Consumption)
Grove Soil Moisture Sensor (Max.
Current Consumption)
Series 2 XBee (Max. Current Consumption)
Dragino LoRa Shield (Max. Current
Consumption)
CC3000 WiFi Shield (Max. Current
Consumption)
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
Disadvantages
54 Mbit/s
Table 3: Estimated max. current consumption and min. lifetime of
monitoring node for each wireless technology.
Table 2: System parameters corresponding to components used in
monitoring nodes.
350
Advantages
Value
Wireless
Technology
6600 mAh
WiFi-based
node
Zigbee-based
node
LoRaWANbased node
45 mA
35 mA
40 mA
10 mA
1 Hz
Tranmission Interval
1s
Transmission Power
-10 dBm
capable of reaching up to 15000 m in LoS. Zigbee was
the second furthest in LoS with 120 m, while WiFi has
the lowest transmission range only capable of reaching
50 m in LoS. In terms of throughput, WiFi can transmit
the most amount of data reaching speeds of 54 Mbit/s.
Zigbee is the next highest with 250 kbit/s, followed by
LoRaWAN with 50 kbit/s.
At the same time, parameters such as the current consumption, the sampling frequency, transmission interval, and transmission power are important in measuring
the power consumption of a device. Table 2, summarizes the current supplied by the battery, maximum current utilized by the various components, and parameters
such as the sampling frequency, the transmission interval, and the transmission power that affect the power
consumption. The power values were measured with
the Monsoon power monitor.
In addition to the current draw of the components,
while greatly affecting the power consumption, the
270
24
120
55
90
73
Table 4: Estimated relay node current consumption and lifetime for
each wireless technology.
190 mA
Sampling Frequency
Estimated Max.
Estimated
Current
Min.
Consumption (mA) Lifetime (h)
Wireless
Technology
WiFi-based
node
Zigbee-based
node
LoRaWANbased node
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
6
Estimated Current Estimated
Consumption (mA) Lifetime (h)
190
35
40
165
10
660
transmission ranges of the wireless technologies are important when designing a system for agricultural monitoring. When the devices have a large monitoring range,
a smaller number of nodes can be used in the monitoring
of a field. A monitoring node would require an Arduino
Uno, a sensor, a battery, and a communication unit. Table 3 shows the expected lifetime of each monitoring
node when each wireless technology is used with maximum current consumption. Again, the Monsoon power
monitor was used for the energy measurements.
At the same time, when the nodes act as relay nodes
to forward the data, the larger the transmission range
the smaller the number of the required units to reach
the destination. To achieve the maximum transmission
range, the maximum current consumption is required.
Table 4 shows the estimated lifetime, when each of the
423
transmit the information every 1 s. Note that these
times were used in order for the systems to consume
a larger amount of power and therefore cease functioning sooner. If the systems were to be placed in an actual
environment for agricultural monitoring the times could
be greatly reduced since actual conditions do not rapidly
vary in a short period of time. Before starting the experiments, the batteries were fully charged.
424
4.2. Outdoor Experiments
416
417
418
419
420
421
422
425
426
Figure 3: Green roof lab at University of Guelph.
427
428
385
386
387
communication units is used as a relay node, and forwards the sensor data from the monitoring node towards
the destination.
429
430
431
432
399
3.3. Cost Analysis
In the design of a system using prototypes, the cost
is an important factor. For an analysis of the cost for
each of the nodes, the unit price for a single component is considered. The total cost for the design of the
monitoring nodes was found to be $155.59 USD, the
most expensive out of all the three nodes. For the relay node, the total cost was found to be $142.70 USD,
and the cheapest was found to be the destination costing $87.85 USD. While the cheapest, the price does not
include the cost of a computer or server that would be
needed in order to store and process the data.
400
4. Experimental Procedure
388
389
390
391
392
393
394
395
396
397
398
433
434
435
436
437
438
439
440
441
402
403
404
405
406
407
408
409
410
411
412
413
414
415
4.1. Experimental Setup
For testing purposes, nodes were configured to sample the charge left on the battery every 1 Hz and
• Experiment 2 - With energy harvesting. The solar
panel was connected to each node. This experiment took place during August 2018.
• Experiment 3 - With energy harvesting. The solar
panel was connected to each node. This experiment took place during December 2018.
