ETSI TR 103 395 V1.1.1 (2016-12)

Smart Body Area Networks (SmartBAN); Measurements and modelling of SmartBAN Radio Frequency (RF) environment

DTR/SmartBAN-006

General Information

• Status: Published
• Publication Date: 19-Dec-2016
• Technical Committee: SmartBAN - Smart Body Area Network
• Current Stage: 12 - Completion
• Due Date: 20-Dec-2016
• Completion Date: 20-Dec-2016

Standards Content (Sample)

TECHNICAL REPORT
Smart Body Area Network (SmartBAN);
Measurements and modelling of SmartBAN
Radio Frequency (RF) environment

2 ETSI TR 103 395 V1.1.1 (2016-12)

Reference
DTR/SmartBAN-006
Keywords
MAC, measurement, network
ETSI
650 Route des Lucioles
F-06921 Sophia Antipolis Cedex - FRANCE

Tel.: [phone]  Fax: [phone]

Siret N° 348 623 562 00017 - NAF 742 C
Association à but non lucratif enregistrée à la
Sous-Préfecture de Grasse (06) N° 7803/88

Important notice
The present document can be downloaded from:
http://www.etsi.org/standards-search
The present document may be made available in electronic versions and/or in print. The content of any electronic and/or
print versions of the present document shall not be modified without the prior written authorization of ETSI. In case of any
existing or perceived difference in contents between such versions and/or in print, the only prevailing document is the
print of the Portable Document Format (PDF) version kept on a specific network drive within ETSI Secretariat.
Users of the present document should be aware that the document may be subject to revision or change of status.
Information on the current status of this and other ETSI documents is available at
https://portal.etsi.org/TB/ETSIDeliverableStatus.aspx
If you find errors in the present document, please send your comment to one of the following services:
https://portal.etsi.org/People/CommiteeSupportStaff.aspx
Copyright Notification
No part may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying
and microfilm except as authorized by written permission of ETSI.
The content of the PDF version shall not be modified without the written authorization of ETSI.
The copyright and the foregoing restriction extend to reproduction in all media.

© European Telecommunications Standards Institute 2016.
All rights reserved.
TM TM TM
DECT , PLUGTESTS , UMTS and the ETSI logo are Trade Marks of ETSI registered for the benefit of its Members.
TM
3GPP and LTE™ are Trade Marks of ETSI registered for the benefit of its Members and
of the 3GPP Organizational Partners.
GSM® and the GSM logo are Trade Marks registered and owned by the GSM Association.
ETSI
3 ETSI TR 103 395 V1.1.1 (2016-12)
Contents
Intellectual Property Rights . 5
Foreword . 5
Modal verbs terminology . 5
1 Scope . 6
2 References . 6
2.1 Normative references . 6
2.2 Informative references . 6
3 Symbols and abbreviations . 7
3.1 Symbols . 7
3.2 Abbreviations . 8
4 Introduction and Background . 10
5 Coexistence . 10
5.0 Introduction . 10
5.1 Bands . 11
6 Measurements . 11
6.1 Background & Motivation . 11
6.2 Spectrum Occupancy Evaluations (SOEs) . 11
6.3 Measurement Campaigns . 13
6.3.0 Introduction. 13
6.3.1 Measurement campaigns in Oulu, Finland . 13
6.3.1.0 Introduction . 13
6.3.1.1 Daily Surgery SOEs (Campaign 1) . 14
6.3.1.2 Accident & Emergency Ward SOEs (Campaign 2) . 24
6.3.1.3 X-Ray & Radiology Ward SOEs (Campaign 3). 29
6.3.2 Analytical Stochastic Model for Spectrum Occupancy . 31
6.3.3 Extracting Mathematical Interference model . 34
6.3.4 Measurement Campaigns in Florence, Italy . 37
6.3.4.0 Introduction . 37
6.3.4.1 Occupancy . 38
6.3.4.1.0 Introduction . 38
6.3.4.1.1 Percentiles . 39
6.3.4.2 PDF . 40
6.3.4.3 Interference as a function of time and frequency . 41
6.3.4.4 Parameters characterizing the distribution . 42
6.3.4.5 Home and office environments . 42
6.3.4.6 Extract the mathematical model . 45
6.3.4.6.0 Introduction . 45
6.3.4.6.1 First results of CNIT-UNIFI . 45
6.4 Statistical model of the interference . 53
6.4.0 Introduction. 53
6.4.1 Cluster dimension . 54
6.4.2 Inter-arrival time . 55
6.4.3 Interfering cluster amplitude . 57
6.4.4 Conclusions. 60
6.5 Extracting the mathematical model of the interference . 60
7 SmartBAN communication system simulator . 69
7.0 Introduction . 69
7.1 Getting started . 69
7.2 Simulator model . 72
7.2.0 Introduction. 72
7.2.1 Node. 72
7.3 Hub . 73
ETSI
4 ETSI TR 103 395 V1.1.1 (2016-12)
7.3.0 Introduction. 73
7.3.1 Simulation parameters . 74
7.4 PHY layer . 75
7.4.0 Introduction. 75
7.4.1 PHY transmitter . 75
7.4.2 Channel, interference and noise . 75
7.4.2.0 Introduction . 75
7.4.2.1 Interference . 76
7.4.3 PHY receiver . 77
7.5 MAC - Frame retransmission . 77
7.6 Verification results . 79
8 Simulation results . 80
8.0 Introduction . 80
8.1 Simulation parameters . 80
8.2 AWGN channel . 81
8.3 Fading channel . 83
8.4 Fading channel and interference . 85
8.5 Discussion . 87
Annex A: Spatial Sample Clustering Algorithm . 88
History . 91

