Wireless Sensor Network, 2010, 2, 571-583
doi:10.4236/wsn.2010.28069 Published Online August 2010 (http://www.SciRP.org/journal/wsn)
A Survey on Context-Aware Sensing for Body
Sensor Networks
Barbara T. Korel1, Simon G. M. Koo2
1Department of Computer Science, Brown University, Providence, USA
2Department of Mathematics and Computer Science, University of San Diego, San Diego, USA
Received May 14, 2010; revised May 25, 2010; accepted May 27, 2010
Abstract
Context awareness in Body Sensor Networks (BSNs) has the significance of associating physiological user
activity and the environment to the sensed signals of the user. The context information derived from a BSN
can be used in pervasive healthcare monitoring for relating importance to events and specifically for accurate
episode detection. In this paper, we address the issue of context-aware sensing in BSNs, and survey different
techniques for deducing context awareness.
Keywords: Context Awareness, Body Sensor Networks, Artificial Neural Networks, Bayesian Networks,
Hidden Markov Models
1. Introduction
Context is defined as “any information that can be used to
characterize the situation of an entity, where an entity can
be a person, place or physical object” [1]. Context aware-
ness can then be defined as detecting a user’s internal or
external state. Context-aware computing describes the
situation of a wearable or mobile computer being aware of
the user's state and surroundings, and modifying its behav-
ior based on this information [2]. Context awareness plays
a significant role in Body Sensor Networks (BSNs) be-
cause it interprets physical and biochemical signals coming
from the BSN, based on information regarding the current
state of the user and the state of the environment. Con-
text-aware sensing is an integral part of the BSN design to
achieve the ultimate goal of long-term pervasive health
care monitoring.
There are three main approaches that have been ap-
plied to deduce context in a sensor network: Artificial
Neural Networks, Bayesian Networks and Hidden Markov
Models. Research in context awareness or activity rec-
ognition using these methods has primarily been done in
wireless sensor networks or wearable sensor networks,
so the application of context-aware sensing in BSNs is
still new and faces many technical challenges. This paper
will address some of these issues raised, describe the
characteristics of each method, and discuss how these
algorithms handle the challenges that need to be faced in
context sensing for BSNs. As a note of credit, this paper
is inspired and largely based upon [3].
2. Context Awareness in BSNs
Wireless medical body sensor devices, either implantable
or wearable, are used to monitor a patient’s physiological
state including EKG, heart rate, blood pressure, oxygen
saturation and sweat volume/rate. The wireless BSN
framework is designed to provide such pervasive moni-
toring of the human body; this ultimately has a huge im-
pact on medical healthcare and monitoring vital signs of
elderly patients or patients with chronic cardiac disease.
BSNs present a method to continuously monitor physio-
logical parameters to detect life threatening abnormali-
ties that could lead to mortality. In addition to a patient's
vital signs, a person is physiologically very sensitive to
external context or environmental change. Such contex-
tual factors include the person's activity, current tem-
perature of the outside environment, time of day, etc. For
instance, if a body sensor detects a rapid increase in a
patient's heart rate, the patient might not be experiencing
a cardiac episode, but rather undergoing a change in his
physical activity such as jogging. By incorporating con-
text awareness into the BSN, environmental factors and
the state of the patient can be evaluated. Ultimately,
changes in the physiological state of the body can be
rationalized according to the events that triggered such
This work is supported in part by the Faculty Research Grant, College
of Arts and Sciences, University of San Diego.

B. T. KOREL ET AL. 572
changes.
There are various algorithms for context-aware sens-
ing that can deduce context in a BSN; each has different
characteristics and accomplishes different tasks. In many
applications studied, these approaches are actually used
in combination with one another to achieve context from
the environment. The first step in achieving context
awareness in a sensor network is to gather the low-level
sensor readings from all sensor nodes; these data read-
ings will always constitute as the input for context sens-
ing. It is often beneficial for the input data to be ordered
in some way, thus the data is typically clustered into
subgroups such that distance is small among data entries
in the same cluster and distance is large among data en-
tries from different clusters [4]. This is accomplished by
a clustering algorithm. To actually achieve context from
the sensor nodes, the input or clustered data must be as-
sociated with a context. In the domain of sensor net-
works, this process is known as classification, which
associates the input vectors to a context profile through
the means of user labels. Some classification algorithms
are only able to recognize context at a given instance in
time but not continuously. So a supervising layer is in-
troduced on top of the classification layer to extract con-
stant recognition of context. Algorithms performing at
the supervising layer are able to classify context transi-
tions and more closely model context events occurring in
natural human behavior.
3. Challenges of Context-Aware Sensing
In practice, BSNs for pervasive healthcare monitoring
will result in network applications operating in a variety
of different environments including a hospital operating
room, an elderly health clinic or a personal home setting.
Each of these environments varies substantially from one
another and yet the BSN framework must be adaptable
and distributed to accommodate for such different set-
tings. Due to the diversity of context-aware environ-
ments, the range of physiological conditions a patient
may experience, and the dynamic nature of BSNs them-
selves, many challenges arise for context-aware sensing.
Specific issues include overcoming sensor noise, node
failure and motion artifact in the network, integrating
multi-sensory data, allowing for smooth context recogni-
tion, providing long term/continuous usage of the appli-
cation, appropriately structuring the network in terms of
number of sensors, and selecting relevant features in the
BSN. These are the challenges context-aware sensing
faces and the various methods, discussed in later sections,
handle some of these issues in various ways.
3.1. Noise Resilience and Detection
Noise in a sensor network may result from sensor noise,
node failure or motion artifact [3]. The presence of noise
in the network from any of these sources may introduce
significant errors into the input data of the sensor net-
work. This data may contain missing sensor information,
malicious sensor readings or uncertain information; the
resultant input vectors will not contain an accurate rep-
resentation of the sensor readings. For context sensing in
BSNs, this could have unfortunate consequences because
in pervasive healthcare monitoring, detrimental actions
are usually taken based upon the sensed values. Specifi-
cally the cost of any “unclean” data can be very signifi-
cant since it is used for critical decisions [5]. The quality
of input sensor readings is crucial and the presence of
noise degrades this quality of data obtained from the
sensor network. Thus a context sensing algorithm must
be able to detect such malicious noise and reduce the
effect of noise in sensor data to appropriately model the
network.
3.2. Introduction of Smoothness Constraint
Human activity innately involves body movement that is
continuous in nature; if such a smoothness constraint were
enforced in BSNs, context could be recognized with a
higher accuracy based on natural human behavior [6].
Typically context is classified at a given instance in time,
but introducing a smoothness constraint means the system
must sense the transitions between individual context or
sequences of context. Context sensing in BSNs must be
able to capture such transitions to accurately recognize
context from the continuous flow of human movement in
time.
3.3. Adaptive On-Line Learning
Recognizing context in the real-world domain of BSNs is a
function that needs to remain flexible as new context may
be continuously added to the system and old context may
no longer be perceived. Thus the system needs to remain
adaptive to learn new context online from the sensor net-
work. Since real BSN applications will function over
longer periods of time, it is important that the system not
only be able to learn new context as it is presented to the
system, but also to not forget previously encountered con-
text that was learnt, so it doesn’t have to re-learn such con-
text again.
