Wireless Sensor Networks for Smart Grid

Applications
Melike Erol-Kantarci, Hussein T. Mouftah
School of Information Technology and Engineering
University of Ottawa, Ottawa, ON, Canada
e-mails:{melike.erolkantarci, mouftah}@uottawa.ca

AbstractElectrical power grid is among the critical infrastructures of a nation. In the past several years, the power
grids have experienced several major failures which have caused
large financial losses in various countries around the globe.
In a close future, the imbalance between the growing demand
and the diminishing fossil fuels, aging equipments, and lack
of communications are anticipated to negatively impact the
operation of the power grids. For this reason, governments
and utilities have recently started working on renovating the
power grid to meet the power quality and power availability
demands of the 21st century. The opportunities that have become
available with the advances in Information and Communications
Technology (ICT) have paved the way to this modernization. The
new grid empowered by ICT is called as the smart grid. The
natural extension of the smart grid applications to the consumer
premises can be through Wireless Sensor Networks (WSNs)
which are able to provide pervasive communications and control
capabilities at low cost. WSNs have broad range of applications
in the smart grid. In this paper we discuss the application of the
WSNs in the home energy management services. We evaluate
the performance of WSNs in terms of delivery ratio, delay and
Packet Delay Variance (PDV) for varying interarrival times and
varying network sizes. We also provide numerical results on the
reduced cost, load and carbon emissions by our home energy
management application.
Index Termssmart grid, wireless sensor networks

I. I NTRODUCTION
The electricity grid is a network with various components
which are mainly responsible for energy generation, power
transmission and electricity distribution processes where consumption forms the last mile of the delivery services. The
techniques used in energy generation, power transmission
and electricity distribution have been designed and implemented almost a century ago, while the growing population
worldwide, and the huge consumer electronics market has
increased nations energy dependence. The demand for energy
is expected to be doubled by 2020 [1] and the diminishing
fossil fuels together with the aging equipments will negatively
impact the resilience of the power grid in a close future.
Already, in the last few years, North America has experienced
several blackouts which have caused major financial losses. To
increase the efficiency of the grid, renovation has become a
necessity. The renovation process targets energy independence,
higher security, high penetration of renewable generation,
efficient buildings and vehicles, and reduced greenhouse gas

Fig. 1: Illustration of the smart grid with communications.

emissions. All of these targeted changes call for efficient communications, coordination and control mechanisms which can
be implemented by the help of the advances in Information and
Communications Technology (ICT). The new grid empowered
by ICT is called as the smart grid.
The objectives of the smart grid are as follows: i) allowing
two-way flow of information and electricity, ii) being selfhealing, iii) being environment-friendly iv) improving energy
storage, v) implementing consumer energy management and
vi) being future proof. An illustration of the smart grid is
given in Figure 1.
Wireless Sensor Networks (WSNs) provide extensive opportunities for the smart grid. Especially, smart grid and pervasive
communication technologies offer diverse energy management
applications [2]. Smart grid enables consumers to have more
control on their consumption. Pervasive communications enable the adoption of the energy management applications
in the daily routines of the consumers while making them
personalized and available anywhere/anytime. In this context,
the natural extension of the advanced technologies to the
consumer premises can be through WSNs which are able to
provide pervasive communications and control capabilities at
low cost.
In this paper, we focus on the home energy management
applications and evaluate the use of WSNs. We first present the

