Professional Documents
Culture Documents
3, JUNE 2017
Abstract—The world has entered into a “smart” era. One area is an intelligent control, or an interface that allows for manage-
becoming smart is the place where we live—homes. Smart homes ment of the system. The final component is home automation,
are expected to be equipped with numerous sensors to con- where any product contained in the home is connected in some
tinually monitor, sense, and actuate the space. The data from
these sensors can be used to provide various types of services by way to a service or system from an outside source [2]. In this
automating common tasks while causing minimal disruption to type of smart home, the main point of emphasis is the use
daily life. In order to provide these services, a system must have of remote access to activate or control a part of the smart
sufficient intelligence to predict future events based on its obser- home. However, what about having a smart home that can
vations. This paper first examines the requirements for smart act on its own? This is the direction that researchers have
home predictions. It then comprehensively reviews prediction
algorithms and variations that have been proposed and inves- been leaning toward, so there must be some way to allow the
tigated in smart environments, such as smart homes. It is these smart home to think for itself in order to complete certain
prediction algorithms that provide the intelligence required by a tasks.
smart home. Comparisons are also made upon these prediction When smart homes were first introduced, the idea was to
algorithms on their features and models. have a system in place for a resident to interact with the
Index Terms—Intelligent systems, Internet of Things, system through an interface to change some aspect according
prediction methods, smart homes. to their own preference. Though this is still a key component
for a smart home, progress continues to be made to make
I. I NTRODUCTION the system more automated. This has led to researchers trying
to implement an artificial intelligence into the environment to
PREDICTION algorithm uses information from past
A experiences to anticipate future events. These algorithms
can be used in situations where data from a specific sample
make decisions based on previously gathered information or
prior inputs from the resident [3]. With artificial intelligence
in play, this should allow a smart home to act on its own.
space can be collected over a period. This makes prediction
This all seems ideal, but there still needs to be a way for the
algorithms ideal for use within smart environments. The term
smart home to make suitable decisions. To do so, prediction
smart environment can be used to refer to homes, office build-
algorithms have been introduced in many smart home systems
ings, power grids, etc. In this paper, the home context should
to analyze information and make a choice that is appropriate
be assumed unless stated otherwise.
for situation at hand.
A smart home can be defined in many ways. One defini-
One example of a smart home that has been developed
tion is dwelling that contains sensors and device controllers in
to act more or less independently is the managing an adap-
order to enhance one or more aspects of inhabitants’ lives [1].
tive versatile home (MavHome) [1], [4]. The researchers at
However, this definition does not fully encompass what a smart
the University of Texas at Arlington created this smart home
home is or how it operates. Another potential way to describe
project to have the home act as a “rational agent,” meaning
a smart home is that it consists of three main components.
that the MavHome wants to make the resident as comfort-
One component is an internal network, which is a way of
able as possible while operating at maximally efficiency. The
communication such as wires or cables. Another component
MavHome architecture is similar to many other smart homes in
Manuscript received October 6, 2016; revised December 6, 2016; accepted that it employs four layers: decision, information, communica-
January 31, 2017. Date of publication February 13, 2017; date of current tion, and physical layers. The decision layer uses knowledge
version June 15, 2017. This work was supported by the National Science given to it from the information layer to select what action
Foundation under Grant 1408165.
S. Wu is adjunct with the School of Computer and Software, Nanjing needs to take place. The information layer gathers and stores
University of Information Science and Technology, Nanjing 210044 China, all data pertinent to any decision making process. The com-
and also with the Department of Computer Science, Ball State University, munication layer enables the flow of information or request
Muncie, IN 47306 USA (e-mail: swu@bsu.edu).
J. B. Rendall, M. J. Smith, S. Zhu, and J. Xu are with the Department of between any two parts of the system. Lastly, the physical
Computer Science, Ball State University, Muncie, IN 47306 USA. layer is all the hardware that is included within the smart
H. Wang is with the Department of Electrical and Computer Engineering, home environment. MavHome uses this structure, along with
University of Massachusetts Dartmouth, Dartmouth, MA 02747 USA (e-mail:
hwang1@umassd.edu). prediction algorithms, to enable the smart home to be a more
Q. Yang is with the Department of Computer Science, Montana State automated space. Some other work on this interest can be
University, Bozeman, MT 59717 USA (e-mail: qing.yang@montana.edu). found in [5]–[14].