3.4. Input Data Dimensionality
A large number of sensors in the BSN may be necessary
in order to achieve accurate context recognition of a pa-
tient’s state/activity or if the system is to recognize a
large number of different contexts. By adding more sen-
sors to the network, context recognition can be achieved
with a higher accuracy [7]. However, two significant

B. T. KOREL ET AL.573
problems may arise from a high dimensionality of input
sensors. First, a substantial burden may be placed on the
power consumption and bandwidth of the system as
more sensors are added to the network. This is rather an
issue of reducing the transmission range and required
bandwidth; clustering data transmission will also reduce
power consumption [8].
The second problem is known as the curse of dimen-
sionality: as the number of sensor inputs increases, the
learning rate of the algorithm significantly slows down.
For a BSN to maintain good performance, a system com-
posed of a large number of sensors should not slow down
or decrease fault tolerance.
3.5. Feature Selection
If only the relevant sensor readings were applied to the
current context, irrelevant or redundant sensors could be
filtered out. This is beneficial because typically the
number of features in the BSN is numerous, especially if
there is a large number of sensors. However, only a small
subset of those features is necessary or relevant to recog-
nize the context [9]. Figure 1 illustrates how context is
extracted in a sensor network using feature selection.
Thus if the input data that is not useful in the decision
process of context recognition were not sent across the
wireless network, the dimensionality of data will be re-
duced. Essentially this allows for a decrease in data
transmission (implying less power consumption and
bandwidth) and efficient data mining of the BSN as only
relevant information is used in the sensor network [10].
4. Techniques for Context Recognition
This section presents a survey of various techniques that
have been used in context-aware sensing. The work by
Yang [3] served as a major inspiration to our study.
4.1. Artificial Neural Networks
The Artificial Neural Network (ANN) is used as a solid
clustering algorithm for context awareness in sensor net-
works. It is based on the biological nervous system of the
SensorASensorB
FeatureI
SensorMSensorN
FeatureK
Conte x t
...
...
Figure 1. Context extraction in a sensor network using fea-
ture selection.
brain that consists of a large number of small and simple
interconnected components: neurons. Each neuron can
perform its own set of computations, yet the network is
capable of performing powerful computations by combin-
ing the limited processing power of each element [7].
For sensor networks used in practice, the low-level
sensors will produce some level of noise no matter what.
One of the key advantages of using ANNs is that they are
still able to perform well despite the presence of this in-
evitable noise coming from data sensors. Another bene-
ficial characteristic of neural networks includes unsuper-
vised training of the input data, that is, the user of a
wearable computer does not need to spend much time
training it and the context learning for the system is not
limited to just the training phase. So as the user transi-
tions from context to context, the algorithm should learn
autonomously the context from the new input it receives
by recalculating its internal representation of the context
(known as on-line adaptation) [11]. Thus because the
data is able to approximate itself, the neural network can
feasibly add new context to the system when necessary,
without intervention from the user.
The following discusses two types of ANNs: Kohonen
Self-Organizing Map (KSOM) and KSOM with k-means
clustering. Table 1 presents the pros and cons of using
ANNs for context awareness and Table 2 summarizes
some studies and applications of ANNs in sensor net-
works.
4.1.1. Kohonen Self-Organizing Maps
The Kohonen Self-Organizing Map (KSOM) is a type of
unsupervised neural network which is used to cluster the
input vectors (low-level sensor readings from nodes) to a
discrete output space that is in the form of a grid-like
map. Just as described by ANNs, the outcome is similar
signals are mapped close to each other on the map and
dissimilar signals are mapped at greater distances from
each other [18].
Algorithm: The structure of the KSOM consists of an
input layer and an output layer: the input layer is essen-
tially the input vector of data and each input node is as-
signed a map-unit to introduce order among the input
vectors. The output layer is a grid of interconnected neu-
rons, usually as a one or two-dimensional array. Each
neuron in the output layer is connected to every single
neuron from the input layer and this connection is as-
signed a particular weight. In addition, every neuron is
also connected with its nearest neighbor nodes on the
grid-map. Figure 2 depicts the structure of a bi-dimen-
sional KSOM.
The KSOM is a competitive network in that each unit
in the output layer competes with the other output units
for a particular kind of input. So when a new input value
is presented to the KSOM, the input vector is compared
to each output neuron's weight vector. The output neuron
that has a weight vector closest to the input vector is se-

B. T. KOREL ET AL. 574
Table 1. Benefits and drawbacks of the ANN approach to context awareness.
Technique Advantages Limitations
ANN KSOM
1) Provides a means of visualizing the behavior of the
network.
2) Unsupervised training of input data, i.e., doesn't
require user participation.
3) Provides an efficient means to cluster data.
1) Stability-plasticity dilemma: KSOM does not re-
main adaptive over time.
2) Curse of dimensionality: high dimensionality of
input data results in a slow and less fault tolerant
algorithm.
ANN KSOM
with k-means
clustering
Overcomes stability-plasticity dilemma so it remains
adaptive and stable over time.
1) Requires labeled input vectors, meaning
2) User participation is necessary.
Table 2. Techniques in context-aware sensing.
Application Experiment Results
Schmidt et al.
(1999) [12]
Presents a layered real-time
architecture for context-aware
adaptation based on redundant
collections of low-level sensors;
the context is derived using
KSOM.
A prototype board consisting of 8
sensors, a PDA and a mobile phone
were used to demonstrate situational
awareness.
Experiments show that it is feasible
to recognize context using sensors
and that context information can be
used to create new interaction meta-
phors.
Lagerholm et al.
(2000) [13]
Uses unsupervised self-organizing
NNs to cluster ECG signals into
25 groups.
Cluster ECG complexes into classes
which are not predefined.
Using the MIT-BIH arrhythmia
database, the resulting KSOM clus-
ters exhibit a very low degree of
misclassification (1.5%).
van Laerhoven
et al. (2000) [14]
Shows an integrated approach
using KSOM for clustering,
along with K-nearest neighbor
for classifying different activi-
ties.
A pair of pants with accelerometers,
connected to a laptop to interpret raw
sensor data, are used to recognize
activities such as walking, sitting,
climbing stairs, etc.
The user has the ability to decide
what activities are learned at what
time, while the system remains
autonomous enough such that the
interaction is kept very minimal.
van Laerhoven
et al. (2001) [7]
Shows that Neural Network is
an ideal algorithm to analyze
data coming from a large num-
ber of small and simple sensors;
specifically KSOM with k-means
clustering.
A wearable system consisting of
several simple sensors used to learn
different, simple activities like sitting,
standing and walking; also used to
automatically start processes or tasks
depending on the current context.
Determining different context a
wearable computer can encounter, by
merely labeling them as they occur,
is still difficult to realize without
setting harsh constraints on system,
usually in terms of the available
context.
van Laerhoven
et al. (2001) [11]
Uses KSOM with k-means
clustering for classification of
incoming sensor data in a
real-time fashion.
Two accelerometers placed above the
knee to recognize simple, everyday
activities such as sitting, standing,
walking, running and bicycling.
KSOM can be very unstable in initial
phases; k-means algorithm added to
KSOM to overcome overwriting
previous inputs and to create a stable
topological mapping of sensor data.
Simelius et al.
(2003) [15]
Presents SOM for spatiotemporal
analysis and classification of
body surface potential mapping
(BSPM) data.
86 cardiac depolarization (QRS) se-
quences paced by a catheter in 18
patients, in which spatial BSPM dis-
tributions at every 5ms over the QRS
complex were presented to an un-
trained SOM.