ETSI
5 ETSI TR 103 395 V1.1.1 (2016-12)
Intellectual Property Rights
IPRs essential or potentially essential to the present document may have been declared to ETSI. The information
pertaining to these essential IPRs, if any, is publicly available for ETSI members and non-members, and can be found
in ETSI SR 000 314: "Intellectual Property Rights (IPRs); Essential, or potentially Essential, IPRs notified to ETSI in
respect of ETSI standards", which is available from the ETSI Secretariat. Latest updates are available on the ETSI Web
server (https://ipr.etsi.org/).
Pursuant to the ETSI IPR Policy, no investigation, including IPR searches, has been carried out by ETSI. No guarantee
can be given as to the existence of other IPRs not referenced in ETSI SR 000 314 (or the updates on the ETSI Web
server) which are, or may be, or may become, essential to the present document.
Foreword
This Technical Report (TR) has been produced by ETSI Technical Committee Smart Body Area Network (SmartBAN).
Modal verbs terminology
In the present document "should", "should not", "may", "need not", "will", "will not", "can" and "cannot" are to be
interpreted as described in clause 3.2 of the ETSI Drafting Rules (Verbal forms for the expression of provisions).
"must" and "must not" are NOT allowed in ETSI deliverables except when used in direct citation.
ETSI
6 ETSI TR 103 395 39 V1.1.1 (2016-12)
1 Scope
The present document specifies the state-of-ththe-art and the future investigations on coexistence forr al allowing smart body
area network (SmartBAN) devices to properlerlyy work and co-operate in the Industrial, Scientific and M Medical (ISM)
band. Interference appears to be one of the mmaajor threats as well as coexistence with other existing s syystems radiating in
the same portion of the frequency spectrum. T The present document describes the coexistence measuurreements and
analysis that need to be considered in order tr too specify the requirements for the SmartBAN compatibleible devices.

