A Preliminary Study of Low Power Wireless

Technologies: ZigBee and Bluetooth Low Energy

Jin-Shyan Lee and Ming-Feng Dong Yuan-Heng Sun

Department of Electrical Engineering Information & Communications Research Labs
National Taipei University of Technology (Taipei Tech.) Industrial Technology Research Institute (ITRI)
Taipei, Taiwan Hsinchu, Taiwan
[email protected] [email protected]

Abstract—Low power design of wireless technologies is a of power demand, bit rate, and latency. Nevertheless, BLE was
critical point for future Internet of things (IoT) applications. In found to be more robust to obstacles.
this paper, we provide a preliminary study of low-power wireless
communication standards: ZigBee and BLE, evaluating their On the other hand, since most of the popular wireless
main features and behaviors in terms of various metrics, technologies occupy the same 2.4 GHz ISM frequency bands.
including the transmission time, data coding efficiency, power In [6], Lin et al. compared received signal strength indicator
consumption, and delivery ratio. It is believed that the study (RSSI) of ZigBee and BLE under Wi-Fi and Bluetooth
presented in this paper would benefit application engineers in interferences. The results of the experiments suggest that BLE
choosing an appropriate low-power wireless protocol. outperforms ZigBee in the context of intra-vehicular
communication when Wi-Fi interference exists in the car.
Keywords— Low-power wireless standards, Internet of things Moreover, Shahzad and Oelmann [7] presented a comparison
(IoT), ZigBee, Bluetooth Low Energy (BLE). of ZigBee, BLE, and Wi-Fi about transceivers’ specifications
and power consumption. The results show that the BLE has
I. INTRODUCTION less minimum energy consumption. In [8], Siekkinen et al.
studied the energy consumption of BLE by measuring real
Recently, Internet of things (IoT) is a fast-growing
devices with a power monitor and derived models of the basic
technology to provide the Internet communication with
energy consumption behavior observed from the measurement
physical objects which are not the traditional networked
results. They compared their results by performing similar
devices. Thus, several wireless communication solutions have
measurements on ZigBee/IEEE 802.15.4 devices. Their results
been developed to enable objects transferring small amount of
showed that BLE indeed consumes extremely little energy and
messages to the Internet under low-power requirements.
has a very attractive ratio of energy per bit transmitted.
ZigBee [1] and Bluetooth Low Energy (BLE) [10] are two
famous wireless standards for this kind of low-power Dementyev et al. [9] provided experimental data to
applications. compare power consumption of BLE, ZigBee, and ANT for a
cyclic sleep scenario on a low-power wireless sensor node with
For ZigBee and Bluetooth, Baker [2] studied their strengths
short-range communication. The results determined a sleep
and weaknesses for industrial applications, and claimed that
interval at which the trade-off between power consumption and
ZigBee over IEEE 802.15.4 protocol can meet a wider variety
data rate is optimized. In [10], Gomez et al. presented an
of real industrial needs than Bluetooth due to its long-term
overview and evaluation of BLE. They studied the protocol
battery operation, greater useful range, flexibility in a number
stack and performed experiments for performance evaluation
of dimensions, and reliability of the mesh networking
of energy consumption and latency measurements. Also, they
architecture. Lee et al. [3] provided a study of Bluetooth, UWB,
showed the main characteristics of BLE as compared with
ZigBee, and Wi-Fi wireless communication standards,
other wireless technologies.
evaluating their main features and behaviors in terms of
various metrics, including the transmission time, data coding Higuera et al. [11] studied the feasibility of Bluetooth,
efficiency, complexity, and power consumption. IEEE 802.15.4 and ZigBee in high-speed railway scenarios.
The experiments were conducted on a high-speed rail line to
In [4], Georgakakis et al. presented a comparison of
evaluate the connectivity time and throughput at different
ZigBee, Bluetooth and BLE about applications, topology,
speeds. The result showed that Bluetooth and IEEE 802.15.4
power consumption, data rate, data encryption, authentication
are suitable for ground to train connectivity in high-speed train
and modulation. Also, the advantages and disadvantages of
environments. In [12], Mikhaylov et al. discussed and
these technologies are summarized. Tabish et al. [5] compared
compared the maximum peer-to-peer throughput, the minimum
the main wireless technologies according to power requirement,
frame turnaround time, and the energy consumption for BLE,
throughput, and latency. The result showed that BLE and
IEEE 802.15.4 and SimpliciTI. The results presented that BLE
6LoWPAN have great potentials for such applicability in terms
provides an inexpensive and power-efficient solution for
This paper was supported by the Ministry of Science and Technology
(MOST) of Taiwan (Republic of China, ROC) under grant MOST-103-
2221-E-027-087.