This method has potential for deter-
mining abnormal ventricular activity.
Gao et al.
(2004) [16]
Presents a diagnostic system for
cardiac arrhythmias from ECG
data, using an ANN classifier
(based on a Bayesian frame-
work).
Bayesian ANN-based arrhythmia dia-
gnostic system determines a patient's
current condition in real-time, using
ECG signals.
At least 75% prediction accuracy for
classification in both the training and
test phases. Greater than 90% false
rate (cry wolf dilemma of medical
monitoring devices–a false alarm is
raised yet patient needs no attention)
prediction accuracy in both phases.
Thiemjarus et al.
(2006) [17]
Proposes a spatio-temporal SOM
that minimizes the number of
neurons involved while main-
taining a high accuracy in class
separation for both static and
dynamic activities.
Four accelerometers were placed on
left and right ankles and legs for a
simple physical exercise sensing ex-
periment involving sitting, standing,
steps, dem-plie, galloping, skipping,
etc.
Using standard SOMs, the average
performance was about 58%, and an
increase in the number of neurons
from 100 to 400 did not make a no-
ticeable difference. The use of spatio-
temporal SOM with a relatively
small number of neurons shows a
great improvement in performance.

B. T. KOREL ET AL. 575
InputvectorX
X
1
X
n
W
1
W
n
OutputLayer
...
Figure 2. The structure of a bi-dimensional KSOM.
lected as the winner node and is able to adapt itself more
towards the input. This means the winning output node
updates its weight vector to more closely reflect the val-
ues of the input vector. To introduce topological ordering
among related units, the neighboring nodes of the win-
ning node are also permitted to update their weight vec-
tors towards the input vector, but to a lesser degree. After
the first iteration of this algorithm with the given input
data, errors will typically exist when new signal readings
are introduced to the mapping [18]. However, after only
a few iterations of this algorithm, the data organizes it-
self in a structured and topological way such that similar
sensor signals activate neighboring units and different
signals activate different neurons [14]. Thus the KSOM
clusters n-dimensional input data from the sensors into
an array of neurons in an adaptive (meaning the neurons
in the grid “learn” to respond better for particular input),
and unsupervised fashion.
Application to Context Awareness: After the input
from sensor readings has been clustered by a clustering
algorithm, it becomes significantly easier to process or
classify the data. In terms of context awareness, KSOM
is in general a universal approach to process sensor data
because it does not require a priori knowledge of the
context and is able to perform learning without explicit
user supervision [14]. Thus using the KSOM to topo-
logically map sensor data is appropriate in applications
which may not contain labeled training data or where
activities are not well-defined. In terms of BSNs, this is
beneficial because it enables the system to not only de-
tect context that has not yet been defined by the user, but
it also allows the system to capture context that is unpre-
dictable and randomly appears in the system.
Any system in the real-world that is to deduce context
awareness from body sensors must have the requirement
of remaining adaptive over time. KSOMs have the capa-
bility to be adaptive; however this comes with its limita-
tions. The KSOM algorithm starts out highly adaptive
with a large learning rate, and then over time becomes
fixed so that it is no longer capable of learning anymore
[14]. This problem is known as the stability-plasticity
dilemma and is one of the main drawbacks to KSOMs
such that they are not adaptive over time. If the system
were designed to remain flexible, then overwriting pre-
viously learned instances would occur as neurons be-
longing to a learned cluster would gradually change to
other clusters. This is true because the KSOM has a fixed
structure and cannot infinitely grow over time, (i.e., it has
a limited number of context it can cluster, so the number
of clusters remains the same while the neurons adapt to
different clusters within the same set of clusters). Overall
the stability-plasticity dilemma prevents a system from
using KSOM for long-term use since the system either
would gradually become fixed or would become unstable
since previously learned context would be forgotten [7].
Limitations: A BSN should not be limited by the
number of sensors or the number of inputs in the network.
In practice, if the number of input increases in the system,
it should be able to handle a large number of input data
and not slow down. Unfortunately, KSOM suffers from
this curse of dimensionality and causes a high dimen-
sionality of input data to be a problem for this algorithm
[14]. As the number of sensors increases in the system,
the clustering algorithm must map the high input space to
a large output space and uses a lot of resources to do so,
resulting in a slow and less fault tolerant algorithm. This
problem becomes especially bad if there is a lot of noise
or irrelevant sensor nodes in the system, in which the
algorithm maps irrelevant context to its output space and
wastes many resources in the process. Thus the KSOM
proves to have limitations in the real-world domain
where the input space could be very large and irrelevant
context is undoubtedly present, causing a serious digress
in maintaining performance speed and fault tolerance.
4.1.2. KSOM with K-Means Clustering
The traditional KSOM has the disadvantage of “unlearn-
ing” context or overwriting prototypes on the map if the
algorithm is to remain adaptive, known as the stabil-
ity-plasticity dilemma. This is a major shortcoming for
context awareness in BSNs since the system should be
able to learn new context over a long period of time and
not forget old context. This problem can be overcome by
introducing a k-means clustering algorithm to the KSOM.
The KSOM still clusters the sensor input and preserves
map topology; the k-means clustering algorithm then
clusters labeled input vectors a second time and adds a
second layer to the structure [11]. Figure 3 illustrates
KSOM with the k-means clustering layer.
Advantages over KSOM: The k-means clustering
algorithm has two main differences from the traditional
KSOM: it is not topology preserving and it requires la-
beled input vectors (meaning the user needs to partici-
pate in the training phase to label incoming sensor data
with a context description) [11]. With just the KSOM,
clusters are unlabeled making the algorithm unaware of

B. T. KOREL ET AL. 576
X
1
X
n
W
j1
...
X
2
...
Input KSOMk‐means Classification
W
j2
W
j3
W
jk
walking
walking
walking
walking
...
Figure 3. KSOM with k-means clustering layers.
relevant context associated with clusters. However, in
k-means the user specifies the labels making the algo-
rithm aware of existing context so it knows not to over-
write clusters which contain relevant context. Also, be-
cause k-means is not topology preserving, the hierarchi-
cal structure allows the k-means sub-clusters to preserve
already clustered data when KSOMs’ topological map-
ping begins to overwrite previously learned prototypes.
Overall, the addition of a second layer with k-means
clustering allows a context-aware system in BSNs to
function over a long period of time while remaining
adaptive and stable.
4.2. Bayesian Networks
Bayesian Networks (BNs) are an appropriate method for
deducing context awareness by classifying context from
the associated sensor readings in the system. The Bayes-
ian Network is a form of a graphical probabilistic model
in which the structure of a BN is a directed acyclic graph.
The nodes in the graph signify random variables and the
directed arcs between nodes represent their causal de-
pendencies. Thus the set of random variables is the do-
main of interest and all of the direct causal or influential
relationships are encoded by the arcs. BNs follow an
independence assumption that every node in the graph is
strictly independent of any other variable except its de-
scendants. Theoretically, simple BNs are considered to
be ideal, such that they obtain the highest accuracy only
when this independence assumption holds. This inde-
pendence assumption along with the graphic structure
representing unambiguous interdependent relationships
allows for an important feature of BNs: representation of
joint probability distributions.
