Professional Documents
Culture Documents
Ko: Design and Implementation of Intelligent Home Control Systems based on Active Sensor Networks 1177
Abstract —The ubiquitous home network has gained wide- sensing information on the surrounding environment, to
spread attentions due to its seamless integration into everyday support more intelligent and adaptive home services. It has the
life. This innovative system transparently unifies various home potential to control consumer home devices used in everyday
appliances, smart sensors/actuators and wireless life. Eventually, users will experience the convenience of
communication technologies. The ubiquitous home network performing ordinary activities and increased satisfaction
gradually forms a complex system to process various tasks. offered by adaptive home services.
Developing this trend, we suggest a new intelligent home Several conditions are required to reap advantages from the
control system based on a wireless sensor/actuator network ubiquitous home network. For instance, computing systems
(we call it as an “Active sensor network”). The proposed should integrate diversified sensing information to perceive
intelligent home control system divides and assigns various the current situation in the home area. Also, they should be
home network tasks to appropriate components. It can able to control various consumer home devices. The home
integrate diversified physical sensing information and control
system may become complex, as the number of sensors and
various consumer home devices, with the support of active
devices offered increases. Therefore, home network systems
sensor networks having both sensor and actuator components.
We develop a new routing protocol LQIR (Link Quality should be designed distributing various tasks into proper
Indicator based Routing) to improve the performance of our computational units to reduce complexity.
active sensor networks. This paper introduces the proposed In this work, we design an intelligent home control system
home control system’s design that provides intelligent services that can assign tasks to suitable components. Using a wireless
for users. We demonstrate its implementation using a real sensor network with actuator functionality [2], our system can
testbed.1 automatically gather physical sensing information and
efficiently control various consumer home devices. We call
Index Terms — Ubiquitous home networks, Intelligent home this network the Active sensor network. The system can
control systems, Active sensor networks efficiently distribute various tasks related to home network to
corresponding components and implement real ubiquitous
home services via smart sensors and actuators deployed in
I. INTRODUCTION home areas.
Recent progress in low-powered micro device designs and Fig. 1 gives an overview of the proposed intelligent home
wireless communication technologies enables the Ubiquitous control system. The system, being supported by underlying
Computing paradigm [1] to become a reality. This paradigm active sensor networks (i.e., wireless sensor/actuator
promises to provide adaptive information and services to users networks) and additional wired/wireless communications
anywhere anytime using various devices. Numerous technologies, can control consumer home devices such as
applications have been proposed in different domains of our lamps, gas valves, curtains, TVs, and air conditioners. It also
everyday life such as in homes, offices, streets, buildings and interoperates with various mobile devices, such as PDA and
shops. Ubiquitous computing in the home domain can be mobile phones, using IEEE 802.11/802.3, RFID and CDMA
distributed in different scales throughout our daily lives. technology.
Developing killer applications and services for ubiquitous Our intelligent home control system based on active sensor
home will confer important commercial value. networks consists of various software components as follows.
A number of projects and research [3]-[9] have developed
ubiquitous home network applications. Compared to z Service Components: These components represent some
traditional home networks, the in-progress ubiquitous home ubiquitous home services provided by our system.
network collects user activity patterns, as well as physical Examples include services for home automation, home
security and home management.
1
Changsu Suh was with the Graduate School of Information & Computer z Decision Component: The component recognizes the
Engineering, Ajou University. Now he is with the R&D department of current home environment via active sensor networks, and
Hanback Electronics, Daejeon, Korea (e-mail: scs@ajou.ac.kr) enables the service components to select the appropriate
Young-Bae Ko is an associate professor in the School of Information &
Computer Engineering, Ajou University, Korea (e-mail: youngko@ajou.ac.kr) service.
Contributed Paper
Manuscript received July 7, 2008 0098 3063/08/$20.00 © 2008 IEEE
Authorized licensed use limited to: University of Greenwich. Downloaded on August 21,2023 at 20:04:14 UTC from IEEE Xplore. Restrictions apply.