Figigure 0: Scope of a SmartBAN
2 References
2.1 Normative refereennces
Normative references are not applicable in th thee present document.
2.2 Informative referrencese
References are either specific (identified byy d date of publication and/or edition number or version numumber) or
nonspecific. For specific references, only thhe e cited version applies. For non-specific references, the lae latest version of the
referenced document (including any amendmments) applies.
NOTE: While any hyperlinks includeedd in this clause were valid at the time of publication, ETTSSI cannot guarantee
their long term validity.
The following referenced documents are noot nt necessary for the application of the present document bu but they assist the
user with regard to a particular subject area.
[i.1] ETSI TS 103 326 (V1.1.1.1) (04-2015): "Smart Body Area Network (SmartBANN); En) hanced Ultra-
Low Power Physical LLaayyer".
[i.2] Void.
TM
[i.3] IEEE 802.11 : "IEEE SE Standard for Information technology--Telecommunicatatiions and
information exchange b beetween systems Local and metropolitan area networks----Specific
requirements Part 11: W Wireless LAN Medium Access Control (MAC) and Phyyssiical Layer (PHY)
Specifications".
ETSI
7 ETSI TR 103 395 V1.1.1 (2016-12)
[i.4] Valenta, V. (2010): "Survey on spectrum utilization in Europe: Measurements, analyses and
observations", 5th International Conference on Cognitive Radio Oriented Wireless Networks
Communications.
[i.5] ITU-R (2011): "ITU-R handbook for spectrum monitoring".
[i.6] Report Recommendation ITU-R SM.2256: "Spectrum occupancy measurements and evaluation".
[i.7] Report Recommendation ITU-R SM.2180 (2010): "Impact of ISM equipment on radio
communication services".
[i.8] Virk, M. H., Vuohtoniemi, R., Hämäläinen, M., Iinatti, J., & Mäkela, J.-P. (2015): "Stochastic
Spectral Occupancy Modeling: A Body Area Network Perspective in ISM Band", 9th International
Symposium on Medical Information & Communication Technology (ISMICT). Kamakura, Japan.
[i.9] J. J. Lehtomäki, e. a. (2012): "Energy detection based estimation of channel occupancy rate with
adaptive noise estimation", IEICE Transactions on Communications.
[i.10] Virk, M. H., Vuohtoniemi, R., Hämäläinen, M., Iinatti, J., & Mäkela, J.-P. (2014): "Spectrum
Occupancy Evaluations at 2.35-2.50 GHz ISM Band in a Hospital Environment", International
Conference on Body Area Networks (BodyNets'14). London, UK.
[i.11] ETSI TS 103 325 (2014): "Smart Body Area Network (SmartBAN); Low Complexity Medium
Access Control (MAC) for SmartBAN".
[i.12] Matlab, Product help, R2011b.
[i.13] Yazdandoost, K.Y. and Sayrafian-Pour, K.: "Channel Model for Body Area Network (BAN),"
IEEE P802.15-08-0780-09-0006, 2009.
[i.14] Proakis, J.G.: "Digital Communications", McGraw-Hill, 2001.
[i.15] Griffin, A.: "Coding CPFSK for Differential Demodulation." University of Canterbury
Christchurch, New Zealand, 2000.
[i.16] IEEE 802.15.6™ (2012): "IEEE Standard for Local and metropolitan area networks - Part 15.6:
Wireless Body Area Networks".
[i.17] Rahman M., Elbadry, M and Harjani R.: "An IEEE 802.15.6 Standard Compliant 2.5 nJ/Bit
Multiband WBAN Transmitter Using Phase Multiplexing and Injection Locking" IEEE Journal of
Solid-State Circuits, Vol. 50, No. 5, May 2015, pp. 1126 -1136.
3 Symbols and abbreviations
3.1 Symbols
For the purposes of the present document, the following symbols apply:
C Channel Number
.
, E( ) Expected Value
f Centre Frequency
c
H Null hypothesis
H Alternative Hypothesis
i Channel Identifier
K Number of Samples Collected from the Band in One Sweep
k Shape Parameter
Maximized Value Of Likelihood Function
n Number of Samples Collected in the Channel
Observed Value
Probability of False Alarm
(()) Sample Power j at Channel i
T Number of Sweeps
ETSI
8 ETSI TR 103 395 V1.1.1 (2016-12)
Threshold for Consecutive Mean Excision

