ArticlePDF Available

Bluetooth - the universal radio interface for ad hoc, wireless connectivity

Authors:

Abstract and Figures

Bluetooth is a universal radio interface in the 2.45 GHz frequency band that enables portable electronic devices to connect and communicate wirelessly via short-range, ad hoc networks. Each unit can simultaneously communicate with up to seven other units per piconet. Moreover, each unit can simultaneously belong to several piconets. Bluetooth technology - which apart from Ericsson, has gained the support of Nokia, IBM, Toshiba, Intel and many other manufacturers - eliminates the need for wires, cables and connectors for and between cordless or mobile phones, modems, headsets, PDAs, computers, printers, projectors, local area networks, and so on, and paves the way for new and completely different devices and applications. Before guiding us through frequency-hopping technology and the channel, packet and physical-link definitions that characterize the Bluetooth air interface, the author briefly describes the conditions that led up to the development of Bluetooth. He then acquaints us with the networking aspects of Bluetooth technology, describing piconets and scatternets, connection procedures, and inter-piconet communication.
Content may be subject to copyright.
Imagine a cheap, power-efficient radio chip
that is small enough to fit inside any elec-
tronic device or machine, that provides local
connectivity, and that creates a (worldwide)
micro-scale web. What applications might
you use it in?
In 1994, Ericsson Mobile Communica-
tions AB in Lund, Sweden, initiated a study
to investigate the feasibility of a low-power,
low-cost radio interface between mobile
phones and their accessories. The intention
was to eliminate cables between phones and
PC cards, wireless headsets, and so forth.
The study was part of a larger project that
investigated multi-communicators con-
nected to the cellular network via cellular
telephones. The last link in the connection
between a communicator and the cellular
network was a short-range radio link to the
phone—thus, the link was called the multi-
communicator link or MC link. As the MC
link project progressed, it became clear that
there was no limit to the kinds of applica-
tion that could use a short-range radio link.
Cheap, short-range radios would make wire-
less communication between portable de-
vices economically feasible.
Current portable devices use infrared
links (IrDA) to communicate with each
other. Although infrared transceivers are in-
expensive, they
• have limited range (typically one to two
meters);
are sensitive to direction and require
direct line-of-sight;
can in principle only be used between two
devices.
By contrast, radios have much greater range,
can propagate around objects and through
various materials, and connect to many de-
vices simultaneously. What is more, radio
interfaces do not require user interaction.
In the beginning of 1997, when design-
ers had already begun work on an MC link
110 Ericsson Review No. 3, 1998
BLUETOOTH—The universal radio interface for
ad hoc
, wireless connectivity
Jaap Haartsen
Bluetooth is a universal radio interface in the 2.45 GHz frequency band that
enables portable electronic devices to connect and communicate wirelessly
via short-range,
ad hoc
networks. Each unit can simultaneously communi-
cate with up to seven other units per piconet. Moreover, each unit can
simultaneously belong to several piconets.
Bluetooth technology—which apart from Ericsson, has gained the support
of Nokia, IBM, Toshiba, Intel and many other manufacturers—eliminates the
need for wires, cables and connectors for and between cordless or mobile
phones, modems, headsets, PDAs, computers, printers, projectors, local
area networks, and so on, and paves the way for new and completely differ-
ent devices and applications.
Before guiding us through frequency-hopping technology and the channel,
packet and physical-link definitions that characterize the Bluetooth air inter-
face, the author briefly describes the conditions that led up to the develop-
ment of Bluetooth. He then acquaints us with the networking aspects of
Bluetooth technology, describing piconets and scatternets, connection pro-
cedures, and inter-piconet communication.
Printer
WORLD
11:33
CLR
123
456
789
*
0
#
e
ON/OFF
GH337
Cellular network
Mobile phone
Laptop
Laptop
Mouse
Headset
LAN Access point
Piconet
Figure 1
User model with local wireless connectivity.
Applications envisioned for the near future.
Ericsson Review No. 3, 1998 111
transceiver chip, Ericsson approached other
manufacturers of portable devices to raise in-
terest in the technology—for the system to
succeed, a critical mass of portable devices
must use the short-range radio. In February
1998, five promoters—Ericsson, Nokia,
IBM, Toshiba and Intel—formed a special
interest group (SIG). The idea was to achieve
a proper mix of business areas: two market
leaders in mobile telephony, two market
leaders in laptop computing, and a market
leader in core, digital-signal-processor
(DSP) technology. On May 20 and 21, 1998,
the Bluetooth consortium announced itself
to the general public from London, England;
San Jose, California; and Tokyo, Japan. Since
then, several companies have joined as
adopters of the technology (Box B).
The purpose of the consortium is to
establish a de facto standard for the air inter-
face and the software that controls it,
thereby ensuring interoperability between
devices of different manufacturers. The first
products to use MC link technology will
emerge at the end of 1999 in mobile phones,
notebook computers and accessories
(Figure 1).
Box A
Abbreviations
ACL Asynchronous connectionless
ARQ Automatic retransmission query
CVSD Continuous variable slope delta
DSP Digital signal processor
FEC Forward error correction
