Systems, Man & Cybernetics - April 2016 - 18

accuracy of the inferred context increases with an
increase in the number and type of sensors utilized in
the inferencing process. As a simple example, a combination of data from a body-worn accelerometer and
ceiling-mounted motion sensors provides a more accurate estimation of whether a person is immobile after a
fall, compared to deductions based solely on one sensor
or the other.
2) The quality of the inferred context is not just a function of the set of chosen sensors but also of the permitted inaccuracy in the data values associated with each
individual sensor. For example, the quality of the estimation of the heart activity context will be less accurate if the blood pressure sensor's tolerance range is
! 20% (indicating that the true reading may be up to
20% higher or lower than the reported value) in comparison to a tolerance range of ! 5% .
Mathematically, we can then say that the quality of
inference function, denoted as QoINF, for any given context will be a function of i (the set of sensors used in the
context inferencing process) and the qi values (called the
tolerance range) associated with each sensor si. Conceptually, the job of our matchmaking algorithms is to find,
given a specific QoINF function, the set i and the associated qi values (for those selected sensors) that minimize
the communication energy overhead.
Quality of Inference for a Single Sensor
To simultaneously model the context accuracy and communication overheads associated with different values of
qi for a single sensor, we assume that each sensor utilizes
the widely adopted event-driven reporting strategy, where
it transmits a new sample only when its sensed value deviates from the previously transmitted sample by ! q i . In
effect, this means that, at any instant, the context inferencing process is not aware of the sensor's true current value
but knows that this value will be within ! q i of the last
value transmitted by the sensor. Of course, a larger tolerance range results in a reduction in a sensor's reporting
rate (frequency) and thus dramatically lowers its communication energy overheads [1], [2].
While many functional forms of the QoINF function are
possible, we initially advocate and explore an inverse
exponential functional model, where the accuracy of

Activity
State
0.9

0.98

Respiratory
Sensor

context inference or QoINF (for a specific application) for
a specific sensor si and its associated qi value are related
via the model
1
-1
qoinf (i) = 1 - o i exp a h i q i k,

where h i, o i are simply scalar constants. The choice of
this inverse-exponential model is both mathematically
motivated and empirically validated: not only does this
functional model make our eventual goal of multicontext
matchmaking tractable, it is also consistent with experimental results we have conducted using a variety of sensors (such as light, accelerometer, and motion sensors).
Context Inference with Multiple Sensors
and Applications
Having established the formal relationship between a single
sensor and a single context attribute, we now consider our
target scenario: multiple applications, each requiring different context inferences, potentially utilizing data from multiple available sensors. To precisely elucidate our approach,
we assume an underlying set S of sensors. Determining
the value of some context metric C may be viewed as a
multidimensional mapping that uses the values sensed by
some subset i of the available sensors (formally, i 3 S )
and maps them to one of the values associated with the
context metric. To understand this relationship better, consider the case illustrated in Figure 2, which depicts nine different sensors that may be used to support smart
health-care applications. An application that senses heart
activity may choose some subset of these sensors to assess
its context; for example, heart activity can be assessed by
the combination of a blood-pressure and a blood-flow sensor. A domain-specific inference function uses these two
low-level values and outputs a measure of heart activity.
To capture the reality that the same context may be
inferred to varying degrees of accuracy using different
sensor subsets, we associate a function that represents the
accuracy of a certain subset of sensors with respect to a
given context metric. That is, QoINFC (i) gives the expected accuracy of inferring a context metric C using the sensors in the subset i. Figure 2 (where we implicitly assume
that each of the sensors has a predefined tolerance range
of 0.10) further illustrates this notion of multiple sensors

Body
Movement
0.8
ECG

Heart
Activity
0.94

0.9
Accelerometer

0.8
EMG

0.7

0.6
Video
Camera

1.0
EEG

ECG

0.7
0.8
0.7
Blood
Blood
SpO2
Flow
Pressure

Figure 2. the impact of different sensor subsets on QoInF (without considering tolerance ranges).
18

IEEE SyStEmS, man, & CybErnEtICS magazInE A pri l 2016

(1)



Table of Contents for the Digital Edition of Systems, Man & Cybernetics - April 2016