t Time
X Sample Space
α Significance Level
Γ( ) Gamma Function
Arrival Rate
Location Parameter
Scale Parameter
Noise Threshold
shape parameter
σ The log-normal variance of the measured data between path loss and K-factor
k
σ The log-normal variance in dB around the mean, representing the variations measured at different
p
body and room locations. This parameter will depend on variations in the body curvature, tissue
properties and antenna radiation properties at different body locations.
E /N Energy per bit to noise power spectral density ratio
b 0
h Modulation index
I Implementation losses in dB
dB
K The fit with measurement data for the K-factor for low path loss
K K factor of Ricean distribution in dB
dB
L Pulse length
L Length of slot
slot
m Numerator of modulation index
m The average decay rate in dB/cm for the surface wave traveling around the perimeter of the body
m The slope of the linear correlation between path loss and K-factor
k
M M-ary number
NF Noise figure in dB
dB
n Zero mean and unit variance Gaussian random variable
k
n Zero mean and unit variance Gaussian random variable
p
p Denominator of modulation index
P The average loss close to the antenna
P The average attenuation of components in an indoor environment radiated away from the body and
reflected back towards the receiving antenna
P Bit error probability
b
PL Path loss in dB
dB
PPDU Times of PPDU repetition
rep
Q( ) Q function
R Data rate
S Receiver sensitivity
dBm
T T /L
min s slot
3.2 Abbreviations
For the purposes of the present document, the following abbreviations apply:
ACK Acknowledgement
AIC Akaike Information Criterion
ANL Average Noise Level
ARA Antenna Research Associate
AWGN Additive White Gaussian Noise
BAN Body Area Network
BCH Bose, Chaudhuri, and Hocquenghem
BER Bit Error Rate
BIC Bayesian Information Criterion
BLE Bluetooth Low Energy
BPF Bandpass Filter
BT BlueTooth
CCA Clear Channel Assessment
CCA-ED Clear Channel Assessment Based On Energy Detection
CDF Cumulative Distribution Function
CM Channel Model
CO Channel Occupancy
ETSI
9 ETSI TR 103 395 V1.1.1 (2016-12)
CSRR Clean Sample Rejection Rate
DSSS Direct Sequence Spread Spectrum
ED Energy Detection
EGC Equal Gain Combining
FBO Frequency Band Occupancy
FCME Forward Consecutive Mean Excision
FER Frame Error Rate
FH Frequency Hopping
GEV Generalized Extreme Value
GEVD Generalized Extreme Value Distribution
GFSK Gaussian Frequency Shift Keying
HI High Interference
ICT Information and Communication Technology
ISM Industrial, Scientific and Medical
ITU-R Telecommunication Union - Radio Communication Sector
JPG Joint Photographic Experts Group
KS Kolmogorov-Smirnov
LI Low Interference
LNA Low Noise Amplifier
MAC Medium ACcess
MATLAB Matrix Laboratory
NOTE: A multi-paradigm numerical computing environment and fourth-generation programming language. A
TM
proprietary programming language developed by MathWorks .
MC Measurement Campaign
Med-FCME Median Forward Consecutive Mean Excision
MLE Maximum Likelihood Estimate
MLSD Maximum-Likelihood Sequence Detector
MPDU MAC Protocol Data Unit
MRI Magnetic Resonance Imaging
OBW Occupied BandWidth
OFDM Orthogonal Frequency Division Multiplexing
OYS Oulun Yliopistollinen Sairaala (Oulu University Hospital)
PDF Probability Distribution Function
PHY PHYsical layer
PLCP Physical Layer Convergence Procedure
PPDU Physical-Layer Protocol Data Unit
PSDU Physical-layer Service Data Unit
RBW Resolution BandWidth
RF Radio Frequency
SA Spectrum Analyser
SNR Signal-to-Noise Ratio
SOE Spectrum Occupancy Evaluation
SRO Spectrum Resource Occupancy
SSC Spatial Sample Clustering
TC Technical Committee
TCME Threshold for Consecutive Mean Excision
TLSD t Location-Scale Distribution
TS Technical Specification
UHF Ultra High Frequency
UWB Ultra WideBand
WBAN Wireless Body Area Network
WI Work Item
WLAN Wireless Local Area Network
WPAN Wireless Personal Area Networks
ETSI
10 ETSI TR 103 395 V1.1.1 (2016-12)
4 Introduction and Background
Modern medical and health monitoring equipment is moving towards the trend of wireless connectivity between the
data collection or control centre and the medical devices or sensors. Therefore, the need for standardized
communication interfaces and protocols between the actors is required. This network of actors performing some
medical monitoring or functions in this context is called a Smart Body Area Network (Smart BAN).
Most emerging radio technologies for Wireless Personal Area Networks (WPAN) are designed to operate around the
2,4 GHz ISM band. Since both standardized (such as Bluetooth and IEEE 802.11™ [i.3]) and non-standardized
(proprietary) devices use the same frequency band, interference may lead to significant performance degradation of
medical (and other) devices operating in the band. The main goal of this work item (WI) is to describe the interference
problem, and to highlight a coexistence framework for the medical information and communication technologies (ICT)
to operate in a proximal environment. In the present document a synthesis of the problem of interference and
coexistence around the 2,4 GHz ISM band is given. Measurements carried out in hospital and campus will be described
in order to have a better insight on the problem. Then, the measurement campaigns exhaustively accumulated data in
order to formulate a mathematical model of the interference at the channel in the 2,36 - 2,5 GHz band will be described.
5 Coexistence
