ICC2017: WT04-5thIEEE International Workshop on Smart Communication Protocols and Algorithms (SCPA 2017)

Efficient Low Cost Supervisory System for Internet

of Things Enabled Smart Home
Md.Sarwar Kamal∗ , Sazia Parvin , Kashif Saleem¶ , Hussam Al-Hamadi§ , Amjad Gawanmeh§
∗ Depart. of Computer Science and Engineering, East West University, Dhaka, Bangladesh, [email protected]
 University of New South Wales, Canberra, Australia, [email protected]
¶ Center of Excellence in Information Assurance, King Saud University, KSA, [email protected]
§ Department of Electrical and Computer Engineering, Khalifa University, Abu Dhabi, UAE

{hussam.alhamadi,amjad.gawanmeh}@kustar.ac.ae

Abstract—Internet of Things (IoT) is a hot and debated subject Smart homes are augmenting more interest among all people
in current digitalized era. IoT enables internet connectivity for irrespective of rich and poor throughout the world [11], [12].
all kinds of devices and physical objects. The current world is The research community is putting lot of efforts in enhancing
approaching towards virtualization of different types of systems
that enables performing activities without direct physical inter- the IoT from every aspect to make it more and more efficient
action. The combination of high-speed internet and intelligent and reliable [9], [13]–[17]. In this scenario, an intelligent al-
devices makes it easier to manage multiple jobs smoothly without gorithm might help to handle IoT based devices to accomplish
the limitation of distances. The outstanding advantages of these the activities of daily routine in an optimal manner [8], [11].
promising technologies require the deployment and utilization of In this work, an algorithm approach has been proposed
proper methods to handle the difficulties arise with these new
applications. in the real world. This paper propose an efficient for home automation and metering system to increase the
low cost supervisory system for smart home automation that can efficiency and reliability of power consumption control and
be managed using IoT. The proposed system is based on Apriori supervisory. Initially all data is collected from the sensors
algorithm and will help to monitor and control all the home and organized with the help of Apriori algorithm [18], [19]
appliances and electronic devices through a supervisory system that is used to perform dynamic analysis. Based on the
in a most efficient and reliable manner. Both the consumers
and the suppliers will get the opportunity to manage the power algorithm criteria, the common rules are generated to control
distribution by monitoring the electricity consumption. the connected devices. The pivotal benefit of these rules is
to train the system for upcoming events. A web portal is
I. I NTRODUCTION designed and developed in a manner to control IoT embedded
Internet of things (IoT) develops a new era of technological appliance and other electronic devices at home from any part
world with a new innovation of intelligence computing [1], of the world in a reliable manner. Consequently, the electricity
[2]. IoT can be termed as the connection among various kinds consumption can be determined and the automated billing can
of digital and analog devices like smart phones, personal com- also be generated with the help of the web portal. At this stage,
puter, personal digital assistance, and tablets to the internet, frequent used devices in a home can be categorized according
that create in very new approach of communication among to the number of hours the device is being used.
things and people and consequently among also things [3]. The Section II reviews the related state of the art literature
The pivotal aim of IoT is to monitor and connect the physical available yet. Section III presents and explains the proposed
objects/devices around us in a more smooth and smart way system architecture, Section IV illustrates Apriori algorithmic
[4]. IoT offers to modify the shape of living by ensuring cost followed by the operation of the system that is discussed in
effective living including safety, security, and entertainment. Section V. Finally, the paper is concluded in Section VI.
Rapid development of technology makes the human daily
life more comfortable with the help of recently designed smart II. R ELATED W ORK
systems and devices [5], [6]. Due to the rapid improvements In [20], the authors have developed a system using Rep-
in internet and the development of smart embedded systems, resentational State Transfer (REST) based Web services that
people are getting more customized in using the internet to does not require a dedicated server PC for monitoring and
avail numerous services around the globe [7], [8]. Daily life controlling environment. The authors implemented server ap-
utilities are mostly enabled with internet protocol (IP) [9], and plication within Arduino with the ability to configure and
in the near future, many of these will smart devices that are enable the communication and to extract the messages in
communicating over IoT [10]. The pivotal aim of IoT is to JavaScript Object Notation (JSON) format. The authors have
monitor and connect the physical objects/devices around us in utilized JAVA programming language based the Android Soft-
a smart manner [4]. IoT offers to modify the shape of living ware Development Kit (SDK) to develop the Smartphone
by ensuring cost effective living including safety, security, and application that provides the functionalities to the user as,
entertainment. This technology can be applied in multiple daily remote connection, then the control on devices, monitoring the
life disciplines in order to enhance the life [8]. devices, and schedule management. The application have the