Bayesian networks can also be dynamic, which means
that the BN determines the activity being performed
based on variables from the sensory data [19]. Essen-
tially, dynamic BNs represent sequences of variables (i.e.
nodes in the graph of a BN). In a dynamic BN, these
sequences are activities that are being performed, and
these activities are determined by certain variables. The
activities are viewed as hidden variables, and the ob-
served variables include the set of objects seen and the
time elapsed. Using the observed variables (i.e., time
elapsed), it becomes possible to probabilistically estimate
the most likely activities and their intensity from sensor
data by using Bayes filtering. A sequential Monte Carlo
approximation is also used to solve for the most likely
activities based on sensor data.
Along with recognizing what activity the person is
performing, it is also important to figure out the person’s
emotional state [20]. However, deciphering a person’s
emotional state through sensory data is extremely diffi-
cult, as it is not always the case that the human perfectly
knows someone else’s emotion. Nevertheless, research-
ing general emotions that accompany certain activities
helps to figure out someone’s current state. It should be
noted that this may not always be accurate. Using sen-
sory data, the BSN can detect emotions for every in-
stance. By recognizing human emotion in these instances,
the BSN can help to assess stress, anger, and other emo-
tions that pertain to the individual’s health. Furthermore,
relating stress to the use of a product may give signifi-
cant information to help developers redesign and im-
prove their products to better fit humans needs.
Another technique that can be used to determine hu-
man activities is a decision tree [21]. With this method,
there are several possible activities that can occur in one
given place. The activity is then determined by sensing
what the human is doing in the current environment. Es-
sentially, there is a preset number of possible activities
given in a certain situation, but sensory data determines
which one of these activities is actually taking place.
The mixture model method brings another approach to
determining human activities [22]. Mixture models clus-
ter observations into event types, and activities are con-
sidered human behaviors. After mixture models organize
observations into clusters, a density function, which can
be found in [9], is applied to check the significance of
the clusters. Density represents a calculation of the con-
sistency of a given activity. The higher the density, the
more significant a cluster is. Therefore, activities with
the highest density calculations are the most commonly
occurring activities, and those with low density calcula-
tions are classified as random events. Based on this in-
formation, it is easier for the BSN to determine what
event is most likely occurring.
The following sections describe naïve Bayes Classifi-
ers, and BNs with hidden nodes. Table 3 discusses the
advantages and limitations of BNs for context-aware
sensing, and Table 4 summarizes various applications of
BNs in context-aware sensing.
4.2.1. Naïve Bayes Classifiers
A naïve Bayes classifier is a probabilistic classifier ad-
hering to Bayes’ rule and in context awareness is used
for classification. More specifically, activity recognition
may be reduced to a classification problem where classes
correspond to activities and Bayes classifiers predict the
activity labels after training examples are generated.

B. T. KOREL ET AL. 577
Table 3. Benefits and drawbacks of the Bayesian Network approach to context awareness.
Technique Advantages Limitations
Naїve Bayes Classifiers
1) Helpful in predicting consequences of
interaction within the domain, since BNs
provide the ability to gain an understanding
of the problem domain at hand by the exis-
tent relationships.
2) Computationally efficient, which means
viable for real-time recognition.
3) Noise resilient. It does not model am-
biguous or noisy information from multiple
sensors; handles classifying context where
missing or incomplete data exists.
1) If there is confusion in labeled training
data between similar activities, or if there is
an extremely low number of training data,
activity recognition suffers greatly.
2) Requires labeled training data, thus
needs user participation.
3) Large scope of real-world domains may
cause naїve Bayes to perform poorly since
not all objects and relations may be repre-
sented.
4) Problem of genericity in recognition due
to ambiguity of some features referring to
multiple context.
5) Enforces mutual exclusivity so when
two different activities occur simultane-
ously, detection of one could preclude
detection of the other.
Feature Selection and Smoothness
Constraint
1) Additional noise resilience and detec-
tion, which in turn increases model accu-
racy.
2) Allows for feature selection.
3) Neutralizes redundancy in the network.
4) Smoothness constraint introduced to
model continuous human movement.
5) Decreases power consumption of net-
work due to feature selection.
6) Allows for distributed computing across
network to local subnets which reduces
stress on central processor in terms of
bandwidth and computational load.
1) The distinction between very similar
context may be blurred or lost.
2) Requires user participation.
Table 4. Examples of some BN techniques in context-aware sensing.
Application Experiment Results
Madabhushi
et al. (1999)
[23]
Incorporates a Bayesian frame-
work to human activity recogni-
tion in order to automatically
identify human action.
Classifies ten different human actions
from visual information by tracking the
position of the head in pictures.
The system had a success rate of 80%
for recognizing activity. The system
has limitations such as only recogniz-
ing one action in a sequence and not
performing in real time.
Korpipaa et al.
(2003) [24]
Applies naїve Bayesian net-
works to classify the context of a
mobile device user in her normal
daily activities.
The classification was based mainly on
audio features measured in a home sce-
nario.
Situations can be extracted fairly well,
but most of the context is likely to be
valid only in a restricted scenario;
naїve Bayes framework is feasible for
context recognition.
Tapia et al.
(2004) [25]
Proposes a system based on a
naїve Bayesian framework for re-
cognizing activities in the home
setting.
Uses a set of small and simple state-
change sensors that can be quickly and
ubiquitously installed in the home envi-
ronment to recognize activities.
Results from a small dataset show that
it is possible to recognize activities
such as toileting, bathing and groom-
ing with detection accuracies ranging
from 25% to 89%.
Elnahrawy
et al. (2004)
[26]
Proposes a technique based on
Bayesian classifiers for model-
ing and learning statistical con-
textual information in sensor
networks.
Analyzes the approach in two applica-
tions, tracking and monitoring. Intro-
duces applications of the model in out-
lier detection, approximation of missing
values and sampling.
Once the contextual information is
learned, these applications reduce to an
inference problem. Evaluations show
the applicability and a good perform-
ance of the approach.
Thus naïve Bayes classifiers require labeled training data
to recognize clearly defined activities, which has the
downside of requiring more effort in the data recording
phase [18]. However, with the given training sample,
naïve Bayes classifiers are able to optimally predict a
class of examples that have not been previous seen by
the system [27]. Generally speaking the theory of prob-
ability provides a solid ground for the task of classifica-
tion [24]; since the naïve Bayes classifier is a probabilis-
tic induction algorithm, this is an approach to classifica-

B. T. KOREL ET AL. 578
tion that performs with high accuracy and an attainable
recognition rate for activity recognition in specific do-
mains.
As mentioned previously the Bayesian classifier is
thought to perform optimally when it adheres to the in-
dependence assumption in which there are no dependen-
cies between attributes. However, in many domains in-
cluding the real-world, this assumption does not and
cannot hold true; thus it seems that either this assumption
is undermined or the accuracy of the Bayesian classifier
is not in fact optimal. However in [27], Domingos and
Pazzani showed that the Bayesian classifier is optimal
despite strong attribute dependencies existing in the sys-
tem. This is partially due to Bayes classifiers not de-
pending on the original independence assumption to
perform at their best. Another explanation could be that
the high bias that exists due to the strong independence
assumption could be neutralized by the low variance of
the classifier [25].