1178 IEEE Transactions on Consumer Electronics, Vol. 54, No. 3, AUGUST 2008
Fig. 1 Overview of the Proposed Intelligent Home Control System based on Active Sensor Networks
z Sensing Component: The component gathers sensing data II. RELATED WORK
and special event information from the active sensor In this section, we briefly survey the existing works for
networking infrastructure deployed in home environments. intelligent home network systems and, based on their main
This sensing component provides this information to the contributions, try to classify them into three types: Decision
decision component. The decision component then Support oriented, Service Provision oriented and Real
adaptively selects the correct home services based on the Implementation oriented.
current home state of affairs. First, some work has focused on how to make the decisions
z Control Component: The component instructs special for the home networks more efficiently. For example, the
control commands to the deployed actuators, such as relay MavHome project [3] focused on the design of intelligent
switches or Infra-Reds. These provide methods to control agents and the role of prediction algorithms, based on the
and handle various consumer home devices such as lamps, inhabitants’ repetitive actions to automatically provide
gas valves, TVs and air conditioners. adaptive home services. The Aware Home project [4] deals
z Interaction Component: The component manages with decisions about the inhabitants’ location and activity
interaction between our home control system (i.e., the recognition within a house, utilizing multi-camera tracking,
control component) behind a sink node and the active audio/video sensors and automated separation of sound
sensor networks deployed in a home domain. sources.
z Additional Communication Components: These Secondly, the research effort for providing more convenient
components manage some other networking technologies and intelligent services for home environments has also been
(beyond the active sensor networks), such as WLAN, conducted. Sival et al.[5] utilized pressure-based floor sensors
Ethernet, RFID and CDMA. These technologies leverage to achieve efficient retrieval and summarization of video and
our provided home services to become more effective. audio data. When a pressure-based floor sensor detects the
person's movement, the information is analyzed to index the
This paper details our proposed system and its video and audio data. This demonstrates efficient multimedia
implementation. The contributions of our work are as follows: services for the human movement patterns. Bae et al.[6]
First, we propose a more intelligent home control system proposed TV-oriented architecture to manage data
based on active sensor networks and design it to be more broadcasting services, based on a common application
flexible with a software component-based approach. Second, platform middleware. Stankovic et al.[7] presented
we develop a novel sensor routing protocol to gain adaptive challenging issues for the provision of healthcare services in
control of smart sensors and actuators. Third, we implement the home arena.
hardware and software utilized in our system. Fourth, we Last, more practical research to emphasize a real
show how real ubiquitous home scenarios operate and are implementation has been done as well. Mori et al.[8]
implemented successfully in our system. suggested a ubiquitous sensing room equipped with cameras,
Authorized licensed use limited to: University of Greenwich. Downloaded on August 21,2023 at 20:04:14 UTC from IEEE Xplore. Restrictions apply.
C. Suh and Y.-B. Ko: Design and Implementation of Intelligent Home Control Systems based on Active Sensor Networks 1179
A. Smart Sensors and Actuators C. LQIR – Link Quality Indicator based Routing Protocol
We have developed the smart sensor nodes that are classified Each node should have multi-hop routing protocol to
into two different types of a generic sensor and an actuator. automatically establish the wireless network between smart
Generic sensor typed nodes try either to detect the general nodes. We develop a new tree-based routing protocol named
physical sensing measurements such as temperature, humidity as “LQIR (Link Quality Indicator based Routing)”.
and light or check for the special events such as gas leaks, Conventional routing protocols, such as the one in [10], are
human movement and window status detection. Whereas, difficult to accommodate to dynamic topology variations and
actuator typed nodes can directly control consumer home to interact with our home control system. We design a new
devices. In our system, some actuators are deployed near the routing protocol specifically for home networks. Our
consumer home devices and connected to their electronic proposed protocol establishes the wireless network, based on
switch by the relay switch module (which is also developed by the LQI value measured from the RF radio. This idea of using
ourselves). The control ability of actuator is limited to turn LQI value for the routing purpose has been presented in the
on/off actions, because the relay switch module simply works ZigBee WPAN standard as well, but it is different from our
as an electronic switch. Our actuator typed node utilizes IR LQIR protocol in that an on-demand approach is utilized there
(Infra-Red) communication also in order to instruct complex to make a routing path and hence many control packets are
consumer devices such as TV, Air Conditioners and Air often required. The proposed tree-based LQIR works as
Cleaners. This collaboration among actuator nodes and follows. When forwarding the data packet to the sink, the
Authorized licensed use limited to: University of Greenwich. Downloaded on August 21,2023 at 20:04:14 UTC from IEEE Xplore. Restrictions apply.