• Experiment 4 - With energy harvesting. The solar
panel was connected to each node. This experiment took place during May 2019.
454
455
5. Results and Discussion
443
444
To evaluate the proposed system, extensive experimentation was conducted. For each wireless technology, identical nodes were configured. The experiments
were conducted at an outdoor environment where the
solar panels for each of the nodes would obtain a similar
amount of solar energy throughout the day. The testing
area was a roof research lab at the University of Guelph
Engineering Building, shown in Fig. 3. To measure the
charge left on the battery, probes from the power converter were connected and measured on the Arduino and
transmitted to a computer that was functioning as the
destination.
• Experiment 1 - No energy harvesting capabilities.
In this experiment, the solar panel was not connected to the node, to examine and characterize the
performance of the battery alone, without the solar
panel.
Due to uncontrollable weather conditions performing
three experiments would guarantee results that demonstrate the system performing with varying amounts of
sunlight. The current consumed by the nodes was also
measured. In order to measure the current consumption
of the devices, the Monsoon power monitor was used.
Monsoon is a monitoring tool that is capable of supplying an input voltage, measuring the current drawn by the
device, and can display the average measurements. To
measure the current consumption of the devices, nodes
were first powered and warmed up until the system was
fully operational. The current was then measured for
two minutes and the values were recorded.
442
401
Four experiments were performed, and each experiment lasted until the power supply in one of the monitoring nodes was drained.
445
446
447
448
449
450
451
452
453
456
457
7
In this section, the experimental results are presented
followed by a discussion.
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
5.1. Results
According to the experimental results obtained, the
system functioned as required with each wireless technology, capable of transmitting information until the
battery of a monitoring node was depleted. Due to the
large amount of data that was gathered throughout the
experiments, only a fraction is shown in the following
figures.
The voltage charge remaining on the battery over
time for the first experiment, with no energy harvesting
capabilities, can be seen in Fig. 4, while Fig. 5, Fig. 6,
and Fig. 7 shows the remaining battery levels over time
for experiments 2, 3, and 4, respectively. An overall
summary of the results and a comparison of the technologies can be seen in Table 5.
The average current consumption of the different devices along with the estimated lifetime was also measured. In terms of current consumption, WiFi consumed the most requiring 171.17 mA, followed by Zigbee which required 69.36 mA and lastly, LoRaWAN
which uses the lowest amount of current by consuming
only 29.33 mA, on average. Using these values along
with the battery size of 6600 mAh, the expected lifetime
of the devices could be calculated. A LoRaWAN based
system would be expected to run for 225.00 h, a Zigbeebased system should function for 95.15 h, while a WiFibased system has an expected lifetime of 38.56 h. According to the experimental results, as seen in Fig. 4,
the LoRaWAN system lasted for 166.23 h before failing, the Zigbee system functioned for 80.28 h, and WiFi
stopped after 29.06 h.
During the second experiment, shown in Fig. 5, again
the device using LoRaWAN technology was the most
optimal capable of lasting 228.20 h on a single battery charge. The Zigbee-based device was the second
to cease operating stopping after 95.15 h. WiFi was determined to be the worst operating technology in terms
of energy consumption, only functioning for 38.56 h.
In the third experiment, shown in Fig. 6, the results
are similar to the previous experiments with fewer sunlight hours, forcing the nodes to use their batteries for
power. This experiment took place during December
2018. The LoRaWAN-based system saw a large reduction in lifetime only functioning for 174.64 h, the
Zigbee-based system was able to last 92.19 h, while the
WiFi system experienced a similar runtime of 30.67 h.
Results from the fourth experiment can be seen in
Fig. 7. This experiment took place during May 2019
and the solar panels manage to collect a greater amount
of energy than the third experiment, but less than the
amount gathered from the second experiment. Using LoRaWAN the system was capable of running
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
8
for 189.22 h, the Zigbee system only functioned for
88.25 h, and the WiFi-based system stopped after
28.85 h.