978-1-5090-1525-2/17/$31.00 ©2017 IEEE

 ICC2017: WT04-5thIEEE International Workshop on Smart Communication Protocols and Algorithms (SCPA 2017)

option for user under settings to enter the network details to on ZigBee network that connects and communicate with IoT
connect with the micro Web-server in their application. The server via gateway that translates packets from ZigBee to IP
application is synchronized with the Web-server in order to gateway.
retrieve the data from actuators and sensors. In this work, no One of the recent study [24] has designed a smart home
comparison of results to justify the system functionality and approach with moderate security and low development cost
performance are presented by the authors. under IoT. However, this process only considers sequential
The authors in [2] proposed Internet of Things based low approach that cannot handle congestion generated by large
cost ubiquitous sensing system to monitor regular domestic number of devices. Power utilization and conservation in
conditions. The system provides basic operations as remote homes in considered in another research [3] IoT device are
management and control of domestic devices to the user. The merged in a common line. In this development, they applied
system framework and the functionality depends on the com- image processing system to identify human functionalities.
bination of information system for data aggregation, pervasive But image processing consumes longer time to select desired
distributed sensing units, and reasoning and context awareness. devices. In the same analysis, another research is designed
The authors utilized XBee-S2 modules built by Digi equipped based on WiFi and GSM technology. The problem of this
with sensors and is connected with a coordinator using wire- work is to collect data from sensors and control according to
less communication. The coordinator is then attached to the the desired devices [25]. Though there are quite a few types
router runs an open source embedded Linux (OpenWRT). In of research are carried out elsewhere, still, there is the scope
this system, still internet protocol (IP) v6 is working until the for more improvement of home automation system using IoT.
router that works as gateway for packet transformation from
ZigBee network based devices to IPv6 format, which means III. S UPERVISORY S YSTEM FOR S MART H OME
that the gateway is connected to internet not the devices. The supervisory system is shown in Figure 1. The super-
In another work, the authors in [21] considered IoT-based visory system enables the user to select the way he would
home automation using low-cost android phones. The authors like to operate the smart home. It will enable the user to
propose two types of home automation using Bluetooth and control and monitor several home appliances such as light,
Ethernet. Nonetheless, data sources in this designing is not fan, TV, air cooler, washing machine etc. can be controlled.
clear. As a consequence, a deadlock situation arises when To accomplish the operations, a supervisory system has been
multiple phones try to access common web portal. For experi- designed. In database part of the web site, Apriori algorithm
mentation, the authors utilized micro-controller based Arduino is implemented under database quarry language. Apriori al-
board and a customized Android based mobile phone with gorithm will be used in order to validate the most frequent
customized application. data recorded on the database. Then the user can access the
The flexible standalone, low cost smart home system is supervisory system through a dedicated webpage, such as
presented in [22]. This system is based on Arduino platform (http://ha.dcoders.net/), where he/she will send a command
that communicates over the internet with a customized app through a Wi-Fi router/Internet. Then, the command will
installed on Android mobile. The Arduino device is enabled receive by the microcontroller via Wi-Fi module. Afterwards,
with sensors and functionalities such as intrusion detection according to the command send by the user the electronic
sensors, humidity sensors, temperature sensors, smoke/gas devices can be controlled. From Fig. 1, it can also define the
sensors, power plugs, and light switches to be monitored and system consists of two-way communication. The first is to
controlled by the mobile app. The heart of the system is control the devices using this dedicated supervisory system,
the micro web-server programmed and run on Arduino Mega as previously depicted.
2560. The authors have tested and have analyzed the efficiency The supervisory system enables the user to obtain in-
of the application and its connectivity over the internet. The formation about the power consumption by the devices, in
system generates alerts based on the given thresholds only in addition, the meter reading can be obtained directly. That
the form of emails. means the devices will send the power consumption rating to
In [23], the authors emphasis more on customer-centric the supervisory system with the help of a dedicated hardware
technique to make the system more scalable and based on this that can be implanted using a simple microcontroller.
technique have presented an architecture for the IoT based
IV. P ROPOSED A LGORITHM
last-meter smart grid. Their architecture provides four main
advantages first of which is seamless integration of smart A. Apriori Classifier
grid with smart home applications. Secondly, the collection Apriori is a classic algorithm for learning and generating
and analyzation of the sensed data through communication association rules from large data set. Association rules are
protocols. Furthermore, the system can be only be accessed defined as IF-THEN statements that help to discover the
by the authorized persons. Fourth one is to simplify interac- relationships between unrelated data in relational database and
tion with nontechnical users, the system allows mapping of other nature of data and properties. The Apriori algorithm
sensor and actuator to the layer of abstraction. This system is is considered as a moderately depth preference selection
deployed and is tested to provides monitoring and managing approach that handles frequent item sets, which are then
power consumption in real time. The experimentation depends combined with Association Rues (AR). The association rules
 ICC2017: WT04-5thIEEE International Workshop on Smart Communication Protocols and Algorithms (SCPA 2017)