Limitations: Activity recognition accuracy suffers
based on naïve Bayes when either there is confusion in
the labeled training examples between very similar ac-
tivities or there is an extremely low number of training
examples for an activity. In many applications which
have implemented naïve Bayes for context awareness in
wireless sensor networks, testing was performed in a
restricted scenario; had testing been done in a real-world
scenario the accuracy rates would most likely have been
lower. This indicates that naïve Bayes classifiers may
perform poorly in real domains. One reason could be that
in the real-world, naïve Bayes classifiers hold the as-
sumption that all attributes that influence a classification
decision are observable and represented [25], which may
be the case for a limited test scenario but not in the scope
of the real-world. In the real-world domain there are
many different kinds of objects, classes of objects and
numerous relations among them. BNs are still limited in
not being able to exhaustively represent all the objects
and relations that exist in the real-world [28]. Problems
of genericity in recognition will also arise for naïve
Bayes classifiers due to the ambiguity of some features
referring to multiple contexts [24]. An additional disad-
vantage of the naïve classifier is it enforces mutual ex-
clusivity, thus when two different activities occur simul-
taneously, detection of one could preclude the detection
of the other [25].
Advantages: Key advantages of naïve Bayes classifi-
ers include providing a noise resilient classification
framework by not modeling ambiguous or noisy infor-
mation from multiple sensors in the network. In addition
they appropriately handle classifying situations where
missing or incomplete data exists. BNs have also been
shown to be computationally efficient, making them vi-
able for real-time recognition of context (an important
criterion for body sensor networks) [24]. Lastly naïve
Bayesian networks provide the ability to gain an overall
understanding of the problem domain at hand from
learning the causal relationships that exist. This can then
be used for predicting the consequences of interaction
within the domain [29].
4.2.2. Hidden Nodes
In a BN, there is a considerable amount of dependency
between parent and child nodes which violates the inde-
pendence assumption that BNs are based on. This de-
pendency may increase when redundant nodes are added
to the network to overcome motion artifact and sensor
failure [6]. Hidden nodes may be added to the BN as
unobserved variables to represent the dependencies
among children. They effectively compensate the extra
weight dependant children add to the network, neutraliz-
ing redundancy in the network. Figure 4 shows the
structure of a BN with hidden nodes.
Advantages of Hidden Nodes: By inserting hidden
nodes to the BNs, subnets are formed (representing the
redundant nodes) to increase robustness and benefit the
network in many ways. First the subnets provide addi-
tional noise resilience by filtering out noise present
within each subnet, in turn increasing model accuracy [8].
Additionally, hidden nodes are able to detect node fail-
ures by identifying an asynchronous child-parent de-
pendency, thus providing noise detection [6]. The capa-
bility of BNs with hidden nodes to detect noise and ac-
curately classify context despite sensor noise is crucial in
BSNs where detrimental actions are taken based on the
context recognized. The smoothness constraint necessary
in modeling continuous human movement is possible
using BNs with hidden nodes. This can be achieved by
adding a fixed size temporal window over instantaneous
model beliefs of the network [6]. Inserting hidden nodes
to indicate redundant sensors is a form of feature selec-
tion in which irrelevant sensor data is not sent across the
network, which has the advantage of decreasing net
power consumption. Finally, stress on the central proc-
essor of the network can be diminished in terms of
bandwidth and computational load by distributing com-
putations across the network to the local subnets [8].
D
HH
S5 S6S4S3S2S1
D‐Decis i onNode
H‐HiddenNode
S‐SensorNode
Hiddennodesindicatehiddensensors
Figure 4. Bayesian network with hidden nodes.

B. T. KOREL ET AL.579
4.3. Hidden Markov Models
For context sensing in a body sensor system to be appli-
cable in the real-world, context needs to be continually
recognized throughout a duration of time and not just at
exact instances in time (as the case of BNs). Hidden
Markov Models (HMMs) are introduced at the supervis-
ing layer to achieve a model of context transition. The
use of HMMs is an approach to context recognition that
can more accurately model human behavior since the
system is capable of recognizing sequences of activities.
A feature selection technique that can be used for con-
text-aware sensing is the HMM. HMMs transform a set of
sequences of different lengths to a feature space, where
clustering can then be performed [30]. In this feature
space, each sequence is represented as a vector of dis-
tances. Using HMMs, each sensor output is represented
by its distance to HMMs trained for the whole set of sen-
sor outputs instead of comparing sensor outputs directly
as time series of different lengths. This distance refers to
how likely a certain event is to be predicted by that HMM
and is described by the log likelihood. If a sensor’s output
is highly correlated to the trained HMM, the values of the
associated log likelihood would also be high.
Layered HMMs (LHMM) are separate layers of
HMMs connected according to their inferential results
[31]. In this way, each level of the LHMM hierarchy can
be trained in a different way. Instead of having a whole
HMM responsible for all actions, LHMM allow for dif-
ferent layers to be responsible for detecting certain ac-
tions. There are two ways to do inference with LHMMs.
One way is to select the model in which the highest like-
lihood is selected; this is called maxbelief. This informa-
tion is made available as input to the HMM at the next
level of the hierarchy. In the distributional approach, the
full probability distribution over the models is passed to
the higher-level HMMs.
Algorithm: HMMs are probabilistic models used to
represent non-deterministic processes and consist of
states, actions and observations. The outcome or obser-
vation of a state is determined by the conditional prob-
ability distribution of the state and is based on the
Markov property, where the current state of the envi-
ronment depends solely on the intermediate previous
state and the associated action. In HMMs the state se-
quence is hidden and only the observations are visible.
The transition from one state to another is an action, la-
beled with the probability that the transition from one
context to another will occur. In context awareness sens-
ing, the finite set of states represents user-defined context
profiles and the HMM models the transition processes
through the states, or rather the behavior of a user transi-
tioning from one context to another.
Applications that have used HMMs in learning, clas-
sifying and modeling the dynamics of different situations
include tracking the daily activity of residents in an as-
sisted living community [6]. Using wearable sensors to
generate personal contextual annotations from an au-
dio-visual recording of a meeting [32]. Inferring envi-
ronmental context by recognizing classes of sound [33].
Recognizing a person’s situation from a wearable au-
dio-visual system [34]. Identifying bathroom activities
based on sound event classification [35]. And finally an
application that automatically tracks the progress of
maintenance or assembly tasks using body worn sensors
[36].
Just as BNs, HMMs also require a training phase in
order to classify activities. Optimally this learning phase
should occur without intervention from the user, espe-
cially to be applicable in Body Sensor Networks. Clark-
son and Pentland in [33] propose a method using incre-
mental additive learning to avoid relying on the user for
learning. As such, if the learned system is given an event
that cannot be recognized accurately by the current
HMM, the system can recognize this by the indication of
low scores and then generates a new model for the event.
When modeling sequences of events, it is advanta-
geous that HMMs allow for time variance (the event may
be performed at varying speeds), repetition (an event
may be repeated any number of times) [36] and they are
able to deal with variable length sequences. In addition,
due to their probabilistic framework they are able to take
into account noisy sensors and imperfect training data
coming from different sources of uncertainty. Overall,
HMMs build a statistical memory of sequences of events
that prove to be relatively robust with regards to tempo-
ral changes and allow for high-level domain knowledge
to be incorporated into the model [3].
Limitations: One disadvantage of HMMs includes not
being able to extract a valuable model of event sequences
if activity transitions in the data set do not occur often
enough [18]. In addition, if the trained model is too gen-
eral then it will classify a larger set of events with a high
probability, thus not being able to distinguish between
specific events that may be somewhat similar [33].