978-1-4799-8389-6/15/$31.00 2015
c IEEE 135

Authorized licensed use limited to: Harokopio University. Downloaded on May 05,2024 at 18:15:25 UTC from IEEE Xplore. Restrictions apply.
wireless communication. However, BLE still have many B. Data Coding Efficiency
limitations that restrict the throughput and communication In this paper, the data coding efficiency is defined by the
latency. ratio of the data size and the message size (i.e. the total number
In this paper, we attempt to make a preliminary comparison of bytes used to transmit the data). The formula for data coding
and experimental study of ZigBee and BLE. The rest of this efficiency (%) can be described as:
paper is organized as follows. Section II introduces the
comparative study of ZigBee and BLE, including their PcodEff = Ndata /( Ndata + ( Ndata / NmaxPld × Novhd )) (2)
transmission time and data coding efficiency. Next,
The parameters listed in Table I are also used for the coding
experimental evaluations for ZigBee and BLE, including their
efficiency comparison. Fig. 2 shows the data coding efficiency
power consumption and delivery ratio, are described in Section
III and IV, respectively. Finally, Section V concludes this of ZigBee and BLE versus the data size. For small data sizes
paper. (around smaller than 37 bytes), BLE is much better than the
ZigBee. For large data sizes, BLE has only a little better
efficiency of 78.68%, as compared with the 76.52% of ZigBee
II. COMPARATIVE STUDY
(where the data is 10K bytes as listed in Table I). The
A. Transmission Time discontinuities in Fig. 1 and 2 are caused by data fragmentation,
i.e. the maximum data payload, which is 37 and 102 bytes for
The transmission time depends on the data rate, the BLE and ZigBee, respectively.
message size, and the distance between two nodes. The
formula for transmission time (ȝs) can be described as: 80

Ttx = ( Ndata + ( Ndata / NmaxPld × Novhd )) × Tbit + Tprop (1) 70

Data Coding Efficiency (%)

60
where Ndata is the data size, NmaxPld is the maximum payload
size, Novhd is the overhead size, Tbit is the bit time, and Tprop is 50
the propagation time between any two devices. For simplicity,
40
the propagation time is negligible in this paper. The typical
parameters of ZigBee and BLE for transmission time 30
evaluation are listed in Table I. As shown in Fig. 1, the BLE
20
transmission time for the ZigBee is longer than BLE because ZigBee
of the lower data rate (250 Kbit/s). Obviously, the result also 10
shows the required transmission time is proportional to the data
payload size and disproportional to the maximum data rate. 0
0 1 2 3 4
10 10 10 10 10
6 Data Payload Size (bytes)
10
Fig. 2. Comparison of the data coding efficiency versus the data size.
Transmission Time (micro seconds)

5
10
III. EXPERIMENTAL STUDY FOR ZIGBEE
10
4 In this section, several sets of experiments are conducted to
study the ZigBee wireless networks.
3
10 A. Power Consumption
As shown in Fig. 3, two ZigBee development boards are
2 BLE
10 used as a coordinator (left) and an end device (right),
ZigBee
respectively. The boards are based on the Texas Instruments
1
(TI) CC2530 chips [13]. For measuring the current
10 consumption of the end device, a 10-ohm resistance is used
0 1 2 3 4
10 10 10 10 10 [14]. The end device continuously transfers the data request to
Data Payload Size (bytes)
the coordinator so as to perform the measurements. The study
Fig. 1. Comparison of the transmission time versus the data size. is for a steady state network, i.e. after the end device finish
channel scanning and the association procedure to join the
TABLE I coordinator.
TYPICAL PROTOCOL PARAMETERS
Fig. 4 shows the power consumption as the end device
Standard ZigBee BLE performing the polling sequence when the coordinator has no
Max data rate (Mbit/s) 0.25 1 buffered data. The total power consumption per polling cycle is
Max data payload (bytes) 102 37 roughly estimated as 103.9 (mA*ms). On the other hand, the
Max overhead (bytes) 31 10 power consumption during a polling sequence where the
Coding efficiency (%) 76.52 78.68
coordinator has data for the end device is measured as Fig. 5.

