Energy Performance of High Data Rate and Low Power Transceiver
based Wireless Body Area Networks
Nanhao Zhu and Ian O’Connor
Université de Lyon, Institut des Nanotechnologies de Lyon (INL) – UMR5270 – CNRS,
Ecole Centrale de Lyon, 36 av. Guy de Collongue, 69134 Ecully, France
Keywords: Energy Consumption, WBANs, Energy Performance, SystemC, Enhanced ShockBurst (ESB), ShockBurst
(SB), nRF24L01+, Energy-efficient Strategies.
Abstract: Emerging Wireless Body Area Networks (WBANs) are receiving increasing interest from researchers and
designers. Specific requirements for small-scale dimensions, low-latency, lightweight and limited power
capacity mean that the key challenge in WBANs design is in the adoption of energy-efficient strategies for
better system performance, and in the efficient use of high data-rate and ultra-low-power transceivers. This
paper presents a high-level energy-aware SystemC-based model and simulation of Nordic’s Enhanced
ShockBurst (ESB) and ShockBurst (SB) baseband protocol engine. The model includes data from Energy
consumption experiments using nRF24L01+ transceiver, enabling detailed exploration of energy
conversation strategies. With this model, we show that a high data-rate ESB and SB transmission at 2Mbps
can save more than 60% and 80% energy respectively, and it has 3x higher lifetime expectancy than the
250Kbps low data-rate communication with a payload collecting strategy.
1 INTRODUCTION
In recent years, a new type of network known as
wireless body area networks (Chen et al., 2011);
(Latré et al., 2011) has attracted increasing interest
from academia and industry alike. Generally, a
WBAN node consists of ultra-low-power, small-
size, lightweight sensors coupled to RF transceiver
for wireless communication as well as
microcontroller for processing. WBANs are usually
used in patient monitoring, sports and fitness
applications. With the wearable and implantable
features, the design of a WBAN sensor node is
strictly constrained to small dimensions and low
weight. Batteries which account for a significant part
of these metrics in the overall system, are
consequently limited in size and weight, which in
turn results in reduced energy capacity. For
implanted devices which must operate over a
lifetime of months or more typically years, the
battery replacement or recharging is inconvenient
and costly. Hence, energy-efficient devices and
sensing/communication strategies are crucial
considerations for the design of a WBAN sensor
node.
The focus on energy consumption optimization is
mainly on the radio transceiver, since wireless
communication is typically the most power
consuming part (Latré et al., 2011); (Weder, 2010).
With high energy consumption, existing Bluetooth-
based device is not suitable for WBANs while
ZigBee-based device cannot satisfy latency
requirements of some WBAN scenarios because of
its low data-rate. High data rate and ultra-low-power
transceiver can be a good solution.
The Nordic nRF24L transceiver series (Nordic
homepage) can support up to 1Mbps/2Mbps high
data rate and they are ultra-low-power compared
with other frequently used devices (Zhu et al., 2011),
and its built-in MultiCeiver function (nRF24L01+
datasheet) supports a simple star network topology,
which is typically preferred by most WBAN
applications.
In order to predict in an accurate way of the
system behavior and network performance of the
designed WBAN in an application context, and to
further explore energy-efficient communication
strategies, it is necessary to build a high-level and
energy-aware model of such a transceiver. We first
present a SystemC-based simulation models for both
Enhanced ShockBurst (ESB) and ShockBurst (SB)
mode (nRF24L01+ datasheet), which are the Nordic
141
Zhu N. and O’Connor I..
Energy Performance of High Data Rate and Low Power Transceiver based Wireless Body Area Networks.
DOI: 10.5220/0004310901410144
In Proceedings of the 2nd International Conference on Sensor Networks (SENSORNETS-2013), pages 141-144
ISBN: 978-989-8565-45-7
Copyright
c
2013 SCITEPRESS (Science and Technology Publications, Lda.)
embedded protocol engines for its 2.4GHz
transceivers. Then we use this SystemC based model
to simulate the energy consumption of various
WBAN scenarios and extract a detailed analysis of
energy performance.
The organization of the rest of the paper is as
follows. In section 2, some related works are
described. In section 3, we first briefly present the
modeling concept and then show some experimental
simulation results with detailed analysis. Finally, we
conclude in section 4.
2 RELATED WORKS
Nordic’s ESB- and SB-based nRF24L transceiver
series are widely used in WBAN application
scenarios by many research communities.
By using nRF24L01, (Sonavane et al., 2009)
propose an adaptive power control algorithm by
simply configuring the programmable output power
on transceiver.
A CSMA/CA based MAC protocol (Zhurong et
al., 2008) is proposed and tested in a nRF24L01-
based pulse oximetry sensor with some experimental
results, however no energy consumption data.
(Weder, 2010) presents a system level simulation
model of nRF24L01 in MiXiM, and simulation
results on energy consumption are given only at
1Mbps data rate and 0dBm power scenario.
In this work, our experimental results provide a
comprehensive energy consumption analysis in both
ESB- and SB-modes.