HMMs may be rather limited to represent models in the
real-world domain because their notion of state lacks the
structure that exists in the real-world [24]. Lastly, HMMs
can be computationally expensive in regards to perform-
ance because the HMM requires enumerating all possible
paths through the model [3].
Next, we present a hierarchical structure of HMMs.
Table 5 describes the advantages and limitations of the
HMM approach, and Table 6 shows examples of HMMs
used in sensor networks.
4.3.1. Hierarchical Hidden Semi-Markov Models
In context awareness sensing there is a natural hierarchy
of human activities, especially as the activities become
more and more complex. So it is beneficial to mode vari-
ous sequences of events using HMMs at higher levels

B. T. KOREL ET AL. 580
Table 5. Benefits and drawbacks of the Hidden Markov Model approach to context awareness.
Technique Advantages Limitations
Hidden
Markov Model
1) Smoothness constraint: allow for continual
recognition of context or sequences of activities.
2) Allows for time variance (an action may be
performed at varying speeds) and repetition (an action
may be repeated any number of times).
3) Handles noisy sensors and imperfect training data
due to their probabilistic framework.
1) Requires a training phase to classify activities, so needs
user participation.
2) If activity transitions in the dataset don't occur often
enough, HMMs might not extract a valuable model of event
sequences.
3) If trained model is too general, can't distinguish between
very similar events.
4) Limited representation of models in the real-world
domain.
5) Computationally expensive in regards to performance.
Hierarchical
Hidden
Semi-Markov
Model
1) Reasons about relative order, duration and ab-
straction of general activity from sequences of sub-
activities.
2) Allows for sequences of sub-activities to be re-
cognized without the system having to learn a se-
parate HMM for each sequence.
3) Improves computational performance by not
processing the training of useless sub-activities.
1) May not scale to environments that contain hundreds of
sensors.
2) Requires user participation.
Table 6. Examples of HMM techniques in sensor networks.
Application Experiment Results
Kern et al.
(2002) [32]
Uses wearable computers and sensor
systems to generate personal contex-
tual annotations in audio visual reco-
rdings of meetings.
HMMs used to segment an audio stream
into sequences assigned to different
speakers.
Use of only HMMs results in a re-
cognition error of 18%. Study shows
that many issues in recognition of
complex postures and gestures remain.
Clarkson
et al. (1998)
[33]
Uses HMMs as a statistical/pattern
recognition framework to recognize
environmental context by audio
classification.
The HMM classification framework
is integrated with Nomadic Radio, a
wearable messaging system, to create
a wearable platform that is aware of
its audio environment.
System proved to be robust towards
noise and reliably detected speech by
and around the user.
Clarkson
et al. (2000)
[34]
Presents modeling with HMMs to
recognize a person’s situation; situa-
tions are coarse locations (at work,
in a grocery store) and coarse events
(having a conversation, walking on
the street).
A wearable camera and microphone
are used to model 12 dimensions of
video features and 12 dimensions of
audio features, in order to recognize a
person’s context.
Results show a surprisingly high ac-
curacy considering the lack of context
and the coarse features; however can’t
assume this method will work for a wide
variety of conditions and situations.
Chen et al.
(2005) [35]
Uses HMM parameters for accurate
and robust recognition and classifi-
cation of major activities occurring
within a bathroom based on sound.
Experiments first performed in a con-
strained setting and then in an actual
trial involving real people using their
bathroom in the normal course of
their daily lives.
Preliminary results show an accuracy
rate of above 84% for most sound
categories.
Lukowicz
et al. (2004)
[36]
Automatically tracks the progress of
maintenance or assembly tasks using
body worn sensors; uses body worn
microphones and accelerometers.
Technique is applied to activities in a
wood shop.
On a simulated assembly task, the
system can successfully segment and
identify most shop activities in a conti-
nuous data stream with 84.4% accuracy.
Kautz et al.
(2003) [28]
Proposes hierarchical hidden semi-
Markov models for tracking the daily
activities of residents in an assisted
living community.
Implements HHSMM to link location
and movement information to a sub-
ject’s behavior, and reasons about the
hierarchical relationship between ab-
stract actions and sub-actions, and both
qualitative and quantitative metric con-
straints.
The semi-Markov structure allows the
network to distinguish different activities
solely based on their duration. Shows
that better algorithms and representations
are needed to scale up to larger, more
detailed models and fine-grained low-
level input data.
Chambers
et al. (2002)
[37]
Uses an extension of HMM for
hierarchical recognition of complex
human gestures for sports video
annotation.
Uses a hierarchy of HMMs and acce-
lerometers to extract complex gesture
recognition; dataset consists of several
Kung Fu martial art movements acted
out by an instructor in a simulated
training video.
The system can robustly differentiate
between gestures and accurately segment
a gesture into its compromising sub-
gestures.

B. T. KOREL ET AL. 581
of abstraction through a hierarchy. For example if a per-
son is “having a meal,” this activity consists of a variety
of sub-activities such as preparing the food, setting the
table, eating the meal, etc. Reasoning about the relative
order and duration of the sub-activities, and abstracting
the general activity of “having a meal” from the sub-
activities can be accomplished with a hierarchy of
HMMs. Individual movements would be represented by
a number of standard HMMs at the lowest level of the
hierarchy. Higher layers represent the combination of
movements, and at even higher layers sequences of com-
bination movements can be recognized [37]. As such, the
system as a whole is able to model the various levels of
complexity of a general activity. The HMM at each level
only interacts with the levels directly above and below it;
in other words a node at one level may have a transition
to a node in the HMM at the direct higher or lower level
[28]. Some applications of hierarchical HMMs include
recognizing complex human gestures for video annota-
tion (see Figure 5) [37] and tracking daily activities [28].
Advantages of Hierarchical HMMs: An advantage
of the multi-layer HMM is that it allows a rich sequence
of sub-activities to be recognized without the system
having to learn a separate HMM for each sequence;
whereas the standard HMM would require training of a
new model with each new combination sequence [37].
This is especially beneficial because the order of sub-
activities performed to compose the ultimate activity will
differ from person to person, and will differ even when a
single person performs the same activity a number of
times. Yet the system is still able to recognize the highest
level activity. In addition, the activity may be performed
with some sub-activities occurring that are not related at
all to the general activity. Thus if hierarchical HMMs
don't require each new sequence to be learned, the system
avoids training useless sub-activities, hence improves the
computational performance of the system as a whole.
Layered HMMs (LHMM) are separate layers of HMMs
Ges ture
Sub‐ges tu reA Sub‐gestu reC Sub‐ges tu reB
HMM forsub‐gestureAHMMforsub‐gestureB
Figure 5. Hierarchical hidden markov model for complex
gesture recognition.
connected according to their inferential results. In this
way, each level of the LHMM hierarchy can be trained in
a different way. Instead of having a whole HMM respon-
sible for all actions, LHMMs allow for different layers to
be responsible for detecting certain actions. There are
two ways to do inference with LHMMs. One way is to
select the model in which the highest likelihood is se-
lected. This information is made available as input to the
HMM at the next level in the hierarchy. In the distribu-
tional approach, the full probability distribution over the
models is passed to the higher-level HMMs. It may be
beneficial for a system to recognize different activities
solely based on metric time or the duration the system
spends in each sub-step of an activity. This can be
achieved by associating a probability distribution over
the time the system remains in one state before it transi-
tions to another state [28]. Thus metric, non-exponential
time is added to the system resulting in the hierarchical
hidden semi-Markov Model.