P. Qin is with the School of Computer Science and Control Engineering,
North University of China, Taiyuan 030051, China (e-mail: qpl@nuc.edu.cn). This paper is a survey on several different prediction algo-
Digital Object Identifier 10.1109/JIOT.2017.2668061 rithms that have been or are being used at smart homes. It will
2327-4662 c 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
WU et al.: SURVEY ON PREDICTION ALGORITHMS IN SMART HOMES 637
discuss the different areas within the home environment that states; and 3) the observational model which gives the rela-
use prediction algorithms. It will also compare the merits and tion between two states [20]. A variation of HMM is known
flaws of the algorithms presented and possible adaptations to as the task-based Markov model (TMM). In TMM, actions
enhance the overall quality of the prediction algorithms used or observations form a cluster of data that can be treated as
in smart homes. hidden [21]. So, the prediction analysis is still done in a sim-
This paper is organized as follows. We first introduce the ilar fashion in TMM as in HMM, but with a different way of
background of prediction algorithms, including system models gathering initial states.
and data in Section II. Then, Section III extensively reviews Similar to Markov models, BN are also used to represent
the literature prediction algorithms, compares them upon their the probability of an event occurring based on previous obser-
features and models. Next, Section IV discusses the enhance- vations. In fact, HMM is a model that is contained within
ments and variants of the prediction algorithms. Finally, this the dynamic BN class, where the BN contains a time con-
paper is concluded by Section V. straint [17]. Therefore, there must be some way to differentiate
the BN structure from the original Markov model. Note that
a BN is defined as a graphical model that demonstrates the
II. BACKGROUND
joint distribution for any given set of random variables. So
A. System Models in any BN, the system can be expressed as a directed acyclic
In order to better understand how prediction algorithms graph to demonstrate how different processes are connected
function, one must first understand the context in which they and where dependencies exist [22]. It can be seen here how,
are used. The idea of predicting how a person will interact with if certain data has been collected to describe a sequence of
an environment is a hard concept to grasp directly. This is why events, a BN can be used (similar to Markov models) to find
certain statistical models are typically used within prediction out what is the next likely event that will occur.
algorithms to better show how different tasks may predict Though both HMM and BN are useful in modeling a system
future events within a system. A major benefit to modeling with missing information, there are potential problems in either
the system is that based any all data collected by sensors, it is model that can cause an error in the prediction of future events.
not likely that this will provide the exhaustive set of activity For example, in HMM, there is no hierarchical structure and
sequences that can possibly occur within the smart home [15]. has trouble to process through a large amount of data [16].
Similarly, probabilistic models tend to be even better than pure This just means that HMM cannot determine what process or
statistical models since they allow for some randomness, time state is most important or what process must come first with-
complexity, and dependency on other variables, and are there- out being given the information explicitly. As noted before,
fore, more fitting to how human behavior within a home can a BN is a more general way of predicting human behavior
and should be modeled [16]. The main purpose of using mod- than a Markov model, but a BN still has its own pitfalls. A
els is to more accurately predict possible outcomes based on major dilemma in using a BN is that it is rigid and has trouble
previous events occurring, and the most commonly referred to adapting the “exact probabilistic inference” [16]. This refers
models used by prediction algorithms within smart homes are to the inability of BN to determine what type of distribution
the Markov model and Bayesian networks (BNs). is being used in the system. So when a prediction algorithm
The Markov model has been a well-established model for uses either of these types of models, there is a potential for
many years. The model says that at any given point in time of error, though the error should be minimal as long as the model
a system, the state at which the system is just depends upon meets its required specifications.
the direct previous state and independent of all other states
beforehand [17]. This has led many to describe a Markov B. System Data
model as a way to determine what is the “most probable event 1) Activities of Daily Living: Prediction algorithms work on
sequence” for any given system [18]. This basically means historical data to calculate potential future events. Each user’s
the state of the system is recorded and then a sequence of actions within a smart home environment provide the data
states is developed over a certain time interval. From this, a for the prediction algorithm to operate to its full capability. In
probability can be estimated of what is most likely to occur general, these actions conducted by a user fall into the category
next. This type of general Markov model leads to many dif- known as activities of daily living (ADLs).