5.2. Discussion
Through the experiments performed, it can be seen
that LoRaWAN is the optimal technology in terms of energy consumption for communicating information between nodes in a wireless IoT system with energy harvesting capabilities. After measuring the current consumption of the different devices, it could be seen that
LoRaWAN nodes consumed over a fifth the amount
of current of WiFi nodes and over half the current of
Zigbee nodes. Hence, this allows for LoRaWAN to
last a much longer duration using batteries with similar capacities. The benefits LoRaWAN provides can be
easily observed when a battery is selected to compare
the lifetimes of the nodes. With a 6600 mAh battery,
LoRaWAN system is expected to last approximately
225.00 h, which is much greater than the 95.15 h Zigbee
system can provide, or the 38.56 h obtained from WiFi
system.
When compared with the first experimental results,
it can be determined that the estimated time is not always accurate. For instance, LoRaWAN system functioned for only 166.23 h, Zigbee system was only capable of achieving a runtime of 80.28 h, and WiFi system runs for 29.06 h. Based on the real-world results,
a great difference can be noticed between the estimated
and the actual lifetime. One reason for the difference
can be attributed to the nonlinear discharge rate of the
LiPo batteries. With all batteries not being ideal drops
in the charge can occur reducing the lifetime of the device [36, 37, 38]. Another factor is the power converter.
The power converter used to supply power to the device from the battery was designed to prevent the battery
from over-discharging. Therefore, the power converter
would stop the supplying power once the charge on the
battery reached 3.4 V.
To improve the battery life of the devices, the next set
of experiments saw the addition of a solar panel to provide energy harvesting capabilities to the devices. According to the experimental results, adding energy harvesting to a system can greatly increase the lifetime of
the nodes in the network. At the same time, the amount
of sunlight obtained will greatly affect the additional
lifetime that the device will be able to function. The
second experiment saw a large amount of sunlight supplying energy to the devices, with the third experiment
supplying a very little amount of energy, and the fourth
experiment providing energy to be between the previous
two experiments.
(a) WiFi.
(b) Zigbee.
(c) LoRaWAN.
Figure 4: Experiment 1 - No energy harvesting capabilities.
(a) WiFi.
(b) Zigbee.
(c) LoRaWAN.
Figure 5: Experiment 2 - August 2018 with solar energy harvesting.
(a) WiFi.
(b) Zigbee.
(c) LoRaWAN.
Figure 6: Experiment 3 - December 2018 with solar energy harvesting.
(a) WiFi.
(b) Zigbee.
Figure 7: Experiment 4 - May 2019 with solar energy harvesting.
9
(c) LoRaWAN.
Table 5: Summary of results.
Wireless
Average Current
Estimated
Technology Consumption (mA) Lifetime (h) Experiment 1
WiFi
171.17
38.56
29.06
Zigbee
69.36
95.15
80.28
LoraWAN
29.33
225
166.23
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
At the LoRaWAN-based system, a larger amount of
variance between the runtimes could be observed. The
second experiment saw the node run for 228.20 h, the
third for 174.64 h, and the fourth for 189.22 h. Due to
the long base runtime of the device, it can be noticed
that a larger amount of solar harvesting could occur further increasing the runtime of the device. At its peak,
the second experiment saw the device last for an additional 50 h before run out of energy, while in the third
experiment the solar panel provided very little benefits
to the system.
At the Zigbee-based system, solar harvesting was
able to greatly increase the lifetime of the node during the experiment. The second experiment saw Zigbee
system run for 104.80 h, 92.19 h during the third experiment, and 88.25 h in the fourth experiment. When
compared to the first experiment, with energy harvesting a Zigbee node could last for an additional 25 h with a
large amount of sunlight, while for a low amount lasted
for 88.25 h. This is a great improvement for the Zigbee system as being able to function for a greater period
would allow for more data to be gathered before the battery in the node would need to be recharged or replaced.
Lastly, the WiFi-based system using energy harvest
has very little impact on the runtime of the device. During the second experiment, the node runs for 28.3 h,
while in the third experiment for 30.67 h, and for
28.85 h in the fourth experiment. Overall, the largest
impact gained from energy harvest was approximately
1 h. A system using WiFi consumed too much power
draining the battery charge that using energy harvesting
will have almost no impact on the system.
In the WiFi experiments, it can be seen that the solar panel provided little benefits. There was a very
little amount of battery charge recovered over the period that the node functioned. For the Zigbee and LoRaWAN systems, the solar panel provided more benefits
and made a bigger difference in the system lifetime.