5. Conclusions
Context awareness in Body Sensor Networks allows for
the current state of the user and environment to reason
about physical and biochemical sensor signals. Con-
text-aware sensing is an integral part of the BSN design
in order to allow for long-term pervasive health care
monitoring of patients. The outcome of such monitoring
would prevent mortality on the grounds that BSNs could
have detected precursors of the death.
In this paper, we presented context-aware approaches
including Artificial Neural Network (ANN), Bayesian
Network (BN) and Hidden Markov Model (HMM). No
technique is the “best” for deducing context awareness
and each method addresses different issues that arise
from context-aware sensing in Body Sensor Networks.
6. References
[1] A. K. Dey, D. Salber, G. D. Abowd and M. Futakawa,
“The Conference Assistant: Combining Context-Awareness
With Wearable Computing,” Proceedings of the 3rd In-
ternational Symposium on Wearable Computers, San
Francisco, 1999, pp. 21-28.
[2] A. Krause, D. Siewiorek, A. Smailagic and J. Farringdon,
“Unsupervised, Dynamic Identification of Physiological
and Activity Context in Wearable Computing,” Proceed-
ings of the 7th International Symposium on Wearable
Computers, White Plains, 2003, pp. 88-97.
[3] G.-Z. Yang, “Body Sensor Networks,” Springer-
Verlag, London, 2006.
[4] B. Fritzke, “Some Competitive Learning Methods, Artifi-
cial Intelligence Institute,” Dresden University of Tech-
nology, Dresden, 1997.
[5] E. Elnahrawy and B. Nath, “Poster Abstract: Online Data

B. T. KOREL ET AL. 582
Cleaning in Wireless Sensor Networks,” Proceedings of
the 1st International conference on Embedded networked
sensor systems, Los Angeles, 2003, pp. 294-295.
[6] S. Thiemjarus, B. P. L. Lo and G.-Z. Yang, “A Distrib-
uted Bayesian Framework for Body Sensor Networks,”
IEE Proceedings of the 2nd International Workshop on
Body Sensor Networks, 2005, pp. 24-26.
[7] K. van Laerhoven, K. A. Aidoo and S. Lowette, “Real-
time Analysis of Data from Many Sensors with Neural
Networks,” Proceedings of the 5th International Sympo-
sium on Wearable Computers, Los Angeles, 2001, pp.
115-123.
[8] S. Thiemjarus, R. King, B. Lo, D. Gilles and G.-Z. Yang,
“A Noise Resilient Distributed Inference Framework for
Body Sensor Networks,” Proceedings of the 3rd Interna-
tional Conference on Pervasive Computing, Munich,
2005, pp. 13-18.
[9] P. Nurmi, P. Floreen, M. Przybilski and G. Linden, “A
Framework for Distributed Activity Recognition in Ubiq-
uitous Systems,” Proceedings International Conference
on Artificial Intelligence (ICAI), Las Vegas, 2005, pp.
650-655.
[10] S. Thiemjarus, B. P. L. Lo and G.-Z. Yang, “Context Aware
Sensing －What is the Significance?” IEE Proceedings of
the Perspective in Pervasive Computing, London, 2005, pp.
163-170.
[11] K. van Laerhoven, “Combining the Self-Organizing Map
and K-Means Clustering for On-Line Classification of
Sensor Data,” Artificial Neural Networks － ICANN, Vi-
enna, 2001, pp. 464-470.
[12] A. Schmidt, K. A. Aidoo, A. Takaluoma, U. Tuomela, K.
van Laerhoven and W. van de Velde, “Advanced Interac-
tion in Context,” Proceedings of the 1st International
Symposium on Handheld and Ubiquitous Computing,
Karlsruhe, 1999, pp. 89-101.
[13] M. Lagerholm, C. Peterson, G. Braccini, L. Edenbrandt
and L. Sornmo, “Clustering ECG Complexes Using Her-
mite Functions and Self-Organizing Maps,” IEEE Trans-
actions on Biomedical Engineering, Vol. 47, No. 7, July
2000, pp. 838-848.
[14] K. van Laerhoven and O. Cakmake, “What Shall We
Teach Our Pants?” Proceedings of the 4th International
Symposium on Wearable Computers, Atlanta, 2000, pp.
77-83.
[15] K. Simelius, M. Stenroos, L. Reinhardt, J. Nenonen, I.
Tierala, M. Makijarvi, L. Toivonen and T. Katila, “Spati-
otemporal Characterization of Paced Cardiac Activation
with Body Surface Potential Mapping and Self-Organizing
Maps,” Physiological Measurements, Vol. 24, No. 3,
2003, pp. 805-816.
[16] D. Gao, M. Madden, M. Schukat, D. Chambers and G.
Lyons, “Arrhythmia Identification from ECG Signals
with a Neural Network Classifier Based on a Bayesian
Framework,” Proceedings of the 24th SGAI International
Conference on Innovative Techniques and Applications of
Artificial Intelligence, Cambridge, 2004.
[17] S. Thiemjarus, B. P. L. Lo and G.-Z. Yang, “A Spatio-
Temporal Architecture for Context-Aware Sensing,”
Proceedings of the IEEE International Workshop on
Wearable and Implantable Body Sensor Networks, Cam-
bridge, 2006, pp. 191-194.
[18] K. van Laerhoven, N. Kern, H. W. Gellersen and B.
Schiele, “Towards a Wearable Inertial Sensor Network,”
Proceedings of IEE Eurowearable, Stevenage, 2003, pp.
125-130.
[19] M. Philipose, K. Fishkin, M. Perkowitz, D. Fox, H. Kautz
and D. Patterson, “Inferring Activities from Interactions
with Objects,” IEEE Pervasive Computing, Vienna, 2004,
pp. 50- 57.
[20] R. Picard, E. Vyzas and J. Healey, “Towards Machine
Emotional Intelligence: Analysis of Affective Physio-
logical State,” IEEE Transactions on Pattern Analysis
and Machine Intelligence, 2001, pp. 1175-1191.
[21] K. Noguchi, P. Somwong, T. Matsubara and Y. Nakauchi,
“Human Intention Detection and Activity Support System
for Ubiquitous Autonomy,” Proceedings of IEEE Inter-
national Symposium on Computational Intelligence in
Robotics and Automation, 2003, pp. 906-911.
[22] T. Barger, D. Brown and M. Alwan, “Health-Status
Monitoring through Analysis of Behavioral Patterns,”
IEEE Transactions on Systems, Man and Cybernetics:
Part A, Edmonton, 2005, pp. 22-27.
[23] A. Madabhushi and J. K. Aggarwal, “A Bayesian Ap-
proach to Human Activity Recognition,” Proceedings of
the 2nd IEEE International Workshop on Visual Surveil-
lance, Fort Collins, 1999, pp. 25-32.
[24] P. Korpipaa, M. Koskinen, J. Peltola, S. M. Makela and T.
Seppanen, “Bayesian Approach to Sensor-Based Context
Awareness,” Personal and Ubiquitous Computing, Vol. 7,
No. 2, 2003, pp. 113-124.
[25] E. M. Tapia, S. S. Intille and K. Larson, “Activity Rec-
ognition in the Home Using Simple and Ubiquitous Sen-
sors,” Proceedings on the 2nd International Conference
on Pervasive Computing, Vienna, 2004, pp. 158-175.