ferent submodels that are also useful in prediction analysis. An ADL can be one of many different activities, though
The most referenced Markov model in smart home prediction gathering the information from these actions is typically per-
algorithms is the hidden Markov model (HMM) [19]. This spe- formed by using a sensor network. The sensors that are used
cialized model is used to model human behavior and maintains can be either wearable that go directly on a person, also
the Markov property. However, HMM works on observations known as wireless body sensor area network, or nonwear-
where the states of the system is unknown or hidden [17]. The able that remain stationary in the environment. These sensors
idea behind this is that even if a system is missing a piece can collect the information that could include a user’s loca-
of data at any given time, it can use previous knowledge to tion, movements, or activities [23]. The use of ADLs at a
make an educated decision on how to continue processing. smart home has its own challenges. One potential flaw is that
HMM could also be considered to consist of three param- though ADLs represent a sequence of events, some activities
eters: 1) the initial probability of the beginning state; 2) the can occur in different orders or together with a set of other
transition matrix that shows the probability to change between activities [24]. This means that any ADL, though beneficial
638 IEEE INTERNET OF THINGS JOURNAL, VOL. 4, NO. 3, JUNE 2017
to use in predicting inhabitant activities, could lead to com- user [30]. Also, information must be communicated to the
plexity since they are not necessarily identical for the same user by a device that is close enough for the user to notice it.
activity. Another possible problem is the issue of intersubject Roy et al. [30] analyzed daily routines of inhabitants to
variability, which refers to the situation that both inhabitant reveal that patterns exist in daily routines. They also realize
and noninhabitant activities can happen and affect how the that routines likely change over time; however, this change
ADL is recorded, resulting in a potential error [25]. can be classified as infrequent and random. Additionally,
When a sensor gathers information on an ADL, not only Roy et al. [30] noted that a smart home usually contains
does it get data related to the activity but will also record multiple paths between two locations, but they observe
the timestamps of events occurring within the activity. So, that inhabitants usually exhibit a most likely path. With
although any given ADL occurs as a sequence of events, the these observations, they assume inhabitant location to be a
use of the duration of each step along with the sequence will piece-wise stationary, ergodic, and stochastic process. Based
allow for better activity prediction [26]. Since there are times on this assumption, the previously discussed LeZi update
when sensors can malfunction or inhabitants can adjust their algorithm is reasonable to be used for a prediction algorithm
routine, the use of duration can help better compare an activity for this domain.
history. All of the ADLs can then be used by a prediction
algorithm to allow the smart home to both help the user and III. A LGORITHMS
make the house more comfortable. This also can help reduce
The architecture of smart homes is often represented as a
the cost of certain activities by being better prepared. One
layered model. A four-layered model is proposed by [31]. The
example of this would be to reduce energy consumption. An
top layer of this model is the “decision layer,” which bears the
experiment done by Zhuang and Liang [27] demonstrated how
responsibility of determining actions for the smart home to
the use of ADLs can benefit the smart home in its energy use.
perform by using the data from other layers. The other layers
Overall, an ADL is a resource that, when used, allows
generate data for the decision layer through communication
prediction algorithms to be extremely accurate in foreseeing
with physical hardware and storing the information obtained.
when and where future activities could occur within a smart
In order to select actions, the decision layer implements
environment.
one or more prediction algorithms, which will infer inhabitant
2) Location: The ability to predict the current location or
actions to occur next with a large probability. Using the result
the next location of inhabitants in a smart environment is
of the prediction algorithms, the decision layer can choose to
important to the functionality of the system as a whole. It is
take one or more actions. One possible choice is to automate
generally not desirable to manipulate a portion of the environ-
the predicted action. Another possibility is to deliver notifi-
ment based on a prediction if the inhabitants are not currently
cations or reminders if the predicted action is not performed
interested in that part of the environment.
within a specified time threshold. This section reviews some
One example of location prediction is performed by
well-know prediction algorithms used in the literature of smart
Petzold et al. [28] in the context of an office building. Two
environments.
major applications of next location prediction within an office
are proposed. First, a device could be placed outside each
office in the building to display the predicted location of a A. Active LeZi
person when he is away from his office. This would enable Although Active LeZi has been a widely accepted compres-
any visitors to decide if they should go to the predicted loca- sion algorithm, it is used in prediction the context of smart
tion or return at a later time. The second application is call homes. One application of Active LeZi is in the MavHome
forwarding. It could prevent a missed call by forwarding the project at the University of Texas at Arlington [32].