1180 IEEE Transactions on Consumer Electronics, Vol. 54, No. 3, AUGUST 2008
For Decision For Services smart nodes. This component manages the generic sensor
nodes in the active sensor network. Our home system handles
Decision Component three packet types: Set-Command packet, Data packet and
Init Decision Service Event packet. The Set-Command packet is generated either to
Service Components
Init process Decision request the physical sensing data of interest or to configure the
entry 1
... special event to the generic sensor type nodes. The packet
includes the specific sensor device ID and its requested
attributes. The packet is transmitted from the Sensing
Command
Request Interface
Data/Event
Response Interface
Control Request
Interface Additional Component to the smart node’s Interaction Component. The
Control
Communication Data packet includes the general physical sensing information
Sensing Component Components
Component such as temperature, humidity and light. This information is
the response to the Set-Command packet generated from the
Sensing Component. The Event packet informs about specific
Active Sensor Networks events, such as detecting a person’s presence, locking the door
WLAN/Ethernet and detecting the presence of dangerous gases. This packet is
Interaction Interaction Interaction Interaction /RFID/CDMA
Component Component ... Component Component
configured by the Set-Command packet that includes
Light Gas Relay Switch Infra-Red descriptions of the specific event as attributes. Both Data and
Event packets are forwarded from the Interaction Component
Fig. 3 A Block Diagram for The Component-based Architecture. on the smart node to the Sensing Component based on our
LQIR protocol.
node selects the special node having the best LQI value
among neighbors. The routing topology is adjusted
dynamically, since nodes check neighbors’ LQI lists whenever B. Decision Component
transmitting data. Users can easily see the sensor network Appropriate home services are selected by the Decision
topology established in the home, since each packet contains Component, based on the physical sensing information
its forwarded routing path list in the packet header. Using this and events gathered by active sensor networks. The
routing path list our home system can discern the routing path Decision Component is provided the Command Request
from the system to each smart node. and Data/Event Response interfaces by the Sensing
We utilize the B-MAC [11] protocol for shared media access, Component. As shown in Fig. 3, it consists of Init
and a special narrow-band RF device that supports the LQI Decision part and Service Decision part. The “Init
value, based on the IEEE 802.15.4 standard [12]. We also Decision part” is the sum of initial process entries
develop a new topology viewer program to show the configured by householders. When a householder selects
established smart nodes’ topology in our home system. Fig. 2 an initial process entry, the selected entry generates
illustrates a snapshot of active sensor networks, captured by appropriate Set-Command packets to the active sensor
using our topology viewer program. networks though Command Request interfaces. The
corresponding smart node’s Interaction Component may
return the appropriate Data packet or Event packet to the
IV. THE PROPOSED HOME CONTROL SYSTEM home control system to respond to the commands. These
Our proposed system initially instructs some intentional response packets are forwarded to the “Service Decision
commands to smart nodes. It then receives the physical part” through the Data/Event Response interface. The
sensing data of interest, or a special event from the Service Decision part receives the physical sensing
corresponding sensors. These operations are managed by information and events detected in the current home areas
Sensing Component, and the gathered information is without additional operations. Service Decision part
forwarded to Decision Component. Decision Component can recognizes the current situation, and executes the adaptive
select the adaptive home service based on these detected home service from Service Components. For example, if
physical sensing data and special events. The selected services the householder selects the initial process entry for the
are operated with the help of Control Component and gas monitoring service in Init Decision part, the matched
Additional Communication Components according to the entry generates the Set-Command packet to receive the
predefined scenarios in Service Components. Fig. 3 shows the current gas level from gas sensors. The command packet
component architecture in our proposed home control system. is forwarded to active sensor networks through the
In this section, we detail each component’s operations. Command Request interface, and the corresponding gas
sensor nodes periodically transmit the Data packet
including the detected gas level in the house. The Service
A. Sensing Component Decision part performs the gas monitoring service from
The Sensing Component is designed to receive and request the Service Components based on the packets forwarded
the physical sensing data and specific events of interest from by the Data/Event Response interface.
Authorized licensed use limited to: University of Greenwich. Downloaded on August 21,2023 at 20:04:14 UTC from IEEE Xplore. Restrictions apply.