Several other parameters such as the sampling frequency, transmission interval, and transmission power
could also affect the estimated and actual lifetime of the
systems. To decide how much each of the parameters
affects the power consumption additional experimenta-
Experimental Node Lifetime (h)
Experiment 2
Experiment 3
28.3
30.67
104.8
92.19
228.2
174.64
Experiment 4
28.85
88.25
189.22
618
tion would be necessary.
According to the experimental results in Table 5 and
the previous discussion on the characteristics of each
technology as shown in Table 1, WiFi is ideal if a large
amount of information is required to be transmitted between short distances. However, at the cost of such a
high speed, a greater amount of power consumed. On
the other hand, LoRaWAN has a much lower throughput
but it is able to transmit far distances with the minimal
amount of power being consumed. In the middle there
is Zigbee. It has a slightly higher throughput than LoRaWAN, but a greatly reduced transmission range. The
power consumed by Zigbee is still low and a network
can be easily set up with nodes capable of being easily
configurable and meshed together in the network.
619
5.3. Open issues
604
605
606
607
608
609
610
611
612
613
614
615
616
617
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
10
In this work, we discussed experimental results on the
energy requirements of three popular wireless technologies for PA with IoT devices. The results can be used as
an indicator for the selection of the proper wireless technology. However, there are still open issues that should
be addressed in order to improve the performance of an
IoT-based system for agriculture. The proper selection
of the sensor is an important parameter that can affect
the overall energy consumption of the system as well
as the cost of the system. Sensors with high accuracy
can be helpful while they can also increase the total cost
and the energy requirements. On the other hand, less
expensive sensors, with lower power requirements but
less accurate might affect the accuracy of the system.
Moreover, factors such as the deployment environment
and the environmental conditions should always be considered. In a harsh environment, the performance of the
solar panel might not be as expected, while the cost of
the protective case for each node should also be considered. At the same time, the cost of the microcontrollers
keeps decreasing. Hence, advance filtering techniques
can be applied without extra energy requirements, to
clean the data locally at each node and minimize the
need for wireless data transmission.
644
6. Conclusions
699
700
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
In this paper, we provide an experimental analysis between three wireless technologies: Zigbee, LoRaWAN,
and WiFi when they are used in an agricultural monitoring system with energy harvesting capabilities. Three
identical systems were created each using one wireless
technology. The systems were placed outdoors and the
batteries could be recharged through solar panels. The
systems were compared on the lifetime of the nodes,
where the node that functioned for the longest time
would be the most optimal for an agricultural application. Experimental results demonstrated that a LoRaWAN system would be ideal as it was capable of
functioning for the longest period before the first monitoring node stops transmitting due to lack of energy.
Zigbee system was the next ideal, followed by WiFi system.
However, power consumption and device lifetime
are usually not the only parameters that are considered
when designing a system. While WiFi has a poor power
consumption, the throughput is much higher allowing
for a larger amount of information that can be transmitted between devices. The results produced in the paper
can be used as an indicator for the selection of a wireless
technology to be used in an agricultural monitor system
with energy harvesting capabilities.
701
[9]
702
703
704
[10]
705
706
707
[11]
708
709
710
711
[12]
712
713
714
715
[13]
716
717
718
719
[14]
720
721
722
723
[15]
724
725
726
727
[16]
728
729
730
731
732
[17]
733
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
References
[1] A. Al-Fuqaha, M. Guizani, M. Mohammadi, M. Aledhari,
M. Ayyash, Internet of things: A survey on enabling technologies, protocols, and applications, IEEE Communications
Surveys Tutorials 17 (4) (2015) 2347–2376. doi:10.1109/
COMST.2015.2444095.
[2] A. Zanella, N. Bui, A. Castellani, L. Vangelista, M. Zorzi, Internet of things for smart cities, IEEE Internet of Things Journal
1 (1) (2014) 22–32. doi:10.1109/JIOT.2014.2306328.
[3] S. Ivanov, K. Bhargava, W. Donnelly, Precision farming: Sensor
analytics, IEEE Intelligent Systems 30 (4) (2015) 76–80. doi:
10.1109/MIS.2015.67.
[4] C. Brewster, I. Roussaki, N. Kalatzis, K. Doolin, K. Ellis,
Iot in agriculture: Designing a europe-wide large-scale pilot,
IEEE Communications Magazine 55 (9) (2017) 26–33. doi:
10.1109/MCOM.2017.1600528.