state of the art of the home energy management applications,

while providing an essential background on the objectives
and the challenges of the smart grid. Then, we focus on
an energy management service that employs WSNs. In the
WSN, sensor nodes communicate via IEEE 802.15.4. We
evaluate the performance of the WSN in terms of delivery
ratio, delay and Packet Delay Variance (PDV) for varying
interarrival times and varying network sizes. We show that
the delivery ratio, delay and PDV improves as the packet
interarrival times of the smart home applications increase, i.e.
the performance of the energy management services increase
as the intensity of the other monitoring applications that share
the resources of WSNs decrease. Furthermore, the WSN has
higher delivery ratio, lower delay and lower PDV values for
smaller network sizes which is due to less contention and less
overhead existent in a smaller network. We also showed that
our energy management application is able to significantly
reduce consumer expenses, appliance loads during peak hours
and carbon emissions related with electricity usage during
peak periods.
The rest of the paper is organized as follows. In Section II,
we present the possible WSN applications in the smart grid
and summarize the challenges of the smart grids. In Section III,
we present our energy management application and in Section
IV we present our results. Section V concludes the paper and
addresses open issues.
II. BACKGROUND
WSNs have broad range of applications in the smart grid. In
this section, we will divide the power grid into three conceptual segments, namely generation, transmission & distribution
and consumption, and describe the use of WSNs in each part.
Generation: In the traditional power grid, energy generation facilities are generally monitored with wired sensors
which are limited in amount and located only at a few critical
places. One of the objectives of the smart grid is to increase
the penetration of renewable energy sources in the energy
generation cycle. These renewable energy generation facilities
can be in remote areas, and operate in harsh environments
where continuous monitoring with low cost sensors becomes
necessary. WSNs offer an ideal technology for monitoring and
control of the generation facilities in the smart grid.
Transmission and Distribution: In the transmission and
distribution segment of the power grid, substations, overhead
power lines and underground power lines are the components
that need to be monitored in near real-time. An equipment
failure or breakdown may cause blackouts, or it may even be
dangerous for public health. Moreover, these assets can be
easily reached from outside, therefore they can be a target
of malicious attack. WSNs, once again, provide promising
solutions for monitoring and securing the transmission and
distribution segment.
Demand-Side: For the traditional power grid, substations
have been the last mile of the power grid, and the extension of
the grid into the consumer premises has not been considered.
The smart grid will enable two-way flow of electricity and

Fig. 2: Illustration of the daily load profile for a winter

weekday.

information between supplier and consumer by using the smart

meters and the Advanced Metering Infrastructure (AMI). It
will be possible to communicate with the consumers, monitor
and possibly control their power consumption without disturbing their business or comfort. On the other hand, consumers
will require more advanced home automation tools which
again can be implemented by using advanced sensor technologies. Furthermore, in the smart grid time-differentiated billing
schemes will be employed. For instance, by the end of 2011,
Time of Use (TOU) will be activated by most the utilities in
North America.
TOU is a natural result of consumer activity where consumer demand is known to have seasonal, weekly and daily
patterns. An illustration of a winter weekday load profile
is given in Figure 2. The hours when the grid faces high
load, is called on-peak periods, while moderate and low
load durations are called as mid-peak and off-peak periods,
respectively. Today, utilities handle peak load by bringing
peaker plants into use whenever needed. The maintenance of
these plants are costly, they generally use fossil fuels which are
expensive, and additionally, they have high CO2 emissions [3].
Therefore, with the TOU rates, electricity consumption during
peak periods have higher price than consumption during offpeak periods. An example rate chart of an Ontario-based utility
is given in Table I [4]. Note that, TOU hours and rates may
vary from one utility to another based on the local load pattern
and generation cost.
In fact, in the existing grid, utilities use a method called
demand response to control the load on the grid during peak
periods. Demand response directly aims to control the load of
the commercial and the industrial consumers. Commercial and
industrial consumers have high impact on the load depending
on their scale. Demand response term is used for those consumers decreasing their demand following utility instructions,
either to avoid high rates or failure of the grid. However,
demand response has not been implemented for residential
consumers. On the other hand, the smart grid introduces a
number of opportunities for the home energy management.
Smart grid and WSNs can enable consumers to have more
control on their consumption. We will describe a WSN-based
home energy management system in the following sections.
In literature, most of the residential energy management

TABLE I: TOU rates in Ontario as of 2011.