5.0 Introduction
A number of use cases have been identified as potential scenarios for SmartBAN. These use cases serve as scenarios
where the real channel occupancy measurements are needed. The environments to be considered for investigating the
coexistence issues are such as:
• Hospital
• Home
• Office
• Outdoor
These cases include the typical environments where a patient wearing a SmartBAN system lives and stays. However,
the present document is focusing on indoor environments only.
Moreover, existing interferers are classified into two classes based on their usage of the spectrum. Devices
implementing the direct sequence spread spectrum (DSSS) technique constitute one class of interferers that utilize a
fixed channel in the band. Typically this channel is 22 MHz wide, although the width of the signal depends on the
transmitter's implementation. The second class of interferers is represented by devices implementing a type of
frequency hopping (FH) mechanism. Note that the IEEE 802.11 [i.3] specifications include a frequency hopping
technique that uses a deterministic frequency pattern. On the other hand, the Bluetooth specifications define a pseudo-
random frequency sequence based on the Bluetooth device's address and its internal clock. While interference among
systems from the same type, such as Bluetooth on Bluetooth, or IEEE 802.11 [i.3] on IEEE 802.11 [i.3], interference
can be significant, it is usually considered early on in the design stages of the protocol (phenomena is called as
multiuser interference.) A third class can be included, which comprehend the devices using orthogonal frequency
division multiplexing (OFDM) technique. Therefore, the worst realistic interference scenario consists of a mix of
heterogeneous devices, i.e. devices belonging to different classes.
In evaluating the performance with respect to coexistence issues, variations in the operational environment need to be
considered, including both the characteristics of the interfering wireless services and the radio frequency (RF)
propagation characteristics. This ensures that the evaluation takes into account the uncertainty in an installation's
location and in the interfering traffic. Evaluating the performance requirements in terms of coexistence issues provides a
method for quantifying the applications interference susceptibility and assists in establishing usage policies.
The analytical model for evaluating the coexistence in terms of the operational environment is developed based on the
following process:
• Characterize the interference under static conditions, i.e. when both interfering and desired signals remain
stationary. Empirical test results are used to estimate model parameters and to substantiate the model.
ETSI
11 ETSI TR 103 395 V1.1.1 (2016-12)
• Extract a mathematical model of the aggregate interference in all operational environments.
5.1 Bands
For coexistence purposes the typical interference levels are evaluated in the following bands:
• ISM band
• 30 MHz before the ISM band
• 30 MHz after the ISM band
• Option for UWB lower band
6 Measurements
6.1 Background & Motivation
Radio frequency is a finite resource coordinated by regulatory bodies all over the globe. For medical usage, varying
regulations are imposed in different countries involving allocation of various chunks of both licensed and unlicensed
frequency resource [i.4]. Some of the license-free solutions include, e.g. sub-gigahertz ISM band, 2,4 GHz ISM band
and 3 GHz to 10 GHz ultrawide band etc. [i.4]. 2,4 GHz ISM band is an unregulated, license free frequency band where
many communication technologies share the frequency resources, e.g. wireless local area networks (WLAN), Bluetooth
(BT), wireless sensor networks, cordless phones, etc. In a hospital environment if a wireless body area network
(WBAN) is planned to be deployed in the 2,4 GHz ISM band, it will require certain measures in order to co-exist with
other wireless communication technologies.
ETSI SmartBAN physical layer (PHY) defines 40 channels in 2,4 GHz ISM band, including 37 data channels and
3 control channels. Each channel is 2 MHz wide and no guard bands between two adjacent channels are defined [i.1].
There had been other considerations as well for PHY solutions, e.g. channels in 2,36 GHz to 2,40 GHz band, which has
already been allocated for medical use in USA. However, in Europe this particular band has been allocated for LTE
special purpose use.
Increasing deployments of wireless technologies inside hospital premises will significantly increase the electromagnetic
clutter in hospital environments and hence interference will be the limiting factor. Spectrum occupancy evaluations
(SOEs) provide statistical quantification of spectrum utilization patterns and highlight temporal characteristics of the
band under consideration. In essence, SOEs can also be used as a probe to reflect upon the degree a victim network
would suffer under the influence of an aggressor. In other words, SOE can provide an insight regarding the
opportunities a wireless network will have, causing presumably no interference to already existing networks. Yet
another way to look at SOE is to study the suitability a particular frequency band is for deployment of a new kind of
network. In a nut-shell, it is highly motivating to perform SOEs in 2,36 GHz to 2,4 GHz band and 2,4 GHz ISM band in
order to characterize aggregate interference which would potentially be harmful to ETSI SmartBAN compliant devices.