C. Suh and Y.-B. Ko: Design and Implementation of Intelligent Home Control Systems based on Active Sensor Networks 1181
C. Control Component Service Components include these home services. The Decision
Our system can elaborately command and control most Component selects an adaptive service, and it may be performed
consumer home devices, based on the actuator nodes. The using Control Component and Additional Communication
Control Component manages these actuator nodes. If a Components according to the predefined scenarios.
home service selected by Decision Component needs to We have designed and implemented seven different kinds
control the special consumer home device, the of intelligent home services on our proposed system. These
corresponding service component requests an appropriate services are classified into three types: first, home
device type and its actions to Control Component through automation services controlling air conditioners, air
the Control Request interface. On receiving the request cleaners and curtain movements, according to the weather.
from the service component, Control Component searches The second, home security services, detects potential
the special smart node ID to control the corresponding crimes and prevents gas explosions. The last is the home
home device. It inserts the searched node ID, its routing management services via the Internet. More details on these
path and attributes of interest to the Control-Command service operations are discussed in the next section.
packets. The Control-Command packet is a special
command to control home device in actuator nodes. It V. OUR IMPLEMENTATION DETAILS
includes descriptions about the consumer device type and
To show the feasibility of the proposed architecture, we
its actions as attributes. Thus, the packet is forwarded the
implement our home control system and active sensor
special actuator type node, based on our LQIR protocol,
networks with various consumer home devices in a house.
and the device is controlled according to the intentions of
We developed all the related hardware and software for our
the service component. In our home control system, all
system. In this section, we present the potential ubiquitous
tasks, related to controlling consumer home devices, are
home services provided by our system.
handled by Control Component.
Authorized licensed use limited to: University of Greenwich. Downloaded on August 21,2023 at 20:04:14 UTC from IEEE Xplore. Restrictions apply.
1182 IEEE Transactions on Consumer Electronics, Vol. 54, No. 3, AUGUST 2008
Authorized licensed use limited to: University of Greenwich. Downloaded on August 21,2023 at 20:04:14 UTC from IEEE Xplore. Restrictions apply.
C. Suh and Y.-B. Ko: Design and Implementation of Intelligent Home Control Systems based on Active Sensor Networks 1183
These pictures are uploaded to the web-server, so z Gas detecting service: The gas sensor periodically
inhabitants can see from anywhere. checks gas in the atmosphere and reports the values to
z IR communication service via Internet: Infra-red (IR) our home system to prevent gas explosion. On detecting
device can support IR communication with complex a large amount of gas, the gas sensor immediately
consumer home devices. TV, DVD, air conditioner and instructs the gas valve actuator node to turn off the
air cleaner can be controlled by our system. Our home valve using its relay switch module. In this situation,
control system provides controlling methods of the there is no latent time to exchange commands with the
Infra-Red sensors using the HTTP server. The user can home control system. The gas sensor autonomously
control these home devices using the web browser via operates, and later informs the home control system.
Internet.
z Talking service with visitors when nobody in home: V. CONCLUSION
In some cases, friends and family may visit the home Ubiquitous home networks excite new possibilities. We
while the resident is absent. The visitor activate the address a new intelligent home control system based on
doorbell, the doorbell event is generated in the active active sensor networks to make home networks more
sensor network. Our home control system automatically intelligent and automatic. We implement the proposed
supports a connection between inhabitant and visitors system and develop related hardware and software. We
through the CDMA modem. The user can communicate suggest new ubiquitous home scenarios based on the
using a cellular phone, and visitors also respond using a proposed system. We expect that our work contributes
mike and speaker in the door. towards the development of ubiquitous home networks. As
z Automatic control of air conditioner and air-cleaner a part of future work, we will apply IEEE 802.15.4a [13]
service: The home control system can deduce the standard technology in our home network systems to
proximate atmospheric conditions in the house using support location services.
temperature, humidity and pressure information from
sensor nodes. The home control server can adaptively
control the air conditioner and air cleaner using the REFERENCES
Infra-Red node. [1] M. Weiser, "The computer for the 21st century," ACM SIGMOBILE
z Automatic curtain control service: If it is cold or dark, Mobile Computing and Communications Review, vol. 3, no. 3, pp. 3-11, 1999.
the home control system decides that closing the [2] A. Rezgui, M. Eltoweissy, "Service-oriented sensor-actuator networks:
Promises, challenges, and the road ahead," Computer Communications, vol.
curtains is better for the inhabitants. Otherwise the 30, no. 13, pp. 2627-2648, 2007.
curtains are opened or closed according to the time of [3] S. K. Das, D. Cook, A. Bhattacharya, E. O. Heierman and T.-Y. Lin, "The
day (morning or night). The system can adaptively close Role of Prediction Algorithms in the MavHome Smart Home Architecture,"
IEEE Wireless Communications Magazine, vol. 9, no. 6, pp. 77-84, 2002.
or open the curtain with appropriate Control-Command
[4] C. D. Kidd, R. Orr, G. D. Abowd, C. G. Atkeson, I. A. Essa, B. MacIn-
packets. tyre, E. Mynatt, T. E. Starner and W. Newstetter, "The Aware Home: A Living
Authorized licensed use limited to: University of Greenwich. Downloaded on August 21,2023 at 20:04:14 UTC from IEEE Xplore. Restrictions apply.