[5] R. Khan, I. Ali, M. Zakarya, M. Ahmad, M. Imran, M. Shoaib,
Technology-assisted decision support system for efficient water utilization: A real-time testbed for irrigation using wireless
sensor networks, IEEE Access 6 (2018) 25686–25697. doi:
10.1109/ACCESS.2018.2836185.
[6] K. Wang, Y. Wang, Y. Sun, S. Guo, J. Wu, Green industrial
internet of things architecture: An energy-efficient perspective, IEEE Communications Magazine 54 (12) (2016) 48–54.
doi:10.1109/MCOM.2016.1600399CM.
[7] F. TongKe, Smart agriculture based on cloud computing and iot,
Journal of Convergence Information Technology 8 (2).
[8] B. Martinez, M. Montón, I. Vilajosana, J. D. Prades, The power
of models: Modeling power consumption for iot devices, IEEE
734
735
[18]
736
737
738
739
[19]
740
741
742
743
744
[20]
745
746
747
748
[21]
749
750
751
752
[22]
753
754
755
756
[23]
757
758
759
760
761
762
763
11
[24]
Sensors Journal 15 (10) (2015) 5777–5789. doi:10.1109/
JSEN.2015.2445094.
J. Burrell, T. Brooke, R. Beckwith, Vineyard computing: sensor
networks in agricultural production, IEEE Pervasive Computing
3 (1) (2004) 38–45. doi:10.1109/MPRV.2004.1269130.
N. Wang, N. Zhang, M. Wang, Wireless sensors in agriculture
and food industry—recent development and future perspective,
Computers and electronics in agriculture 50 (1) (2006) 1–14.
D. Han, J. Lim, Design and implementation of smart home
energy management systems based on zigbee, IEEE Transactions on Consumer Electronics 56 (3) (2010) 1417–1425. doi:
10.1109/TCE.2010.5606278.
F. Adelantado, X. Vilajosana, P. Tuset-Peiro, B. Martinez,
J. Melia-Segui, T. Watteyne, Understanding the limits of lorawan, IEEE Communications Magazine 55 (9) (2017) 34–40.
doi:10.1109/MCOM.2017.1600613.
M. Ayaz, M. Ammad-Uddin, Z. Sharif, A. Mansour, E. M. Aggoune, Internet-of-things (iot)-based smart agriculture: Toward
making the fields talk, IEEE Access 7 (2019) 129551–129583.
doi:10.1109/ACCESS.2019.2932609.
N. Ahmed, D. De, I. Hussain, Internet of things (iot) for smart
precision agriculture and farming in rural areas, IEEE Internet of
Things Journal 5 (6) (2018) 4890–4899. doi:10.1109/JIOT.
2018.2879579.
W. Chen, Y. Lin, Y. Lin, R. Chen, J. Liao, F. Ng, Y. Chan, Y. Liu,
C. Wang, C. Chiu, T. Yen, Agritalk: Iot for precision soil farming of turmeric cultivation, IEEE Internet of Things Journal 6 (3)
(2019) 5209–5223. doi:10.1109/JIOT.2019.2899128.
S. N. Daskalakis, G. Goussetis, S. D. Assimonis, M. M.
Tentzeris, A. Georgiadis, A uw backscatter-morse-leaf sensor
for low-power agricultural wireless sensor networks, IEEE Sensors Journal 18 (19) (2018) 7889–7898. doi:10.1109/JSEN.
2018.2861431.
P. Spachos, S. Gregori, Integration of wireless sensor networks
and smart uavs for precision viticulture, IEEE Internet Computing 23 (3) (2019) 8–16. doi:10.1109/MIC.2018.2890234.
F. Viani, M. Bertolli, M. Salucci, A. Polo, Low-cost wireless
monitoring and decision support for water saving in agriculture, IEEE Sensors Journal 17 (13) (2017) 4299–4309. doi:
10.1109/JSEN.2017.2705043.
A. A. Aziz, Y. A. Sekercioglu, P. Fitzpatrick, M. Ivanovich, A
survey on distributed topology control techniques for extending the lifetime of battery powered wireless sensor networks,
IEEE Communications Surveys Tutorials 15 (1) (2013) 121–
144. doi:10.1109/SURV.2012.031612.00124.