call to the predicted location when a person is away from Active LeZi is an improvement on the LZ78 algorithm [34]
his or her desk. Petzold et al. [28] have investigated Neural as proposed by Ziv and Lempel [33]. By parsing an input
networks, BNs, Markov, and state predictors for creating a string into one symbol at one time, LZ78 calculates the prob-
system with the described functionality. In addition to location, abilities for the next symbol at each context in the data. LZ78
a prediction of length of stay is necessary for these applica- will generate probabilities for the contexts that it sees from the
tions. Petzold et al. [28] claimed that this could be easily beginning of the data through the current symbol. The algo-
predicted with the same techniques that they have already rithm parses the input string into substrings “w1 , w2 , . . . , wn ”
investigated. In this scenario, it is also important to under- such that for any j > 0, the prefix (consists of all characters
stand when no prediction is better than an incorrect prediction. of wj except the last character) of wj is equal to some wi . As
Too many incorrect predictions could annoy users no longer of this prefix property, the substrings can be stored in a trie
interested in the predictions [28]. structure. Since LZ78 was designed as a compression algo-
Weiser [29] investigated the location prediction in a smart rithm, it has both an encoder and a decoder. For the purpose
home environment. He envisions a smart environment with a of prediction, only the encoder is required because there is no
high concentration of computing and communication capabil- need to produce the original data string from the trie struc-
ities and seamlessly integrated with its human users. In order ture which approximates a Markov model. An example of the
to provide information to the user, the system must be aware results of LZ78 parsing can be found in [35].
of the user’s location. The information needed by the user LZ78 cannot be used in practice for two main rea-
is strongly related to the current or future location of the sons. First, data that crosses phrase boundaries is lost.
WU et al.: SURVEY ON PREDICTION ALGORITHMS IN SMART HOMES 639
TABLE I
S UMMARY OF P REDICTION A LGORITHMS LeZi generates a trie structure to hold the information. In a
similar fashion, the SPEED algorithm also uses a tree struc-
ture that contains the information to predict future events. The
other algorithms, however, do not share this same structure.
In the flocking algorithm, all the data is stored in individual
clusters that share common characteristics. On the other hand,
the Apriori algorithm operates on transactions stored within a
database. An optimization of the Apriori algorithm loads the
database transactions into a matrix structure in the memory
is not suitable for smart homes and proposed improvements. to avoid the time required for disk access. As for the Nash
One of these proposals is made by Zhang and Qi [46] who H-learning algorithm, it is also unique in how it retains infor-
observed that this algorithm takes too long to run in order to mation. Instead of centralizing all the data into a single source,
be practical for a smart environment. In order to overcome this this algorithm keeps all the preferences and information on the
problem, they propose optimizations. First, Zhang and Qi [46] individual users themselves before going through to find what
observed that the standard implementation of the Apriori algo- is most likely to occur.
rithm requires multiple scans of the database in order to 2) Model: One major commonality among any prediction
obtain the required information. To address this, they pro- algorithms is whether any of them relies on a Markov model
pose scanning the database once and storing the pertinent data or BN to function. Of the algorithms discussed in this paper,
in memory. Then the algorithm can run on data in memory both the SPEED and Active LeZi algorithms are based on
instead of taking the extra time to access a database. A Markov models and use a PPM approach to generate predic-
second improvement is proposed based on the idea that a tions. However, they determine the size and order of these
frequent itemset cannot be obtained by adding an additional models in a slightly different manner. For Active LeZi, the
item to a nonfrequent itemset. This reduces the number of order of the Markov model is determined by the length of
iterations that the support must be calculated for an itemset. the longest LZ78 string observed. However, for SPEED, the
Zhang and Qi [46] included experimental data that shows the order is determined by the number of levels in the deci-
execution time for their enhanced Apriori algorithm is only sion tree. Also, note that the Nash H-learning relies on a
a fraction of the execution time of the standard Apriori algo- Markov model since the Q-learning technique is adapted from
rithm. Their results also show that as the size of the problem a Markov model and Nash H-learning is itself an adaptation of
grows, execution time grows less quickly for their enhanced Q-learning. The other two mentioned algorithms operate with-
algorithm than the standard algorithm. out a Markov model or BN and rather work on rules. In order
Rashidi and Cook [47] discovered another problem with to determine its results, the Apriori algorithm evaluates the
the Apriori algorithm within a smart environment. The algo- candidate rules that it generates two metrics, namely support
rithm was originally designed to discover frequent itemsets, and confidence which are defined above. The flocking algo-
but a smart home environment also requires the ability to rithm uses three rules in order group similar objects together:
discover periodic itemsets. In order to address this problem, 1) alignment; 2) separation; and 3) cohesion.