136 2015 IEEE 10th Conference on Industrial Electronics and Applications (ICIEA)

Authorized licensed use limited to: Harokopio University. Downloaded on May 05,2024 at 18:15:25 UTC from IEEE Xplore. Restrictions apply.
The total power consumption is roughly estimated as 610.2 33mA 33mA 33mA
(mA*ms) per cycle. 27mA
27mA
27mA 27mA

23mA 27mA
B. Delivery Ratio 23mA
23mA 8mA
The experiment is run in a one-hop star topology, as shown 8mA

in Fig. 6. The distance between the coordinator and the end 14mA

device is 1 meter. The end device is the main transceiver 8mA

continuously sending or receiving data packet to or from the
coordinator. The other two devices are traffic load generators.
Fig. 7 shows the delivery ratio between the coordinator and the
end device for different data payload sizes with varied traffic Fig. 5. Power consumption when the coodinator has data for end device.
loads. The results show that with the increase of payload size,
the delivery ratio decreased due to the presence of collisions
7UDIILFJHQHUDWRU
and random backoffs. Moreover, with the larger traffic load
from other devices, the delivery ratio is also reduced because
FP
the collision possibility increased.
P

(QGGHYLFH
Coordinator &RRUGLQDWRU FP
10ȍ
7UDIILFJHQHUDWRU
End device
Fig. 6. Experimental configuration for ZigBee develiry ratio.

120

100
Delivery ratio (%)

Fig. 3. Experimental configuration for ZigBee power consumption. 80

60
no traffic load
40
1-node traffic load
27mA 33mA
20 2-node traffic load
27mA
27mA
0
1 20 60 100
Data payload (bytes)
23mA
Fig. 7. Delivery ratio of data payload sizes with varied traffic loads.
7 mA 8mA 8mA
5mA
14mA
IV. EXPERIMENTAL STUDY FOR BLE
In this section, several sets of experiments are conducted to
study the BLE wireless networks.

A. Power Consumption
Fig. 4. Power consumption without buffered data on coodinator. As shown in Fig. 8, one BLE USB dongle and one BLE
development board are used as a master (left) and a slave (right)
device, respectively. The master dongle and slave board is
based on the TI CC2540 [15] and CC2541 [16] chips,
respectively. For measuring the current consumption of the
slave device, a 10- ohm resistance is also used [17].
Fig. 9 (a) shows the power consumption as the slave device
performs broadcasting sequence to find the master device. The

2015 IEEE 10th Conference on Industrial Electronics and Applications (ICIEA) 137

Authorized licensed use limited to: Harokopio University. Downloaded on May 05,2024 at 18:15:25 UTC from IEEE Xplore. Restrictions apply.
power consumption per broadcasting cycle is roughly
estimated as 57.1 (mA*ms). Then, the slave device would
transmit packets to confirm this connection with master device,
as shown in Fig. 9 (b). The power consumption per
confirmation cycle is roughly estimated as 39.5 (mA*ms). On
12mA
the other hand, the power consumption of the slave device to
transmit one-byte data is measured as Fig. 10. The total power 10mA
consumption is roughly estimated as 29 (mA*ms) per cycle.

B. Delivery Ratio
The experiment is run in a one-hop star topology, as shown
in Fig. 11. The distance between the BLE master and slave
device is 1 meter. The slave device is the main transceiver Fig. 10. Power consumption when the coodinator has data for end device.
continuously sending or receiving data packet to or from the
master. Fig. 12 shows the delivery ratio between the master
and the slave device for different transmission time intervals
with varied data payload sizes. The results show that with the P
decrease of time intervals, the delivery ratio reduced because of
the presence of traffic jams. Moreover, with the larger payload
size, the delivery ratio is also decreased since the collision
possibility increased. BLE USB dongle BLE board (slave)
(master)

Fig. 11. Experimental configuration for develiry ratio.

10ɏ VCC(3V)

120

BLE USB dongle 100

(master) BLE board

Delivery ratio (%)

80
(slave)
60
1 byte
Fig. 8. Experimental configuration for BLE power consumption.
40 10 bytes
20 bytes
20

0
10 20 30 40 50 100

17.5mA 17.5mA Transmission Time Interval (ms)

13mA
10mA 11mA 13mA Fig. 12. Delivery ratio of transmission time intervals with varied payload sizes.

Sleep mode
V. CONCLUSIONS
Now or in the near future, IoT applications are seems to
(a) (b) deploy massive devices with low-power and short-range
Fig. 9. Power consumption for a) boradcasting and b) confirmation. wireless communication technologies. This paper has
presented a quantitative evaluation of ZigBee and BLE in
terms of the transmission time, data coding efficiency, power
consumption, and delivery ratio. However, this paper is not to
come to any conclusion about which one is better since the
suitability of wireless selection is greatly affected by practical
situations, such as the realistic environment interferences,
chipset prices, and installation cost. Future work includes a
more comprehensive comparison of ZigBee and BLE for
specific realistic case study.