1184 IEEE Transactions on Consumer Electronics, Vol. 54, No. 3, AUGUST 2008
Laboratory for Ubiquitous Computing Research," Lecture Notes in Computer Changsu Suh received the B.S. and M.S. degree in
Science, vol. 1670, pp. 191-198, 1999. Information and Communication Engineering from
[5] G. C. de Silva, B. Oh, T. Yamasaki and K. Aizawa, "Ubiquitous Home: the Ajou University, South Korea in 2004 and 2006,
Retrieval of Experiences in a Home Environment," IEICE Transactions on respectively. He is recipient of awards for his
Information and Systems, vol. E91-D, no. 2, pp. 330-340, 2008. academic achievement. His interest is in the research
[6] Yu-S. Bae, B.-J. Oh, K.-D. Moon and S.-W. Kim, "Architecture for areas of Wireless Sensor Networks, Home networks,
Interoperability of Services between an ACAP Receiver and Home Networked IEEE 802.11/802.15.4 MAC protocols, ZigBee
Devices," IEEE Transactions on Consumer Electronics, vol. 52, no. 1, pp. networks and implementation of real sensor
123- 128. 2006
platforms. E-mail: scs@ajou.ac.kr.
[7] J. Stankovic, Q. Cao, T. Doan, L. Fang, Z. He, R. Kiran, S. Lin, S. Son, R.
Stoleru and A. Wood, "Wireless Sensor Networks for In-Home Healthcare:
Potential and Challenges," Proc. of the Workshop on High Confidence
Young-Bae Ko is currently an Associate Professor in
Medical Devices Software and Systems, June 2005.
the School of Information and Computer Engineering
[8] T. Mori, H. Noguchi, A. Takada and T. Sato, "Sensing Room: Distributed
at Ajou University, Korea, leading the Ubiquitous
Sensor Environment for Measurement of Human Daily Behavior," Proc. of the
Networked Systems (UbiNeS) Laboratory. Prior to
International Conference on Networked Sensing Systems, June 2004.
joining Ajou University in 2002, he was with the IBM
[9] T. Yamazaki, "Ubiquitous Home," International Journal of Smart Home,
T. J. Watson Research Center, Hawthorne, New York,
vol. 1, no. 1, pp. 17-22, 2007
as a research staff member in the Department of
[10] J. Hill, R. Szewczyk, A. Woo, S. Hollar, D. Culler, and K. Pister,
Ubiquitous Networking and Security. He received his
"System architecture directions for network sensors," Proc. of the
Ph.D. degree in computer science from Texas A&M
International Conference on Architectural Support for Programming
University, and B.S. and M.B.A. degrees from Ajou University. His research
Languages and Operating Systems, November 2000.
interests are in the areas of mobile computing and wireless networking. In
[11] J. Polastre, J. Hill, and D. Culler, "Versatile Low Power Media Access
particular, he is working on mobile ad hoc networks, wireless mesh/sensor
for Wireless Sensor Networks," Proc. of the ACM Conference on Embedded
networks, and various ubiquitous networked system issues. He was the
Networked Sensor Systems, November 2004.
recipient of a Best Paper award from ACM Mobicom 1998. He has served on
[12] IEEE Computer Society, MAC and PHY Specifications for Low-Rate
the program committees of several conferences and workshops. He also serves
Wireless Personal Area Networks (LR-WPANs), IEEE 802.15.4 TM-2003.
on the editorial board of ACM Mobile Computing and Communications
[13] IEEE Computer Society, Low-Rate Alternative PHY Task Group of IEEE
Review. See http://uns.ajou.ac.kr for further details.
802.15.4 (TG4a) for the ranging and location, IEEE 802.15.4a-2007.
Authorized licensed use limited to: University of Greenwich. Downloaded on August 21,2023 at 20:04:14 UTC from IEEE Xplore. Restrictions apply.