6.2 Spectrum Occupancy Evaluations (SOEs)
Generally, spectrum occupancy measurements involve collection of measurement data, processing the measured data
for occupancy assessment and development of models to characterize the spectrum utilization. The key factors
influencing the measurements are measurement bandwidth span, channel bandwidths, number of channels, observation
time per channel, revisit time to a specific channel, total duration of monitoring and statistical integration time if
monitoring is sufficiently long. The International Telecommunication Union - Radio Communication Sector (ITU-R)
guidelines for spectrum occupancy measurements are elaborated in a special handbook [i.5] and also in two short
reports [i.7] and [i.6]. Stochastic parameters regarding spectrum occupancy evaluations include channel occupancy
(CO), frequency band occupancy (FBO) and spectrum resource occupancy (SRO). Spectrum occupancy measurements
usually involve:
• Spectrum sensing, utilizing an antenna coupled with a band pass filter and a spectrum analyser.
• Sample collection and saving the records in hard disk drive.
ETSI
12 ETSI TR 103 395 V1.1.1 (2016-12)
• Processing and analysis of recorded data in order to calculate occupancies of interest.
Spectrum occupancy measurement handbook published by ITU-R [i.5] suggests to divide the sample space, i.e. one full
record of the band (one sweep) into channels of the expected system, which is utilizing the spectrum resource. Then,
each channel is searched for number of samples above a predefined noise threshold. If more than 50 % of the samples in
the channel are above a noise threshold, channel is marked as occupied. In this way, individual channel occupancies
(CO) are calculated for all the channels.
There are two more metrics which can also be calculated, frequency band occupancy and spectrum resource occupancy.
FBO provides statistical information about how much the whole frequency band is used, independently of a particular
system. SRO is a system specific metric, which gives information about the utilization of resources available to a
specific system [i.5]. After intensive review of literature, it was found that there were propositions which lacked the
objectivity in relation to fully unregulated band, like ISM, with so diverse access technologies. Even those studies
which were performed especially for ISM band could not grasp the whole picture. For example:
i) Many of the studies considered revisit time of more than a second for a span of more or less 100 MHz, which
is impractical for bursty transmissions, and a large portion of the band might appear empty during observation.
If a too fast revisit time is used, one might not get enough samples above noise threshold in a certain channel
to declare it occupied with enough confidence.
ii) Another example of ambiguity is the channelization of the band under observation. A number of studies talked
about ISM band but end up finding occupancies for WLAN only. Although ITU-R suggests a method which
involved at first division of the sample space into channels of the wideband system, calculation of occupancies
and then modifying the sample space by deleting the samples identified inside a wideband channel. Secondly,
the modified sample space was supposed to be divided in channels of the narrowband system (or next
wideband system with narrower bandwidth as compared to the previous one) and then following the same
procedure as before. But in real scenarios, sometimes it can lead to misclassification or even no detection,
depending on the quality of samples acquired, which in turn depends on the radio channel conditions.
iii) Overestimation or underestimation of occupancies due to less careful noise threshold setting, insufficient
number of samples in order to obtain statistical confidence, and inability to dig the signal out in case of very
low signal-to-noise ratio (SNR). 50 % premise, used by ITU-R handbook, results in underestimation of true
occupancies, because channel nulls or under-sampling can vanquish formidable part of channel, which would
then appear to be as noise.
iv) A comprehensive study of occupancies in ISM band considering possible coexisting systems had been lacking.
v) Efficient and robust bandwidth and centre frequency estimation algorithms are missing.
In order to dig legitimate signals very close to noise levels, especially in case of low signal-to-noise ratio (SNR) or
spread spectrum scenarios, one cannot rely upon static noise thresholds. From previous research works, a dynamic noise
thresholding algorithm known as median forward consecutive mean excision (Med-FCME) [i.9] is adopted. Contrary to
previous implementations, the algorithm is applied per each sweep. The idea was to establish a noise threshold for every
sweep because after a single sweep, a blind period is encountered, when measurement equipment was saving the sweep
data, and supposedly the noise floor is also fluctuating. Clean sample rejection rate (CSRR) is a measure to quantify the
number of noise-only samples wrongly classified as outliers having signal components by FCME. The concept of
CSRR is presented in [i.9]. CSRR can be preset as a probabilistic limit so that only a given amount of misclassification
is tolerated, i.e. a false alarm. This metric is termed as desired probability of false alarm, . The desired