Winter Weekdays

Summer Weekdays
Weekends

Period
On-Peak
Mid-Peak
On-Peak
Off-Peak
Mid-Peak
On-Peak
Mid-Peak
Off-Peak
Off-peak

schemes follow optimization-based techniques. In [5], the

authors employ a game theoretic approach to reduce the peakto-average electricity usage ratio by finding an optimal consumption schedule. In [6], the authors focus on an automatic
controller that is able to predict the price of electricity during
the scheduling horizon and schedule appliances to provide an
optimum cost and waiting time within that horizon. In [7],
the authors employ particle swarm optimization technique and
schedule the demands of a PHEV, space heater, water heater,
pool pump and generation of the PV system. In [8], the authors
use a neural network-based prediction approach to predict the
next-day demand and according to the predicted demand, the
schedule of the microCHP device in each house is optimized.
Smart homes also employ energy saving and comfort applications that can turn the lights off depending on the occupancy
of the rooms, or dim the lights off based on outside light
intensity and shutter positions, or adjust the thermostat based
on the outside temperature and sensor measurements. These
type of comfort-focused energy management applications date
back to 1990s [9], [10]. However, smart home implementations
have been rare and today most of the residential premises
do not have such energy management systems. Furthermore,
smart home related techniques do not involve communications
with the power grid.
Briefly, there are a large number of opportunities that will
become available with the smart grid however the implementation of the smart grid has several challenges. Regulations and
standardization is one of the major challenges. Currently, various governmental agencies, alliances, committees and groups
are working to provide standards so that smart grid implementations are effective, interoperable and future-proof. Security
is another significant challenge since the grid is becoming
digitized, integrating with the Internet, and generally using
open media for data transfer. Smart grid may be vulnerable to
physical and cyber attacks if security is not handled properly.
Moreover, the utilities need new business models to adapt
to the operation of the smart grid. Consumer participation
especially on the home energy management applications is
a challenge, as well. Successful market penetration of energy
management systems is important for the smart grid to achieve
its goals. Last but not least, the load on the grid is expected to
increase as PHEVs are plugged-in for charging. Unbalanced
and uncoordinated charging may cause failures and the smart
grid calls for novel coordinated PHEV charging mechanisms
[11].

Time
7:00am to 11:00am
11:00am to 5:00pm
5:00pm to 9:00pm
9:00pm to 7:00am
7:00am to 11:00am
11:00am to 5:00pm
5:00pm to 9:00pm
9:00pm to 7:00am
All day

Rate
9.3 cent/kWh
8.0 cent/kWh
9.3 cent/kWh
4.4 cent/kWh
8.0 cent/kWh
9.3 cent/kWh
8.0 cent/kWh
4.4 cent/kWh
4.4 cent/kWh

III. WSN- BASED H OME E NERGY M ANAGEMENT

A PPLICATIONS
In the smart grid, HANs will become the natural extensions
of the grid penetrating into the residential areas, and home
energy management will be a part of the coordination between
generation and load. In [12], we presented an in-Home Energy
Management (iHEM) application that uses WSNs. In this
paper, we present detailed performance evaluation of the WSN
used in iHEM.
The iHEM application employs smart appliances with communication capability, a WSN and a central Energy Management Unit (EMU). In the iHEM application, when a consumer
turns on an appliance, the appliance generates a START-REQ
packet and sends it to the EMU. When the EMU receives
the START-REQ packet, it sends AVAIL-REQ packets to the
energy storage units to retrieve the amount of energy generated
by the renewable resources and stored in their associated
storage units. EMU also communicates with the smart meter
periodically and receives updated price information from the
utility. Upon reception of AVAIL-REQ, the storage unit replies
with a AVAIL-REP packet where the amount of available
energy is sent to the EMU. After receiving the AVAILREP packet, EMU determines the convenient starting time
of the appliance. Basically EMU first tries to accommodate
the appliance requests by the locally generated power. If the
local power is not adequate then it tries to shift the request
to an off-peak hour. After these scheduling attempts, EMU
computes the waiting time as the difference between the
suggested and requested start time, and sends the waiting time
in the START-REP packet to the appliance. The waiting time is
displayed on the LCD monitor of the appliance. The consumer
decides whether to start the appliance right away or wait until
the assigned time slot depending on the waiting time. The
decision of the consumer is sent back to the EMU with a
NOTIFICATION packet. This handshake protocol among the
appliance and the EMU ensures that EMU does not force an
automated start time which avoids degrading the comfort of
the consumers and provides flexibility.
The iHEM application utilizes a WSN to relay its packets.
We assume that this WSN is also utilized for other applications
in the smart home. The WSN continues its regular tasks,
such as inhabitant health monitoring, and at the same time, it
relays iHEM messages. The WSN communicates via Zigbee.
A sample topology of the WSN is given in Figure 3.