Smart Home
Wi-Fi IoT Enabled
(Internet)

Arduino
ino
Supervisory System Controller
l
Processed data
From algorithm

Fig. 1. Overview of Supervisory System Architecture for Smart Home.

correlate the different types of data items with preferable Algorithm 1 Apriori Algorithm for Minimum Support
properties or attributes [1]. procedure Apriori_Algorithm(Ek , Fk )
Input: Ek : Candidate item set of size k //set of electronic device.
For example, let B = {B1 , B2 , . . . Bn } represents a set of Output: Fk : frequent item set of size k //electronic device in candidate item set.
binary attributes called item set, and T = {T1 , T2 , . . . , Tn }is Initialize:
• F1 = {f requentitems}; // initialize set of frequencies items
a set of transactions called the database. Each transaction in •m=1
T has a unique transaction ID and contains a sub-set of items REPEAT
• Em = candidategeneratesf romF m //generates a new candidate
from B. Thus, a rule is defined as implication of the form REPEAT
X → Y where, X, Y ⊆ B and X ∩ Y = ∅ . The set of • ∀t ∈ T // transaction in T, database subgroups of electronic device
• do begin
items X and Y are called antecedent and consequent of the Em+1 = subset(Em , t) // this function generate candidates in transaction
rules respectively. An association rule can simply defined by ∀e · e ∈ Em ∧ e ∈ Em+1 // for all candidates
• do begin
simple notation: e.count + + //determine support
• end
X → Y [S, C] (1) Fm+1 = e|e ∈ Em ∧ e ∈ Em+1 ∧ e.count ≥ minSup // new set
• end
Where X and Y are two different item sets, S is support m=m+1
UNTIL Fm = ∅,
and C is confidence, here the support count refers to the UNTIL m > k, 
frequencies of the rules within the data classification. Support OUTPUT Ek+1 Fk+1
END
count is performed on different data level such as high and low
end procedure
value. High value means that rule involves a highest impact
on database and low value indicates less importance in the
association rules generation. considered depending on the usage, for instance, light and
support(X → Y [S, C]) = P (P ∩ Q) (2) room heater account for 10 hours, and an average less than 1
hour for fan and air condition. Algorithm 1 shows the proposed
Meanwhile, the confidence C defines the percentage of
Apriori algorithm for processing the database.
transactions in the rule generation. We define confidence in
percentage using the following expression: TABLE I
 
X support(X, Y ) M ASTER DATA FILE FOR CLASSIFICATION WITH DIFFERENT ELECTRONIC
conf idence(X → Y ) = P = (3)
Y support(Y ) DEVICES IN WINTER SEASON .