probability of false alarm value is given as a parameter to the FCME algorithm which calculates a threshold for
consecutive mean excision as:

T = -ln(PFA ) , (1)
CME DES
and then performs the iterative algorithm to find out the noise thresholds. In our measurement campaign, a target CSRR
is set to be 5 %, i.e. = 0,05. Hence, = 2,99.

The most important goals achieved through this undertaking can be listed as:
a) CO, FBO and SRO evaluation for the 2,35 - 2,50 GHz band in a hospital environment to characterize potential
interferers in two different countries (in our case Finland and Italy).
b) A novel centre frequency and bandwidth estimation algorithm named as Spatial Sample Clustering (SSC).
c) An overhauled, robust and much more objective mechanism for SOE evaluations with a sufficiently low
desired probability of false alarm.
ETSI
13 ETSI TR 103 395 39 V1.1.1 (2016-12)
d) Mathematical models for quantificatcation of interference.
For various equations including CO, FBO, S SRRO calculation and the complete analytical approach totowwards the problem,
please refer to [i.8].
The spectrum sensing approach used was Enneergy Detection (ED) because of its simplicity of impleemmentation. Other
methods like cyclostationarity or wavelet deecoc mposition cannot be used as any information about ththee signals being
encountered is known. This makes the procesess blind, i.e. any electromagnetic energy being radiated id into the channel has
to be taken into account and no decisions abobout the nature or any feature of the already existing syststeems is taken. This is
exactly the same as Clear Channels Assessmmeent (CCA) based on ED as defined in the recently acceptepted ETSI
SmartBAN PHY document [i.1].
A Neyman-Pearson type of energy detector chr chain is used and a decision statistics based on the dynaammically calculated
noise threshold is formulated. This problem c can be written mathematically as a hypothesis test, i.e. a a null hypothesis
that a channel contains only noise, and an altelternative hypothesis that a channel contains noise alongg w with a legitimate
signal.
:      $ %
:   !  " #  ! & $ %
(2)
where i is the channel identifier, n is the nummbber of samples collected from the channel, %% is ts the sample power j
at channel I, and is the noise threshold. So, o, if the average power in the channel exceeds a certain th threshold, there is a
signal plus noise in a channel (alternative hyyppothesis), otherwise the null hypothesis stands.
6.3 Measurement Caammpaigns
6.3.0 Introduction
Various measurement campaigns were undertertaken in Oulu (Finland) and Florence (Italy) to analyze te the channel usage
patterns in essentially at the 2,35 GHz to 2,5050 GHz band. Different analysis techniques have been appapplied in order to dig
out maximum information regarding varying ng spectrum usage mainly in modern hospital environmeenntts. Office and home
environments were studied in Florence onlyy. T. The process had been evolutionary and the campaignss d differ slightly in
parameter settings as well as in implementatioation perspective. More light will be shed on it in the follollowing clauses. After
evaluations of spectrum occupancy, mathemmaatical models for channel occupancy description were e exxtracted.
The measurement results, in both Finland anndd Italy, had been in accord in general. However, there h haad been slight
variations due to the differing radio environmnments, different measurement equipment and different a annalysis strategies.
The measurement campaigns carried out in O Oulu University Hospital, Oulu, Finland is first presentedted. Later the
corresponding measurement campaigns carrrriieed out in San Giuseppe Hospital in Empoli, Florence, I Ittala y is described.
6.3.1 Measurement camppaigns ina Oulu, Finland
6.3.1.0 Introduction
Oulun yliopistollinen sairaala (OYS, or Oululu University Hospital), situated in the city of Oulu is the e north most of the
five university hospitals in Finland. The hossppital is affiliated to the University of Oulu, Faculty of MMedie cine and
operates with more than a 1 000 beds. The h hoospital is also equipped with state-of-the-art medical eququipment, several
ambulatory bays and a helipad.
Three measurement campaigns were carried od out in OYS premises between December 2013 to June 22014. The following
list describes the locations used in the camppaiaigns:
th th
• Daily Surgery (10 - 16 December er 2013)
th th
• Accident & Emergency Ward (10 - 17 June 2014)
ETSI
14 ETSI TR 103 395 39 V1.1.1 (2016-12)
th th
• X-ray & Radiology (18 - 25 Junene 2014)
6.3.1.1 Daily Surgery SOEs (CCampa aign 1)
Figure 2 shows the map of the ground floor or of Daily surgery ward where the measurements were carrarried out. A red dot
shows the location where the measurement eq equipment was placed. Green circles show the locationss of o the active Wi-Fi
access points installed in the premises. Suchh c channel occupancy measurements are always highly depependent upon the
location or placement of the equipment becauause the measurement equipment itself does not emit annyy probing signal in
the air and the measurements rely only on tthhee energy detection perceived from the radio channel. Itt s should be noted
here that in this kind of measurement no asssuumptions on radio signal propagation characteristics (i.e.e. channel model
model) are needed, as only electromagneticc r raadiations' level in the air is of interest. In other words,, n noo signal decoding
is done here, it is just simple old school blindnd energy detection that was done.
However, sometimes it becomes important to to identify the interfering radiation or signal. That is whhyy in the data analysis
phase a blind detect and identify method (SSCSC) is implemented in order to characterize the specific sc syystems occupying
the frequency band of interest. For more dettails abai out SSC, please refer to [i.8].