where SO is the superframe order and BO is the beacon

order. The range of the SO and BO is defined as 0 SO
BO 14. For beaconless mode these values are set to 15.
In our iHEM application, we configure EMU as the PAN
Coordinator.
IV. S IMULATION R ESULTS

Fig. 3: A typical WSN topology.

Zigbee is a short-range, low-data rate, energy-efficient wireless technology that is based on the IEEE 802.15.4 standard.
Zigbee utilizes 16 channels in the 2.4GHz ISM band worldwide, 13 channels in the 915MHz band in North America and
one channel in the 868MHz band in Europe. The supported
data rates are 250 kbps, 40 kbps, and 20 kbps. Zigbees range
is approximately 30 meters indoors. It supports 16-bit and 64bit addressing modes, and it can support up to 64,000 nodes
(devices). However, when HAN devices need to get integrated
with the Internet, IP addressing is required. IPv6 over LowPower Wireless Personal Area Networks (6LoWPAN), which
is defined in the IETF RFC 4944, aims to integrate IPv6
addressing to LoWPANs like Zigbee. 6LoWPAN adds an
adaptation layer to handle fragmentation, reassembly and
header compression issues, to support IPv6 packets on the
short packet structure of Zigbee.
Zigbee allows two types of devices which are Full Function
Device (FFD) and Reduced Function Device (RFD). FFDs can
communicate with their peers while RFDs are simpler than
FFDs and they can be the edge nodes in a star topology. In
Zigbee, sensor nodes are either organized in a star topology,
mesh topology or a cluster-tree topology. In our model home,
the WSN is organized in a cluster-tree topology. Zigbee
requires a Personal Area Network (PAN) coordinator. PAN
coordinator can operate in beacon-enabled mode or beaconless
mode. The duty cycle of the nodes is defined with the superframe duration (SD) of the superframe structure. A superframe
synchronizes the nodes in the network and nodes. Superframe
has Contention Access Period (CAP) and Contention Free
Period (CFP) slots. During CAP, nodes compete to access the
channel by using the slotted Carrier Sense Multiple Access
with Collision Avoidance (CSMA/CA) technique. During CFP,
nodes that have previously reserved Guaranteed Time Slots
(GTS) transmit their data. One cycle of active and inactive
periods can occur within a Beacon Interval (BI) which starts
at the beginning of a beacon frame and ends at the beginning
of the next beacon frame. SD and BI are calculated as follows
[13]:
SD = aBaseSuperf rameDuration 2SO symbols

(1)

BI = aBaseSuperf rameDuration 2

(2)