User ID L>=10hr H>=10hr F<=1 hr AC<=1hr

B. Apriori Algorithm for controlling electronic device U1 1 0 0 0
In our daily life we use different house hold electronic U2 0 1 0 0
devices. These electronic devices are used based on a fixed U3 0 1 0 1
U4 1 0 0 0
time (Day or night) or season (winter or summer).We consider ... 0 1 0 0
a database D which contains 1000 electronic device users. U500 0 0 1 0
In our experiment we consider five electronic devices such U501 1 0 0 0
... 0 1 0 0
Light (L), Fan (F), Room Heater (H) and Air Cooler (AC). U999 1 0 0 0
We convert the using hours of these electric devices into U1000 0 0 1 0
binary flags (1/0) on different cases such as morning, night,
winter and summer. Database DD records the power usage In winter season maximum user less use Fans (F) and Air
of different electronic devices for 1000 users and their usage Cooler (AC) that is less than 1 hours. Maximum people use
per hour is scanned. In Winter season, minimum hours are room heater and light. We generate a binary table (Table I)by
 ICC2017: WT04-5thIEEE International Workshop on Smart Communication Protocols and Algorithms (SCPA 2017)

Minimum Minimum
Attribute
suport
Attribute
suport heater and light. Here, we generate a binary table (Table I)
{L>=10hr} 826 {L>=10hr} 826 by 1 or 0. The cell value 1 indicates that users satisfy the
{H}>=10hr} 789 {H}>=10hr} 789 field condition, while the value 0 indicates that users use the
{F<=1hr} 645 {F<=1hr} 645 electronic device for a during less than the considered number
{AC}<=1hr 605 {AC}<=1hr 605 of hours. Then we apply Apriori approach on database DD in
E1 Compare candidate support F1
order to generate a set of association rules. We consider our
count with min support minimum support count is 200. Scan the database, DD for
count of each candidate based on minimum support count, as
Fig. 2. First Scan Process of Database DD that generates Candidate Key E1 illustrated in Figure 2. Each electronic device is considered as
and Support Table F1 a member of the set of candidates, from which two electronic
devices with frequent patterns are selected, as illustrated in
1 or 0. The cell value 1 indicates that users satisfies the Figures 3 and 4.
field condition, while the value 0 indicates that users use the
Attribute
electronic device for a during less than the considered number
{L.>=10hr,H>=10hr,AC<=1hr}
of hours. Then we apply Apriori approach on database DD in
{L.>=10hr,H>=10hr,F<=1hr}
order to generate a set of association rules. We consider our
Generate
minimum support count is 200. Scan the database, DD for {L.>=10hr,F<=1hr,AC<=1hr}

count of each candidate based on minimum support count, as {H.>=10hr,F<=1hr,AC<=1hr}

illustrated in Figure 2. Each electronic device is considered a E3 Scan DD for min support
for two items
member of the set of candidates, from which two electronic
devices with frequent patterns are selected, as illustrated in Attribute Minimum support

Figures 3 and 4. {L.>=10hr,H>=10hr,AC<=1hr} 213

{L.>=10hr,H>=10hr,F<=1hr} 245

Attribute {L.>=10hr,F<=1hr,AC<=1hr} 120

{L.>=10hr,H>=10hr} {H.>=10hr,F<=1hr,AC<=1hr} 301

{L>=10hr,F<=1hr}
E3 Compare candidate support
{H>=10hr,AC<=1hr} count with min support
Generate
{H>=10hr,F<=1hr}
Attribute Minimum support
{F<=1hr,AC<=1hr}
{L.>=10hr,H>=10hr,AC<=1hr} 213
{H>=10hr,L<=10hr}
{L.>=10hr,H>=10hr,F<=1hr} 245
E2 Scan DD for min {H.>=10hr,F<=1hr,AC<=1hr} 301
support for two items
F3
Minimum
Attribute
suport
{L.>=10hr,H>=10hr} 434 Fig. 4. Third Scan Process of Database DD that generates Candidate Key
{L>=10hr,F<=1hr} 402 E3 and Support Table F3
{H>=10hr,AC<=1hr} 395

{H>=10hr,F<=1hr} 340
C. Association Rule
{F<=1hr,AC<=1hr} 298 Next, we generate association rules for frequent item sets
{H>=10hr,L<=10hr} 189 from F3 . For every item sets of F3 , we generate all nonempty
E2
Compare candidate support
subsets of frequent item sets by considering a confidence limit
count with min support is 50%. The rule G = {L >= 10hr, H >= 10hr, AC <=
Attribute
Minimum 1hr} has all nonempty subsets of {L >= 10hr}, {H >=
suport
{L.>=10hr,H>=10hr} 434
10hr}, {AC <= 1hr}, {L >= 10hr, H >= 10hr}, {L >=
10hr, AC <= 1hr}, {H >= 10hr, AC <= 1hr}. For a min-
{L>=10hr,F<=1hr} 402
imum confidence threshold of 50%, the resulting association
{H>=10hr,AC<=1hr} 395
rules for winter season are shown in Table II.
{H>=10hr,F<=1hr} 340