Rashidi and Cook [47] designed a variant of the Apriori 3) Data Type: Just as the algorithms structures differ at
algorithm that is called frequent and periodic activity miner times, so does the information that they are gathering. They
(FPAM). It defines one hour and one day granules of time all retrieve data in order to predict future events, but they most
in order to recognize periodic activities. After FPAM deter- definitely do not all get the same data. For the Nash H-learning
mines its candidate itemsets, it classifies them as periodic or algorithm, it is mainly concerned with getting the location of
frequent. It is possible for an itemset to be both periodic and all users, and then uses this to try and conserve energy. Going
frequent; however, FPAM will report it as frequent only. a slightly different route, the flocking algorithm instead tries to
get all information pertaining to each users’ daily activities (or
ADLs). This can allow the smart environment to adapt to fit
F. Summary of Algorithms the users’ needs. The SPEED algorithm instead takes data in
All prediction algorithms that are used within a smart home terms of sequences of events. From this, it attempts to predict
environment have a common goal, which is to allow the system potential events that will occur in the future. However, the
to make an educated prediction based on previous information SPEED algorithm does not take time into consideration when
in order to make the users life easier and more comfort- doing this. Yet, note that the Apriori algorithm is often used in
able. Even though this is the main role for every prediction looking at temporal relations. Now, the Active LeZi algorithm
algorithm, they differ in their approaches to storing, interpret- does not focus on only one area to gather data from. Instead,
ing, and mining data for the system. The following sums up it can be used to get information on location and time that
these five algorithms presented above in both similarity and certain activities occur. As can be observed, although there is
difference in many aspects, which is highlighted in Table I. some overlap, these five algorithms gather data that pertains
1) Information Presentation: The most widely cited to various aspects of the environment.
prediction algorithm used in a smart home is the Active LeZi. 4) Genealogy: It seems common for a prediction algorithm
This algorithm, as previously stated, is an adaptation of the to be a modification or improvement of an existing algo-
LZ78 compression algorithm. As it parses the data, the Active rithm. For example, Active LeZi is an improvement to LeZi
642 IEEE INTERNET OF THINGS JOURNAL, VOL. 4, NO. 3, JUNE 2017
[45] D. Schweizer and H. Wache, “Learning frequent and periodic usage Jacob B. Rendall, photograph and biography not available at the time of
patterns in smart homes,” Ph.D. dissertation, School Bus., Univ. Appl. publication.
Sci. Arts Northwestern Switzerland (FHNW), Muttenz, Switzerland,
Feb. 2014.
[46] Y. Zhang and T. Qi, “Research on mining association behavior of smart
home users based on apriori algorithm,” in Proc. Int. Conf. Mater. Eng.
Inf. Technol. Appl., 2015, pp. 817–821.
[47] P. Rashidi and D. J. Cook, “Keeping the resident in the loop: Adapting Matthew J. Smith received the B.S. degree (summa cum laude) in computer
the smart home to the user,” IEEE Trans. Syst., Man, Cybern. A, Syst., science from Trine University, Angola, IN, USA, in 2014. He is currently
Humans, vol. 39, no. 5, pp. 949–959, Sep. 2009. pursuing the M.S. degree in software engineering at Ball State University,
[48] A. Achar, S. Laxman, and P. S. Sastry, “A unified view of the apriori- Muncie, IN, USA.
based algorithms for frequent episode discovery,” Knowl. Inf. Syst.,
vol. 31, no. 2, pp. 223–250, 2012.
[49] E. O. Heierman and D. J. Cook, “Improving home automation by dis-
covering regularly occurring device usage patterns,” in Proc. 3rd IEEE
Int. Conf. Data Min. (ICDM), Melbourne, FL, USA, 2003, pp. 537–540.
Shangyu Zhu, photograph and biography not available at the time of
[50] M. R. Alam, M. B. I. Reaz, M. A. M. Ali, and S. A. Samad, “Temporal
publication.
modeling of human activity in smart homes,” Informacije MIDEM,
vol. 41, no. 2, pp. 118–121, 2011.
[51] V. R. Jakkula and D. J. Cook, “Enhancing smart home algorithms using
temporal relations,” Technol. Aging, vol. 21, pp. 3–10, Jan. 2008.
[52] V. Jakkula and D. J. Cook, “Learning temporal relations in smart home
data,” in Proc. 2nd Int. Conf. Technol. Aging Canada, vol. 33. 2007.
[53] B. Gu et al., “Incremental learning for ν-support vector regression,” Junhong Xu, photograph and biography not available at the time of
Neural Netw., vol. 67, pp. 140–150, Jul. 2015. publication.
[54] B. Gu, X. Sun, and V. S. Sheng, “Structural minimax probability
machine,” IEEE Trans. Neural Netw. Learn. Syst., to be published.