M. E. E. Alahi, L. Xie, S. Mukhopadhyay, L. Burkitt, A temperature compensated smart nitrate-sensor for agricultural industry,
IEEE Transactions on Industrial Electronics 64 (9) (2017) 7333–
7341. doi:10.1109/TIE.2017.2696508.
F. Viani, M. Bertolli, M. Salucci, A. Polo, Low-cost wireless
monitoring and decision support for water saving in agriculture, IEEE Sensors Journal 17 (13) (2017) 4299–4309. doi:
10.1109/JSEN.2017.2705043.
V. Sharma, U. Mukherji, V. Joseph, S. Gupta, Optimal energy
management policies for energy harvesting sensor nodes, IEEE
Transactions on Wireless Communications 9 (4) (2010) 1326–
1336. doi:10.1109/TWC.2010.04.080749.
C. Alippi, C. Galperti, An adaptive system for optimal solar energy harvesting in wireless sensor network nodes, IEEE Transactions on Circuits and Systems I: Regular Papers 55 (6) (2008)
1742–1750. doi:10.1109/TCSI.2008.922023.
Y. Zhang, S. He, J. Chen, Y. Sun, X. S. Shen, Distributed sampling rate control for rechargeable sensor nodes with limited
battery capacity, IEEE Transactions on Wireless Communications 12 (6) (2013) 3096–3106. doi:10.1109/TCOMM.2013.
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
050613.121698.
[25] Z. Ruirui, C. Liping, G. Jianhua, M. Zhijun, X. Gang, An
energy-efficient wireless sensor network used for farmland soil
moisture monitoring, in: IET International Conference on Wireless Sensor Network 2010 (IET-WSN 2010), 2010, pp. 2–6.
doi:10.1049/cp.2010.1017.
[26] C. Wang, J. Li, Y. Yang, F. Ye, Combining solar energy harvesting with wireless charging for hybrid wireless sensor networks,
IEEE Transactions on Mobile Computing 17 (3) (2018) 560–
576. doi:10.1109/TMC.2017.2732979.
[27] S. Sadowski, P. Spachos, Solar-powered smart agricultural
monitoring system using internet of things devices, in: 2018
IEEE 9th Annual Information Technology, Electronics and Mobile Communication Conference (IEMCON), 2018, pp. 18–23.
doi:10.1109/IEMCON.2018.8614981.
[28] Large 6v 3.7w solar panel.
URL https://www.adafruit.com/product/417
[29] Polymer lithium ion battery - 6600mah 3.7v.
URL
https://www.seeedstudio.com/
Polymer-Lithium-Ion-Battery-6600mAh-3.
7V-p-1711.html
[30] P. Spachos, D. Hatzinakos, Self-powered wireless sensor network for environmental monitoring, in: 2015 IEEE Globecom Workshops (GC Wkshps), 2015, pp. 1–6. doi:10.1109/
GLOCOMW.2015.7414207.
[31] Arduino uno rev3.
URL
https://store.arduino.cc/usa/
arduino-uno-rev3
[32] Grove - moisture sensor.
URL http://wiki.seeedstudio.com/Grove-Moisture_
Sensor/
[33] Sparkfun wifi shield - cc3000.
URL https://www.sparkfun.com/products/retired/
12071
[34] Xbee 2mw wire antenna - series 2.
URL https://www.sparkfun.com/products/retired/
10414
[35] Lora shield for arduino.
URL
http://www.dragino.com/products/module/
item/102-lora-shield.html
[36] M. Chen, G. A. Rincon-Mora, Accurate electrical battery model
capable of predicting runtime and i-v performance, IEEE Transactions on Energy Conversion 21 (2) (2006) 504–511. doi:
10.1109/TEC.2006.874229.
[37] L. Gao, S. Liu, R. A. Dougal, Dynamic lithium-ion battery
model for system simulation, IEEE Transactions on Components and Packaging Technologies 25 (3) (2002) 495–505. doi:
10.1109/TCAPT.2002.803653.
[38] B. S. Bhangu, P. Bentley, D. A. Stone, C. M. Bingham, Nonlinear observers for predicting state-of-charge and state-of-health
of lead-acid batteries for hybrid-electric vehicles, IEEE Transactions on Vehicular Technology 54 (3) (2005) 783–794. doi:
10.1109/TVT.2004.842461.
12
View publication stats