Figure 1: Maapp of ground floor (Daily Surgery Ward)
TM
Measurements were carried out using high p perfe ormance spectrum analyser (SA) Agilent E4446A c connected to a
computer. Instrument Control Toolbox was us used to connect MATLAB directly to the spectrum anaallyyser enabling
control over SA and direct measurement ressuullts' analysis. The spectrum analyser was connected wiitthh a 1 m length cable
to an omnidirectional, wideband antenna ARRAA CMA-118/A. Measurement setup is displayed in Fiigguures 3 and 4. Before
the measurement campaign signal levels weere re measured and dynamic range of the spectrum analysseerr was optimized for
the specific environment. Measurement cammppaigns are affected by certain key parameters which are sre set before the start
of the campaign. Such parameters set for the ce campaign in Daily Surgery are listed in the Table 1.
TM
NOTE: SA Agilent E4446A is (are) ae) an example(s) of a suitable product(s) available commerercially. This
information is given for the coonnvenience of users of the present document and does nnoot cot nstitute an
endorsement by ETSI of this ( (tthhese) product(s).
ETSI
15 ETSI TR 103 39395 V1.1.1 (2016-12)

Figure 3: Actual cabinet keepeping the SA,
Figure 2: Logical Measurementnt Setup
omnidirectional antenna and cnd ontrol PC

Table 1: Parammeter setting for Daily Surgery Campaign
Parameter NaName Value
Frequency band 2,35 to 2,50 GHz
Bandwidth 150 MHz
Number of recorded frreqequency bins 1 601
Resolution bandwidth 300 kHz
Bin-width 93,7 kHz
No. of sweeps 10 000
Sweep time 2 ms approx.
Processing ti
...