[55] B. Gu and V. S. Sheng, “A robust regularization path algorithm for ν-
support vector classification,” IEEE Trans. Neural Netw. Learn. Syst., to
be published.
[56] Z. Pan, Y. Zhang, and S. Kwong, “Efficient motion and disparity esti- Honggang Wang (S’05–A’08–M’10–SM’13)
mation optimization for low complexity multiview video coding,” IEEE received the Ph.D. degree in computer engineering
Trans. Broadcast., vol. 61, no. 2, pp. 166–176, Jun. 2015. from the University of Nebraska–Lincoln, Lincoln,
[57] Z. Fu, X. Wu, C. Guan, X. Sun, and K. Ren, “Toward efficient NE, USA, in 2009.
multi-keyword fuzzy search over encrypted outsourced data with accu- He is currently an Associate Professor with
racy improvement,” IEEE Trans. Inf. Forens. Security, vol. 11, no. 12, the Department of Electrical and Computer
pp. 2706–2716, Dec. 2016. Engineering, University of Massachusetts at
[58] M. Marufuzzaman, M. B. I. Reaz, M. A. M. Ali, and L. F. Rahman, Dartmouth, Dartmouth, MA, USA. He has authored
“A time series based sequence prediction algorithm to detect activi- or co-authored over 100 papers. His current
ties of daily living in smart home,” Methods Inf. Med., vol. 54, no. 3, research interests include wireless health, body area
pp. 262–270, 2015. networks, cyber and multimedia security, mobile
[59] M. T. Moutacalli, A. Bouzouane, and B. Bouchard, “Sensors activation multimedia and cloud, wireless networks and cyber-physical systems, and
time predictions in smart home,” in Proc. 8th ACM Int. Conf. Pervasive big data in mHealth.
Technol. Related to Assistive Environ., 2015, Art. no. 37. Prof. Wang serves as the Chair/Co-Chair for several leading international
[60] S. Choi, E. Kim, and S. Oh, “Human behavior prediction for smart conferences and on the Editorial Board for several leading journals. He is an
homes using deep learning,” in Proc. IEEE RO-MAN, Gyeongju, Associate Editor-in-Chief for the IEEE I NTERNET OF T HINGS J OURNAL.
South Korea, 2013, pp. 173–179.
[61] S. T. M. Bourobou and Y. Yoo, “User activity recognition in smart homes
using pattern clustering applied to temporal ANN algorithm,” Sensors,
vol. 15, no. 5, pp. 11953–11971, 2015.
[62] J. Huang, Y. Meng, X. Gong, Y. Liu, and Q. Duan, “A novel deployment Qing Yang (GS’10) received the Ph.D. degree in
scheme for green Internet of Things,” IEEE Internet Things J., vol. 1, computer science from Auburn University, Auburn,
no. 2, pp. 196–205, Apr. 2014. AL, USA, in 2011.
He is a RightNow Technologies Assistant
Professor with the Gianforte School of Computing,
Montana State University, Bozeman, MT, USA. He
has authored or co-authored over 50 papers in pres-
tigious journals and conferences such as the IEEE
T RANSACTIONS ON D EPENDABLE AND S ECURE
C OMPUTING, IEEE INFOCOM, and IEEE CNS.
Shaoen Wu (M’14–SM’16) received the Ph.D. His current research interests include online social
degree in computer science from Auburn University, networks, trust assessment, vehicular networks, and the Internet of Things.
Auburn, AL, USA, in 2008.
He is currently an Assistant Professor of
Computer Science with Ball State University,
Muncie, IN, USA. He was an Assistant Professor
with the School of Computing, University of
Southern Mississippi, Hattiesburg, MS, USA, a Pinle Qin received the Ph.D. degree in electrical and
Research Scientist with ADTRAN Inc., Huntsville, computer engineering from the Dalian University of
AL, USA, and a Senior Software Engineer with Bell Technology, Dalian, China, in 2008.
Laboratories, Qingdao, China. His current research He is currently an Associate Professor with
interests include wireless and mobile networking, cyber security, cyber- the School of Computer Science and Control
physical systems, and cloud computing. Engineering, North University of China, Taiyuan,
Prof. Wu was a recipient of the Best Paper Award of IEEE ISCC 2008 China. His current research interests include com-
and ANSS 2011. He has served on the chairs and the committees of various puter vision, intelligent control, big data, and
conferences such as the IEEE INFOCOM, ICC, and Globecom, and as an machine learning.
Editor for several journals.