BO

symbols

We conduct simulations using Qualnet Network simulator

[14]. In our simulations, we evaluate the performance of the
WSN for varying packet sizes and varying network sizes.
We analyze the delivery ratio, end-to-end delay and PDV.
The delivery ratio is the ratio of the number of successfully
received packets to the number of sent packets. End-to-end
delay is the interval between the time of sending a packet
from the source and the time of receiving the packet at the
destination. PDV is the variation of the delay experienced by
the packets.
We consider a network topology as described in Figure 3
where one PAN coordinator is engaged for residential energy
management and other types of WSN applications. We use
Zigbee protocol utilizing the 2.4GHz ISM band and the bandwidth is 250kb/s. The nodes are a mixture of RFD and FFD
devices, the appliances are assumed to be associated with RFD
nodes. Deploying a dedicated WSN for relaying iHEM packets
is costly therefore, as we mentioned before, the WSN relays
the packets of a smart home monitoring application, as well.
We show the impact of the packet interarrival time of those
applications on the performance of iHEM. The interarrival
time varies between 100s and 300s.
In Figure 4a, we show that the packet delivery ratio of
the WSN increases as the packet interarrival times of the
monitoring application increases. When the interarrival time
is 100s, the delivery ratio is around 85% and it increases to
almost 90% for interarrival times of 300s. As seen in Figure
4b, the end-to-end delay is less than 1s for all interarrival
times. Furthermore, in Figure 4c, we show that PDV is less
than 0.04s and it decreases as the interarrival time increases.
Lower PDV mean that consumers will experience less variable
delay. Therefore the performance of the energy management
services increase as the intensity of the other monitoring
applications that share the resources of WSNs decrease.
We also evaluate the performance of the WSN for varying
number of nodes between 20 and 30. In Figure 5a, we show
that the delivery ratio is higher for small sized networks.
Delivery ratio reaches to 85% when the number nodes is 20.
In Figure 5b, we present that the end-to-end delay decreases
with decreasing number of nodes. The PDV again decreases
for lower number of nodes as seen in Figure 5c. Therefore, the
performance of the WSN improves for smaller network sizes
which is due to less contention and less overhead present in
a smaller network.
We also analyze the energy savings achieved by the iHEM
application. We consider four appliances and assume average
load per cycle as given in [15]. The washer, dryer, dishwasher
and coffer maker is assumed to consume 0.89kWh, 2.46kWh,
1.19kWh and 0.4kWh while the duration of the appliance

0.9

0.05
0.88

0.85

0.045

0.86
0.8
0.75
0.7
0.65

Packet delay variation (s)

Endtoend delay (s)

Delivery ratio

0.84
0.82
0.8
0.78
0.76

0.04

0.035

0.03

0.6
0.74
0.55
0.5
100

0.025

0.72
200
Interarrival time (s)

0.7
100

300

(a) Delivery ratio

200
Interarrival time (s)

0.02
100

300

200
Interarrival time (s)

(b) End-to-end delay (s)

300

(c) PDV (s)

Fig. 4: WSN performance under varying interarrival times.

0.9

1.5

0.85

1.4

0.2
0.18
0.16

0.8
0.75
0.7
0.65

Packet delay variation (s)

Endtoend delay (s)

Delivery ratio

1.3

1.2

1.1

0.14
0.12
0.1
0.08

1
0.6

0.06
0.9

0.55
0.5
30

25
Network size

0.8
30

20

(a) Delivery ratio

0.04

25
Network size

20

0.02
30

25
Network size

(b) End-to-end delay (s)

20

(c) PDV (s)

Fig. 5: WSN performance under varying network sizes.

TABLE II: iHEM results.
Consumer expenses reduced by iHEM
Peak load reduced by iHEM
Carbon emissions reduced by iHEM

30%
40%
20%

cycles are given as 30, 60, 90 and 10 minutes, respectively.

We use the TOU rates of Table I. In Table II, we provide the
savings, peak load reduction and emission reduction of the
iHEM scheme per month. iHEM is able reduce the energy
expenses of the consumers by 30%. Moreover, it can also
reduce the contribution of the appliances on the peak load
by almost 40% and related carbon emissions can be reduced
by 20%.
V. C ONCLUSION
As the existing energy grid is evolving into a smart grid,
there are vast opportunities to apply WSNs to the generation,
transmission, distribution and consumption of the electricity.
Residential energy management is one of the attractive topics
where the implementation of home energy management applications and their integration with the smart grid plays a key
role in increasing the efficiency of the smart grid.
In this paper, we presented a home energy management
application that employs WSNs. We evaluated the performance
of the WSN under varying interarrival times and varying