138 2015 IEEE 10th Conference on Industrial Electronics and Applications (ICIEA)

Authorized licensed use limited to: Harokopio University. Downloaded on May 05,2024 at 18:15:25 UTC from IEEE Xplore. Restrictions apply.
 REFERENCES [8] M. Siekkinen, M. Hiienkari, J. K. Nurminen, and J. Nieminen, “How
low energy is Bluetooth Low Energy? Comparative measurements with
[1] J. S. Lee and Y. M. Wang, “Experimental evaluation of ZigBee-based ZigBee/802.15.4,” in Proc. IEEE Wireless Communications and
wireless networks in indoor environments,” Journal of Engineering, vol. Networking Conf. (WCNC), Paris, France, Apr. 2012, pp. 232-237.
2013, Article ID 286367, 9 pages, doi10.11552013286367, Feb. 2013. [9] A. Dementyev, H. Steve, T. Stuart, and S. Joshua, “Power consumption
[2] N. Baker, “ZigBee and Bluetooth: Strengths and weaknesses for analysis of Bluetooth Low Energy, ZigBee and ANT sensor nodes in a
industrial applications,” IEE Computing & Control Engineering, vol. 16, cyclic sleep scenario,” in Proc. Int. Wireless Symp., Beijing, china, Apr.
no. 2, pp 20-25, April/May 2005. 2013, pp. 1-4.
[3] J. S. Lee, Y. W. Su, and C. C. Shen, “A comparative study of wireless [10] C. Gomez, J. Oller, and J. Paradells, “Overview and evaluation of
protocols: Bluetooth, UWB, ZigBee, and Wi-Fi,” in Proc. Annual Conf. Bluetooth Low Energy: An emerging low-power wireless technology,”
IEEE Industrial Electronics Society (IECON), Taipei, Taiwan, Nov. Sensors, vol. 12, no. 9, pp. 11734-11753, 2012.
2007, pp. 46-51. [11] J. Higuera, E. Kartsakli, J. L. Valenzuela, L. Alonso, A. Laya, R.
[4] E. Georgakakis, S. A. Nikolidakis, D. D. Vergados, and C. Douligeris, Martinez, and A. Aguilar, “Experimental study of Bluetooth, ZigBee
“An analysis of Bluetooth, ZigBee and Bluetooth Low Energy and their and IEEE 802.15.4 technologies on board high-speed trains.” in Proc.
use in WBANs,” in Wireless Mobile Communication and Healthcare. IEEE Conf. Vehicular Technology (VTC Spring), Yokohama, Japan,
Springer, 2011, pp. 168-175. May 2012, pp. 1-5.
[5] R. Tabish, A. B. Mnaouer, F. Touati, and A. M. Ghaleb, “A comparative [12] K. Mikhaylov, N. Plevritakis, and J. Tervonen, “Performance analysis
analysis of BLE and 6LoWPAN for U-HealthCare” in Proc. IEEE Gulf and comparison of Bluetooth Low Energy with IEEE 802.15.4 and
Cooperation Council (GCC) Conference and Exhibition, Doha, Qatar, SimpliciTI,” J. Sens. Actuator Networks, vol. 2, no. 3, pp. 589-613, Aug.
Nov. 2013, pp. 286-291. 2013.
[6] J. R. Lin, T. Talty, and O. K. Tonguz, “An empirical performance study [13] Texas Instruments, CC2530 Datasheet. Nov. 2010.
of intra-vehicular wireless sensor networks under WiFi and Bluetooth [14] B. Selvig, Measuring Power Consumption with CC2430 & Z-Stack,
interference,” in Proc. IEEE Global Communications Conf. Application Note AN053, Texas Instruments. Jul. 2007.
(GLOBECOM), Atlanta, USA, Dec. 2013, pp 581-586. [15] Texas Instruments, CC2540 Datasheet. Jun. 2013.
[7] K. Shahzad and B. Oelmann, “A comparative study of in-sensor [16] Texas Instruments, CC2541 Datasheet. Jun. 2013.
processing vs. raw data transmission using ZigBee, BLE and Wi-Fi for [17] S. Kamath and J. Lindh, Measuring Bluetooth Low Energy Power
data intensive monitoring applications” in Proc. Int. Symp. Wireless Consumption, Application Note AN092, Texas Instruments. Apr. 2012.
Communication Systems, Barcelona, Spain, Aug. 2014, pp 519-524.

2015 IEEE 10th Conference on Industrial Electronics and Applications (ICIEA) 139

Authorized licensed use limited to: Harokopio University. Downloaded on May 05,2024 at 18:15:25 UTC from IEEE Xplore. Restrictions apply.