3 MODELING AND CASE STUDY
To the best of our knowledge, this paper is the first
to propose a SystemC-based simulation model for
Nordic high data-rate and ultra-low-power nRF
transceivers as used in WBANs. We extend Nordic
nRF24L transceiver series module to a SystemC-
based simulator (Wan Du et al., 2011), our
transceiver module supports both ESB and SB-
modes and as the ESB-supported transceiver (e.g.
nRF24L01) model is backward-compatible with SB-
based transceiver (e.g. nRF24E1) models, so
heterogeneous simulations can be handled.
In SystemC simulation, SC_THREAD process
(Black and Donovan, 2009) is used to model ESB-
and SB-modes in the finite state machine. Figure 1
shows a brief state diagram.
Figure 1: Brief State Diagram ESB and SB Mode.
Another feature of the transceiver module is that
the register configuration method is used, so
different nodes objects could maintain individual
settings during simulation.
In this section, an ECG monitoring application is
studied to analyze the energy consumption, and
some optimization strategies are proposed. The
energy performance is focused on the use of the
Nordic hardware embedded ESB- and SB- modes.
The sensor node of the ECG monitoring application
is equipped with an ECG lead sampling at 200 Hz,
and the size of each data is set to 2 bytes. While the
nRF24L01-based platform is widely used in
WBANs for ECG applications (Weder, 2010);
(Zhurong et al., 2008), we base our experiments on
the more recent nRF24L01+ due to its drop-in
device compatibility and improved RF performance.
3.1 Energy Consumption of ESB and
SB Mode
This experiment analyzes energy consumption with
ESB- and SB- modes for varying output power and
data rate values. Based on our ECG application
scenario, a packet containing 2 bytes of payload data
is transmitted every 5ms by PTX (Primary
Transmitter) and the simulation runs for a
transmission of 500 packets. The PRX (Primary
Receiver) is configured in continuous RX mode, and
default values are used for all other packet
parameters (1 byte preamble, 5 bytes address, 1 byte
CRC, power supply of 3.3 V). Figure 2 shows the
simulation results and the energy consumptions are
compared in Table 1.
SENSORNETS2013-2ndInternationalConferenceonSensorNetworks
142
Figure 2: Energy Consumption of ESB and SB Mode.
Table 1: Energy Consumption of ESB and SB.
Data Rate
ESB (PTX)
0dBm VS -18dBm
SB (PTX)
0dBm VS -18dBm
2Mbps 5.35% higher 9.84% higher
1Mbps 8.45% higher 15.65% higher
250Kbps 15.02% higher 28.06% higher
Data Rate
ESB (PTX) SB (PTX)
2Mbps@PTX
(Energy efficient)
65% total
energy is saved
in 2Mbps
83% total
energy is saved
in 2Mbps
250Kbps@PTX
(Energy Consumption)
3.2 Detailed Consumption and Radio
Task Profiling
By using the same application as in part A, Table 2
shows the detailed energy consumption information
and radio task time profiling.
Table 2: Energy Consumption and Task Time.
Detailed Energy Consumption (mJ) @2Mbps/0dBm
ESB-PTX ESB-PRX SB-PTX SB-PRX
Standy1 0.20 0 0.21 0
RX_settling 1.91 1.91 0 1.91
TX_settling 1.72 1.73 1.72 0
TX 0.75 0.60 0.67 0
RX 0.74 105.14 0 108.32
Task Profiling (ms) @2Mbps/0dBm
Standy1 2339.8 0 2423.2 0
RX_settling 65 65 0 65
TX_settling 65 65 65 0
TX 20 16 18 0
RX 16 2360 0 2431.3
For PTX devices in ESB- and SB- modes, large
amount of energy can be saved in low-power
Standby1 mode (less than 4% total energy, over 90%
task time). For PRX devices in ESB- and SB-
modes, the RX mode takes over 90% task time and
also consumes over 95% total energy. The total
energy consumption of PRX in ESB-mode is about
20x that of PTX, while in SB- mode, it rises to about
42x.
3.3 Energy Consumption, Packet
Payload and Lifetime
The first part of Figure 3 shows the relationship
between energy consumption and packet payload.
When compared with (Weder, 2010), our simulation
results provide more information on different
transceiver data rate scenarios (all in ESB-mode):
transmission of packet with 2 bytes payload every
5ms, then 4 bytes every 10ms until 30 bytes every
75ms. The total payload during transmission is the
same (0.4 bytes/ms), and all the simulations are
launched for the same duration. For implanted
devices, WBANs are likely to have higher packet
error rate (PER) due to the impact of cutaneous and
subcutaneous tissue, as well as body movement
(Chen et al., 2011); (Latré et al., 2011). Hence, PER
is necessary for energy profiling. In the second part
of Figure 3, 5% PER and 10% PER are used
respectively in the channel module.