{F<=1hr,AC<=1hr} 298 D. Data Propagations

F2
The processed data sets from Apriori algorithm send to the
desired hardware, which is implemented in targeted homes.
Fig. 3. Second Scan Process of Database DD that generates Candidate Key
E2 and Support Table F 2
This environment is built under a microcontroller (Arduino Pro
Mini), Wi-Fi module (ESP8266 Wi-Fi chip), some relays and
If we consider the example in Winter season, maximum an LCD (liquid crystal display). The LCD notifies common
users use less fans (F) and air cooler (AC), probably for parameters such as date, time, temperature and the device
less than 1 hour. On the contrary, maximum users use room condition (ON/OFF). All the data is collected directly from
 ICC2017: WT04-5thIEEE International Workshop on Smart Communication Protocols and Algorithms (SCPA 2017)

TABLE II
A SSOCIATION RULES C ONFIDENCE .

Rules Confidence Result Status

R1 : (L >= 10hr)( H >= 10hr) → (AC <= 1hr) 213/434=49.07% Rejected
R3 : (H >= 10hr( AC <= 1hr) → (L >= 10hr) 213/395=53.92% Selected
R2 : (L >= 10hr) → (H >= 10hr ( AC <= 1hr) 213/826=25.78% Rejected
... ... ...

the designed system with standard format. Accurate time is on a database with a set of options to turn on and off the
ensured by analyzing simultaneous simulation of collected devices used in home automation. According to the database,
data and Apriori algorithm. The quality of microcontroller and if any user would like to turn on or off the device, he/she just
Wi-Fi chips is also a critical issue for efficient communication. need to click the appropriate option. A scaled based indicator
Various chips from different companies are analyzed, and with a maximum value of 100 is used to control different types
find out that system efficiency specifically depends on the of devices, for instance, to control the speed of the fan. In
quality of chips. Fogs and other dust factors are also vital this way, the supervisory system can control all the connected
for temperatures assessments or predictions. The temperature home appliances. Figure 5 shows a smart reading obtained
is measured periodically over the day to acquire an average from the supervisory system.
temperature. The second advantage of the proposed system is to obtain
A meter reading by the proposed system that includes the meter reading, device rating, and the billing information.
mainly the power consumption and rating of the devices and As shown in the figure 7, the user can acquire rating of the
further to generate the billing is all automated. Users should devices based on the energy utilization from the supervisory
first configure the supervisory system by selecting the smart system. The device rating particularly helps the user in main-
electric meter to enable the readings. There are several options taining the devices utilization that eventually helps in reducing
for a user for instance, he/she can key in the meter number for the electricity consumption and the bill. The information
getting information and the bills. Moreover, not only newer but maintained by the supervisory system further helps in avoiding
also the previous bills and information can be extracted, like the unfair meter readings that some time happens while taking
real power, apparent power, power factor, supply voltage and meter reading manually due to human errors. Moreover, the
energy consumption by the devices. This dynamic supervisory electricity company can obtain the accurate energy consump-
system calls the data from the meter at the rate of 10kb/s tion information generated by the supervisory system. In the
after every two seconds. Hence, the system requires or utilize result, the proposed system raises the efficiency of the billing
very low bandwidth that does not affect the other applications system and reduces the cost of operation. Additionally, helps
running by the user over the internet. in controlling the power consumption, reduces the illegal usage
and load on the power grid. Suppliers can also identify any
illegal activities by observing the energy consumption at grid
side and the userŠs side. The other main advantage of the
proposed supervisory system is paying the electricity bill
electronically in an easy manner.
Furthermore, the user with the help of supervisory system
can control and maintain the electricity usage according to
the industrial rules and standards. Accordingly, the user can
monitor the units to keep the energy consumptions in the
given range and avoid billing at higher rate. The system
operation performs based on certain association rules that are
generated by Apriori algorithm and helps is enhancing the
system performance. The system performed with maximum
possibilities with the increase in selected association rules.
System association rules are designed according to the en-
vironment of the appliances as given in Table II. System
operations are varying for different times and association rules.
VI. C ONCLUSIONS AND F UTURE W ORK
Fig. 5. Smart Reading in the Supervisory system
In this paper, Apriori algorithm for a smart home automation
and metering system using IoT is presented. The main purpose
V. S YSTEM I MPLEMENTATION AND O PERATION of this system is to control the home appliances and electronic
The supervisory system provides a very easy and simple in- devices with the help of a supervisory system. The supervisory
terface to perform the desired operations. The system depends system is designed in such a way that everyone can access
 ICC2017: WT04-5thIEEE International Workshop on Smart Communication Protocols and Algorithms (SCPA 2017)