[26] E. Elnahrawy and B. Nath, “Context-Aware Sensors,”
Proceedings on 1st European Workshop on Wireless
Sensor Networks, Berlin, 2004, pp. 73-93.
[27] P. Domingos and M. Pazzani, “Beyond Independence:
Conditions for the Optimality of the Simple Bayesian
Classifier,” Proceedings of the 13th International Con-
ference on Machine Learning, Bari, 1996, pp. 105-112.
[28] H. Kautz, O. Etzioni, D. Fox and D. Weld, “Foundations
of Assisted Cognition Systems,” Technical report CSE-
02-AC-01, University of Washington, Department of
Computer Science and Engineering, Washington, D.C.,
2003.
[29] M. Angermann, P. Robertson and T. Strang, “Issues and
Requirements for Bayesian Approaches in Context Aware
Systems,” Location- and Context Awareness: First In-
ternational Workshop, Vol. 3479, 2005, pp. 235-243.
[30] R. King, L. Atallah, A. Darzi and G.-Z. Yang, “A HMM
Framework for Optimal Sensor Selection with Applica-
tions to BSN Sensor Glove Design,” Proceedings of the
Fourth Workshop on Embedded Networked Sensors, Cork,

B. T. KOREL ET AL. 583
2007, pp. 58-62.
[31] N. Oliver, E. Horvitz and A. Garg, “Layered Representa-
tions for Human Activity Recognition,” Proceedings of
the 4th IEEE International Conference on Multimodal
Interfaces, Pittsburgh, 2002, pp. 3-8.
[32] N. Kern, B. Schiele, H. Junker and G. Troster, “Wearable
Sensing to Annotate Meeting Recordings,” Proceedings
on the 6th International Symposium on Wearable Com-
puters, Seattle, 2002, pp. 186-194.
[33] B. Clarkson and A. Pentland, “Extracting Context from
Environmental Audio,” Proceedings of the 2nd Interna-
tional Symposium on Wearable Computers, Pittsburgh,
1998, pp. 154-155.
[34] B. Clarkson, K. Mase and A. Pentland, “Recognizing
User Context via Wearable Sensors,” Proceedings of the
4th International Symposium on Wearable Computers,
Atlanta, 2000, pp. 69-76.
[35] J. Chen, A. H. Kam, J. Zhang, N. Liu and L. Shue, “Bath-
room Activity Monitoring Based on Sound,” Proceedings
of the 3rd International Conference on Pervasive Com-
puting, Munich, 2005, pp. 47-61.
[36] P. Lukowicz, J. A.Ward, H. Junker, M. Stager, G. Troster,
A. Atrash and T. Starner, “Recognizing Workshop Activ-
ity Using Body Worn Microphones and Accelerometers,”
Pervasive Computing, Vol. 3001, 2004, pp. 18-32.
[37] G. S. Chambers, S. Venkatesh, G. A. W. West and H. H.
Bui, “Hierarchical Recognition of Intentional Human
Gestures for Sports Video Annotation,” Proceedings of
the 16th IEEE Conference on Pattern Recognition, Qué-
bec City, Vol. 2, 2002, pp. 1082-1085.
[38] B. T. Korel and S. G. M. Koo, “Addressing Context
Awareness Techniques in Body Sensor Networks,” Pro-
ceedings of the 21st IEEE International Conference on
Advanced Information Networking and Applications
(AINA 2007), Niagara Falls, Vol. 2, 2007, pp. 798-803.
[39] G. D. Abowd and A. K. Dey, “Towards a Better Under-
standing of Context and Context-Awareness,” Georgia
Institute of Technology, Technical Report GIT-GVU-99-
22, Atlanta, 1999.
[40] G. D. Abowd, A. K. Dey, R. Orr and J. Brotherton, “Con-
text-Awareness in Wearable and Ubiquitous Computing,”
Proceedings of the 1st International Symposium on
Wearable Computers, Cambridge, 1997, pp. 179-180.
[41] U. Anliker, H. Junker, P. Lukowicz and G. Troster, “De-
sign Methodology for Context-Aware Wearable Sensor
System,” Pervasive Computing: 3rd International Con-
ference, Seattle, 2005, pp. 220-236.
[42] A. R. Golding and N. Lesh, “Indoor Navigation Using a
Diverse Set of Cheap, Wearable Sensors,” Proceedings of
the 3rd International Symposium on Wearable Computers,
San Francisco, 1999, pp. 29-36.
[43] J. Himberg, J. A. Flanagan and J. Mantyjarvi, “Towards
Context Awareness Using Symbol Clustering Map,”
Proceedings of Workshop for Self-Organizing Maps, Ki-
takyushu, 2003, pp. 249-254.
[44] N. Kern and B. Schiele, “Context-Aware Notification for
Wearable Computing,” Proceedings of the 7th Interna-
tional Symposium on Wearable Computers, Osaka, 2003,
pp. 223-230.
[45] P. Langley, W. Iba and K. Thompson, “An Analysis of
Bayesian Classifiers,” Proceedings of the 10th National
Conference on Artificial Intelligence, San Jose, 1992, pp.
223-228.
[46] B. P. L. Lo, S. Thiemjarus, R. King and G.-Z. Yang,
“Body Sensor Network－A Wireless Sensor Platform for
Pervasive Healthcare Monitoring,” Proceedings of the
3rd International Conference on Pervasive Computing,
Munich, 2005, pp. 77-80.
[47] B. P. L. Lo and G.-Z. Yang, “Key Technical Challenges
and Current Implementations of Body Sensor Networks,”
IEE Proceedings of the 2nd International Workshop on
Body Sensor Networks, London, 2005, pp. 1-5.
[48] D. Macq, M. Verleysen, P. Jespers and J. D. Legat, “Ana-
log Implementation of a Kohonen Map with On-Chip
Learning,” IEEE Transactions on Neural Networks, 1993,
pp. 456-461.
[49] D. Patterson, D. Fox, H. Kautz and M. Philipose, “Ex-
pressive, Tractable and Scalable Techniques for Model-
ing Activities of Daily Living,” Proceedings of the 2nd
International Workshop on Ubiquitous Computing for
Pervasive Healthcare Applications, 2003.
[50] N. Ravi, N. Dandekar, P. Mysore and M. L. Littman,
“Activity Recognition from Accelerometer Data,” Pro-
ceedings of IAAI, Pittsburgh, 2005.
[51] K. van Laerhoven, “Teaching Contexts to Applications,”
Proceedings of the Workshop on Situated Interaction in
Ubiquitous Computing, 2000, pp. 5-8.
[52] K. van Laerhoven, B. P. L. Lo, J. W. P. Ng, S. Thiemja-
rus, R. King, S. Kwan, H. W. Gellersen, M. Sloman, O.
Wells, P. Needham, N. Peters, A. Darzi, C. Toumazou
and G.-Z. Yang, “Medical Healthcare Monitoring with
Wearable and Implantable Sensors,” International Work-
shop on Ubiquitous Computing for Pervasive Healthcare
Applications, 2004.
[53] K. van Laerhoven, A. Schmidt and H. W. Gellersen,
“Multi-Sensor Context Aware Clothing,” Proceedings of
the 6th International Symposium on Wearable Computers,
Seattle, 2002, pp. 49-57.