Finally, in Figure 4, we analyze the radio lifetime
expectancy by using a Li-ion battery with a capacity
of 160mAh and coin cell size, which is used by
Generic Wireless Node of Human++ BANs platform
(Huang et al., 2009). The lifetime estimation is
calculated as follows:
( *160) / _ /1000 / 3600 / 24LT SimT Total mAh
(1)
Where LT is lifetime, SimT represents the simulation
time and Total_mAh denotes the total current
consumption during the simulation period.
To sum up, results in Figure 4 show that 2Mbps
scenario is the most energy-efficient way of using
payload collecting strategy for PTX, which can save
from 64.9% (2 bytes) to 70.3% (30 bytes) energy
when compared to 250Kbps, and for 1Mbps scenario
the ratio are from 21.0% to 25.3%. Even with the
introduced PER, high data-rate (2Mbps) scenario
can still provide better energy performance than the
other two data-rate scenarios, when considering 5%
PER, 6.6% (30 bytes) to 12.1% (2 bytes) more
energy is consumed compared with the ideal 2Mbps
scenario, and with 10% PER, the ratio rises to 17.7%
EnergyPerformanceofHighDataRateandLowPowerTransceiverbasedWirelessBodyAreaNetworks
143
to 23.9%.
Figure 3: Energy Consumption, Payload Size and
PER@ESB mode.
Figure 4: Lifetime Estimation of Payload Collecting
Strategy.
On the other hand, lifetime estimation shows that
with low data rate (250kbps), the PTX radio cannot
last for one month, while with the high data rate
(2Mbps), the radio can run for nearly 3 three
months. Therefore high data rate based transmission
is suggested for lower energy consumption, besides
it can also provide much better latency and less
over-the-air collisions.
4 CONCLUSIONS
In this paper, a SystemC structured ESB/SB based
nRF transceiver simulation model is presented, with
high data rate and low power features. Then
nRF24L01+ transceiver is selected for performance
analysis of energy consumption in a WBANs
scenario. Results demonstrate that more than 60%
and 80% energy are consumed respectively in ESB-
and SB- mode in low data rate (250Kbps)
transmission compared to 2Mbps high data rate
transmission, besides high data rate will be more
energy-efficient in payload collecting strategy, and
can last for about 3 months with a coin cell size
battery. These results can provide valuable
information for WBANs application designers.
REFERENCES
M. Chen, S. Gonzalez, A. Vasilakos, H. Cao, V. C. Leung,
Body Area Networks: A Survey, Mobile Networks and
Applications, Volume 16, Issue 2, pp. 171-191, Apr
2011.
Benoît Latré, Bart Braem, Ingrid Moerman, Chris Blondia,
Piet Demeester, A Survey on Wireless Body Area
Networks, Wireless Networks, Volume 17, Issue 1, pp.
1-18, Jan 2011
Andreas Weder, An Energy Model of the Ultra-Low-
Power Transceiver nRF24L01 for Wireless Body
Sensor Networks, Computational Intelligence,
Communication Systems and Networks),2010 Second
International Conference, pp. 118-123.
Nordic 2.4GHz RF homepage, Avaiable:
http://www.nordicsemi.com/eng/Products/2.4GHz-RF.
nRF24L01+ datasheet, Available: http://www.nordicsemi.
com/eng/content/download/2726/34069/file/nRF24L01
P_Product_Specification_1_0.pdf.
S. S. Sonavane, V. Kumar, B. P. Patil, MSP430 and
nRF24L01 based Wireless Sensor Network Design
with Adaptive Power Control, ICGST-CNIR Journal,
Volume 8, Issue 2, pp. 11-15, Jan 2009.
C. Zhurong, H. Chao, L. Jingsheng, and L. Shoubin,
Protocol architecture for wireless body area network
based on nrf24l01, Automation and Logistics,2008,
IEEE International Conference on, pp. 3050–3054.
D. C. Black, J. Donovan, SystemC: From the Ground Up,
2nd ed., Springer 2009. ISBN 0-387-69957-0.
Li Huang, Maryam Ashouei, Firat Yazicioglu, Julien
Penders, Ultra-Low Power Sensor Design for Wireless
Body Area Networks-Challenges, Potential Solutions,
and Applications, International Journal of Digital
Content Technology and its Applications, Volume 3,
Issue 3, pp. 136-148, Sep 2009.
N. Zhu, F. Mieyeville, D. Navarro, W. Du, I.
O’Connor, Research on high data rate wireless sensor
networks, 14eme Journées Nationales du Réseau
Doctoral de Micro et Nanoélectrionique, May 2011.
W. Du, F. Mieyeville, D. Navarro and I. O’Connor,
"IDEA1: A validated SystemC-based system-level
design and simulation environment for wireless sensor
networks," EURASIP Journal on Wireless
Communications and Networking, 2011:143, Oct 2011.
2 6 10 14 18 22 26 30
0
10
20
30
40
50
60
70
80
Pa
y
load ever
y
transmission
[
b
y
te
]
@
0dBm
Radio Lifetime [days]
@2Mbps
@2Mbps-5%PER
@2Mbps-10%PER
@1Mbps
@250kbps
SENSORNETS2013-2ndInternationalConferenceonSensorNetworks
144