TABLE III
A SSOCIATION RULES FOR SYSTEM PERFORMANCES .

Winter Session
Rules Confidence Result Status
RW1 : (L >= 10hr)( H >= 10hr) → (AC <= 1hr) 49.07% Rejected
RW2 : (H >= 10hr ( AC <= 1hr) → (L >= 10hr) 53.92% Selected
RW3 : (L >= 10hr) → (H >= 10hr ( AC <= 1hr) 25.78% Rejected
... ... ...
Summer Session
RS1 : (L >= 10hr)( H >= 10hr) → (AC <= 5hr) 57.7% Selected
RS2 : (H >= 10hr ( AC <= 5hr) → (L >= 10hr) 53.92% Selected
RS3 : (L >= 10hr) → (F >= 10hr( AC <= 1hr) 45.78% Rejected
... ... ...

it. The system provides users with the ability to control, [10] Somayya Madakam, R Ramaswamy, and Siddharth Tripathi, “Internet
manage the electronic devices, can monitor the consumption of things (iot): A literature review,” Journal of Computer and Commu-
nications, vol. 3, no. 05, pp. 164, 2015.
of electricity, and to pay the electricity bills, securely and [11] Biljana L Risteska Stojkoska and Kire V Trivodaliev, “A review of
reliably. The supplier can also monitor and accordingly and internet of things for smart home: Challenges and solutions,” Journal
maintain the issues in an efficient manner. As a future work, of Cleaner Production, vol. 140, pp. 1454–1464, 2017.
[12] Guanglou Zheng, Rajan Shankaran, Mehmet Orgun, Li Qiao, and Kashif
the work can be extended in different directions, like by Saleem, “Ideas and challenges for securing wireless implantable medical
applying it on a real testbed with multiple smart homes to devices: A review,” IEEE Sensors Journal, 2016.
conduct performance evaluation of the complete system with [13] Raja Wasim Ahmad, Abdullah Gani, Siti Hafizah Ab Hamid, Mo-
hammad Shojafar, Abdelmuttlib Ibrahim Abdalla Ahmed, Sajjad A
enormous load. Moreover, the optimization methods can be Madani, Kashif Saleem, and Joel JPC Rodrigues, “A survey on energy
enhanced according to the performance. Other evolutionary estimation and power modeling schemes for smartphone applications,”
algorithms can be involved to enhance the selection process International Journal of Communication Systems, 2016.
[14] João Santos, Joel JPC Rodrigues, Bruno MC Silva, João Casal, Kashif
and consumption. Saleem, and Victor Denisov, “An iot-based mobile gateway for intel-
ligent personal assistants on mobile health environments,” Journal of
ACKNOWLEDGMENT Network and Computer Applications, vol. 71, pp. 194–204, 2016.
The work of Kashif Saleem is supported by the National [15] João Santos, Joel JPC Rodrigues, João Casal, Kashif Saleem, and Victor
Plan of Science, Technology and Innovation (MAARIFAH), Denisov, “Intelligent personal assistants based on internet of things
approaches,” IEEE Systems Journal, 2016.
King Abdulaziz City for Science and Technology, Kingdom [16] Qi Jing, Athanasios V Vasilakos, Jiafu Wan, Jingwei Lu, and Dechao
of Saudi Arabia, Award Number (12-INF2817-02). Qiu, “Security of the internet of things: perspectives and challenges,”
Wireless Networks, vol. 20, no. 8, pp. 2481–2501, 2014.
R EFERENCES [17] Anam Sajid, Haider Abbas, and Kashif Saleem, “Cloud-assisted iot-
[1] Luigi Atzori, Antonio Iera, and Giacomo Morabito, “The internet of based scada systems security: A review of the state of the art and future
things: A survey,” Computer networks, vol. 54, no. 15, pp. 2787–2805, challenges,” IEEE Access, vol. 4, pp. 1375–1384, 2016.