Systems, Man & Cybernetics - April 2016 - Cover1
Systems, Man & Cybernetics - April 2016 - Cover2
Systems, Man & Cybernetics - April 2016 - 1
Systems, Man & Cybernetics - April 2016 - 2
Systems, Man & Cybernetics - April 2016 - 3
Systems, Man & Cybernetics - April 2016 - 4
Systems, Man & Cybernetics - April 2016 - 5
Systems, Man & Cybernetics - April 2016 - 6
Systems, Man & Cybernetics - April 2016 - 7
Systems, Man & Cybernetics - April 2016 - 8
Systems, Man & Cybernetics - April 2016 - 9
Systems, Man & Cybernetics - April 2016 - 10
Systems, Man & Cybernetics - April 2016 - 11
Systems, Man & Cybernetics - April 2016 - 12
Systems, Man & Cybernetics - April 2016 - 13
Systems, Man & Cybernetics - April 2016 - 14
Systems, Man & Cybernetics - April 2016 - 15
Systems, Man & Cybernetics - April 2016 - 16
Systems, Man & Cybernetics - April 2016 - 17
Systems, Man & Cybernetics - April 2016 - 18
Systems, Man & Cybernetics - April 2016 - 19
Systems, Man & Cybernetics - April 2016 - 20
Systems, Man & Cybernetics - April 2016 - 21
Systems, Man & Cybernetics - April 2016 - 22
Systems, Man & Cybernetics - April 2016 - 23
Systems, Man & Cybernetics - April 2016 - 24
Systems, Man & Cybernetics - April 2016 - 25
Systems, Man & Cybernetics - April 2016 - 26
Systems, Man & Cybernetics - April 2016 - 27
Systems, Man & Cybernetics - April 2016 - 28
Systems, Man & Cybernetics - April 2016 - 29
Systems, Man & Cybernetics - April 2016 - 30
Systems, Man & Cybernetics - April 2016 - 31
Systems, Man & Cybernetics - April 2016 - 32
Systems, Man & Cybernetics - April 2016 - 33
Systems, Man & Cybernetics - April 2016 - 34
Systems, Man & Cybernetics - April 2016 - 35
Systems, Man & Cybernetics - April 2016 - 36
Systems, Man & Cybernetics - April 2016 - 37
Systems, Man & Cybernetics - April 2016 - 38
Systems, Man & Cybernetics - April 2016 - 39
Systems, Man & Cybernetics - April 2016 - 40
Systems, Man & Cybernetics - April 2016 - 41
Systems, Man & Cybernetics - April 2016 - 42
Systems, Man & Cybernetics - April 2016 - 43
Systems, Man & Cybernetics - April 2016 - 44
Systems, Man & Cybernetics - April 2016 - 45
Systems, Man & Cybernetics - April 2016 - 46
Systems, Man & Cybernetics - April 2016 - 47
Systems, Man & Cybernetics - April 2016 - 48
Systems, Man & Cybernetics - April 2016 - 49
Systems, Man & Cybernetics - April 2016 - 50
Systems, Man & Cybernetics - April 2016 - 51
Systems, Man & Cybernetics - April 2016 - 52
Systems, Man & Cybernetics - April 2016 - 53
Systems, Man & Cybernetics - April 2016 - 54
Systems, Man & Cybernetics - April 2016 - 55
Systems, Man & Cybernetics - April 2016 - 56
Systems, Man & Cybernetics - April 2016 - Cover3
Systems, Man & Cybernetics - April 2016 - Cover4
https://www.nxtbook.com/nxtbooks/ieee/smc_202110
https://www.nxtbook.com/nxtbooks/ieee/smc_202107
https://www.nxtbook.com/nxtbooks/ieee/smc_202104
https://www.nxtbook.com/nxtbooks/ieee/smc_202101
https://www.nxtbook.com/nxtbooks/ieee/smc_202010
https://www.nxtbook.com/nxtbooks/ieee/smc_202007
https://www.nxtbook.com/nxtbooks/ieee/smc_202004
https://www.nxtbook.com/nxtbooks/ieee/smc_202001
https://www.nxtbook.com/nxtbooks/ieee/smc_201910
https://www.nxtbook.com/nxtbooks/ieee/smc_201907
https://www.nxtbook.com/nxtbooks/ieee/smc_201904
https://www.nxtbook.com/nxtbooks/ieee/smc_201901
https://www.nxtbook.com/nxtbooks/ieee/smc_201810
https://www.nxtbook.com/nxtbooks/ieee/smc_201807
https://www.nxtbook.com/nxtbooks/ieee/smc_201804
https://www.nxtbook.com/nxtbooks/ieee/smc_201801
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_1017
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0717
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0417
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0117
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_1016
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0716
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0416
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0116
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_1015
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0715
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0415
https://www.nxtbook.com/nxtbooks/ieee/systems_man_cybernetics_0115
https://www.nxtbookmedia.com