network sizes. We show that the delivery ratio, delay and

PDV improves as the packet interarrival times of the smart
home applications increase. Furthermore, the WSN has higher
delivery ratio, lower delay and lower PDV values for smaller
network sizes which is due to less contention and less overhead
existent in a smaller network.
Employing WSNs to the smart grid will provide a large
number of opportunities. As a future work we plan to employ
WSNs for PHEV load control applications as well as monitoring renewable energy generation sites.
R EFERENCES
[1] T. F. Garrity, Getting Smart, IEEE Power and Energy Magazine, AprilMarch, 2008, pp 38-45.
[2] M. Erol-Kantarci, H. T. Mouftah, Pervasive Energy Management for the
Smart Grid: Towards a Low Carbon Economy, Pervasive Communications Handbook, Eds. S. I. A. Shah, M. Ilyas, H. T. Mouftah, CRC Press,
2011.
[3] M. Erol-Kantarci, H. T. Mouftah, The Impact of Smart Grid Residential
Energy Management Schemes on the Carbon Footprint of the Household
Electricity Consumption, IEEE Electrical Power and Energy Conference
(EPEC), Halifax, NS, Canada, August 25-27, 2010.
[4] Hydro Ottawa TOU rates, http://www.hydroottawa.com.
[5] A.-H. Mohsenian-Rad, V. W.S. Wong, J. Jatskevich, R. Schober, Optimal
and Autonomous Incentive-based Energy Consumption Scheduling Algorithm for Smart Grid, IEEE PES Innovative Smart Grid Technologies
Conference, Jan 2010
[6] A.-H. Mohsenian-Rad, A. Leon-Garcia, Optimal Residential Load Control with Price Prediction in Real-Time Electricity Pricing Environments,
IEEE Transactions on Smart Grid, vol.1., no.2, 2010, pp. 120-133.

[7] M. A. A. Pedrasa, T. D. Spooner, I. F. MacGill, Coordinated Scheduling

of Residential Distributed Energy Resources to Optimize Smart Home
Energy Services, IEEE Transactions on Smart Grid, vol.1, no.2, 2010,
pp. 134-143.
[8] A. Molderink, V. Bakker, M. Bosman Johann L. Hurink, Gerard J.M.
Smit, Management and control of domestic smart grid technology, IEEE
Transactions on Smart Grid, vol.1., no.2, 2010, pp. 109-119.
[9] B. Brumitt, B. Meyers, J. Krumm, A. Kern, S. Shafer, EasyLiving: technologies for intelligent environments, in: Proc of Handheld Ubiquitous
Computing, 2000.
[10] V. Lesser, M. Atighetchi, B. Benyo, B. Horling, A. Raja, R. Vincent,
T. Wagner, P. Xuan, S. Zhang, The UMASS intelligent home project,
in: Proc. of the 3rd annual conference on Autonomous Agents, Seattle,
Washington, 1999.

[11] M. Erol-Kantarci, H. T. Mouftah, Management of PHEV Batteries in

the Smart Grid: Towards a Cyber-Physical Power Infrastructure, Workshop on Design, Modeling and Evaluation of Cyber Physical Systems (in
IWCMC11), Istanbul, Turkey, June 2011.
[12] M. Erol-Kantarci, H. T. Mouftah, Wireless Sensor Networks for CostEfficient Residential Energy Management in the Smart Grid, to appear
in IEEE Transactions on Smart Grid, 2011.
[13] IEEE 802.15.4 standart. http://www.ieee802.org/15/pub/TG4.html. Last
accessed: February 2011.
[14] Qualnet Network Simulator. http://www.scalable-networks.com/. Last
accessed: January 20011.
[15] R. Stamminger, Synergy Potential of Smart Appliances, Deliverable
2.3 of work package 2 from the Smart-A project, University of Bonn,
March 2009. [Online] http://www.smart-a.org.