2010. [18] Rakesh Agrawal, Tomasz Imieliński, and Arun Swami, “Mining asso-
[2] Sean Dieter Tebje Kelly, Nagender Kumar Suryadevara, and Sub- ciation rules between sets of items in large databases,” Acm sigmod
has Chandra Mukhopadhyay, “Towards the implementation of iot for record, vol. 22, no. 2, pp. 207–216, 1993.
environmental condition monitoring in homes,” IEEE Sensors Journal, [19] Rakesh Agrawal, Ramakrishnan Srikant, et al., “Fast algorithms for
vol. 13, no. 10, pp. 3846–3853, 2013. mining association rules,” in Proc. 20th int. conf. very large data bases,
[3] Nikhil Singh, Shambhu Shankar Bharti, Rupal Singh, and Dushyant Ku- VLDB, 1994, vol. 1215, pp. 487–499.
mar Singh, “Remotely controlled home automation system,” in Advances [20] Rajeev Piyare, “Internet of things: ubiquitous home control and
in Engineering and Technology Research (ICAETR), 2014 International monitoring system using android based smart phone,” International
Conference on. 2014, pp. 1–5, IEEE. Journal of Internet of Things, vol. 2, no. 1, pp. 5–11, 2013.
[4] Eric Brown, “Who needs the internet of things?,” 2016. [21] Kumar Mandula, Ramu Parupalli, CH AS Murty, E Magesh, and Rutul
[5] Ala Al-Fuqaha, Mohsen Guizani, Mehdi Mohammadi, Mohammed Lunagariya, “Mobile based home automation using internet of things
Aledhari, and Moussa Ayyash, “Internet of things: A survey on en- (iot),” in Control, Instrumentation, Communication and Computational
abling technologies, protocols, and applications,” IEEE Communications Technologies, International Conference on. IEEE, 2015, pp. 340–343.
Surveys & Tutorials, vol. 17, no. 4, pp. 2347–2376, 2015. [22] Shiu Kumar, “Ubiquitous smart home system using android application,”
[6] Bruno MC Silva, Joel JPC Rodrigues, Isabel de la Torre Díez, Miguel arXiv preprint arXiv:1402.2114, 2014.
López-Coronado, and Kashif Saleem, “Mobile-health: A review of [23] Elisa Spanò, Luca Niccolini, Stefano Di Pascoli, and Giuseppe Ian-
current state in 2015,” Journal of biomedical informatics, vol. 56, pp. nacconeluca, “Last-meter smart grid embedded in an internet-of-things
265–272, 2015. platform,” IEEE Transactions on smart grid, vol. 6, no. 1, pp. 468–476,
[7] Dhananjay Singh, Gaurav Tripathi, and Antonio J Jara, “A survey of 2015.
internet-of-things: Future vision, architecture, challenges and services,” [24] S Hasan Hussain, S Geetha, and M AmuthaPrabhakar, “Design and
in Internet of things, IEEE world forum on. 2014, pp. 287–292, IEEE. implementation of an adaptive model for sustainable home automation
[8] Kashif Saleem, Basit Shahzad, Mehmet A Orgun, Jalal Al-Muhtadi, using internet of things (iot),” Int J Adv Engg Tech/Vol. VII/Issue I/Jan.-
Joel JPC Rodrigues, and Mohammed Zakariah, “Design and deployment March, vol. 827, pp. 829, 2016.
challenges in immersive and wearable technologies,” Behaviour & [25] Nathan David, Abafor Chima, Aronu Ugochukwu, and Edoga Obinna,
Information Technology, pp. 1–12, 2017. “Design of a home automation system using arduino,” International
[9] Ejaz Ahmed, Ibrar Yaqoob, Abdullah Gani, Muhammad Imran, and Journal of Scientific and Engineering Research, vol. 6, no. 6, pp. 795–
Mohsen Guizani, “Internet-of-things-based smart environments: state 801, 2015.
of the art, taxonomy, and open research challenges,” IEEE Wireless
Communications, vol. 23, no. 5, pp. 10–16, 2016.