FH Frequency hop
FSK Frequency shift keying
HEC Header error correction
HPC Handheld personal computer
IrDA Infrared Data Association
ISM Industrial Scientific Medical
MAC Media access control
MC Multicommunicator
PC Personal computer
PDA Personal digital assistant
RF Radio frequency
SCO Synchronous connection-
oriented
SIG Special interest group
TDD Time division duplex
TDM Time division multiplex
Ericsson Promoter
Intel Promoter
IBM Promoter
Nokia Promoter
Toshiba Promoter
3Com
Axis
BreezeCOM
Casio
Cambridge consultantsLtd.
CETECOM GmbH
Cirrus Logic
Compaq Computer Corp.
Convergence Corporation
Dell Computer Corp.
InnoLabs Corporation
Jeeves Telecom Ltd.
Lucent Technologies UK Ltd.
Metricom
Motorola
NeoParadigm Labs, Inc.
Plantronics
Psion
Puma Technologies
Quadriga
Qualcomm, Inc.
Samsung Electronics Ltd.
Siemens Forsvarsystem A/S
Symbian
Symbionics Ltd.
T-Span System
Temic Semiconductor
TDK
TTP Communications Ltd.
Universal Empowering Technologies
VLSI Technology, Inc.
Xircom
* The name, Bluetooth, was taken from Har-
ald Blåtand, a Danish Viking king from the
early Middle Ages.
Box B
The Bluetooth consortium—promoters and adopters
The promoters of the Bluetooth* consortium formed a special interest group (SIG) at Ericsson
Inc., Research Triangle Park, North Carolina, on February 4, 1998.
The consortium was announced to the public on May 20 and 21, 1998. Many companies have
since joined the consortium as adopters of the technology (status as of July 11, 1998):
The Bluetooth air interface
The focus of user scenarios envisioned for
first-generation products is typically on
traveling business people. Portable devices
that contain Bluetooth radios would enable
them to leave cables and connectors at home
(Box C). Before the air interface for Blue-
tooth could be designed, however, certain
requirements had to be settled:
The system must operate worldwide.
• The connection must support voice and
data—for instance, for multimedia appli-
cations.
• The radio transceiver must be small and
operate at low power. That is, the radio
must fit into small, portable devices, such
as mobile phones, headsets and personal
digital assistants (PDA).
License-free band
To operate worldwide, the required fre-
quency band must be available globally.
Further, it must be license-free and open to
any radio system. The only frequency band
that satisfies these requirements is
at 2.45 GHz—the Industrial-Scientific-
Medical (ISM) band, which ranges from
2,400 to 2,483.5 MHz in the US and Eu-
rope (only parts of this band are available in
France and Spain), and from 2,471 to
2,497 MHz in Japan. Consequently, the
system can be used worldwide, given that
the radio transceivers cover the frequency
band between 2,400 and 2,500 MHz and
that they can select the proper segment in
this band.
Frequency hopping
Since the ISM band is open to anyone, radio
systems operating in this band must cope
with several unpredictable sources of inter-
ference, such as baby monitors, garage door
openers, cordless phones and microwave
ovens (the strongest source of interference).
Interference can be avoided using an adap-
tive scheme that finds an unused part of the
spectrum, or it can be suppressed by means
of spectrum spreading. In the US, radios op-
erating in the 2.45 GHz ISM band are re-
quired to apply spectrum-spreading tech-
niques if their transmitted power levels ex-
ceed 0 dBm.
Bluetooth radios use frequency-hop (FH)
spread spectrum, since this technology bet-
ter supports low-cost, low-power radio im-
plementations. Frequency-hop systems di-
vide the frequency band into several hop
channels. During a connection, radio trans-
ceivers hop from one channel to another in
a pseudo-random fashion. The instanta-
neous (hop) bandwidth is small in
frequency-hop radios, but spreading is usu-
ally obtained over the entire frequency band.
This results in low-cost, narrowband trans-
ceivers with maximum immunity to inter-
ference. Occasionally, interference jams a
hop channel, causing faulty reception.
When this occurs, error-correction schemes
in the link restore bit errors.
Channel definition
Bluetooth channels use a frequency-
hop/time-division-duplex (FH/TDD) scheme
(Figure 2). The channel is divided into
625 µs intervals—called slots—where a dif-
ferent hop frequency is used for each slot.
This gives a nominal hop rate of 1,600 hops
per second. One packet can be transmitted
112 Ericsson Review No. 3, 1998
Three-in-one phone—use the same phone
everywhere
When you are at the office, your phone func-
tions as an intercom (no telephony charge).
At home, it functions as a cordless phone
(fixed-line charge). When you are on the move,
it functions as a mobile phone (cellular
charge).
Internet bridge—surf the Internet regardless
of the connection
Use your portable PC to surf the Internet any-
where, regardless of whether you are con-
nected wirelessly through a mobile phone (cel-
lular) or through a wired connection (PSTN,
ISDN, LAN, xDSL).
Interactive conference—connect every par-
ticipant for instant data exchange
In meetings and at conferences, you can
share information instantly with other partici-
pants. You can also operate a projector
remotely without wire connectors.
The ultimate headset—a cordless headset
keeps your hands free
Connect a headset to your mobile PC or to any
wired connection and free your hands for more
important tasks at the office or in your car.
Portable PC speakerphone
Connect cordless headsets to your portable
PC and use it as a speakerphone regardless
of whether you are in the office, your car, or
at home.
Briefcase trick (hidden computing 1)
Access e-mail while your portable PC is still in
the briefcase. When your portable PC
receives an e-mail message, you will be noti-
fied by your mobile phone. You can also use
the phone to browse incoming e-mail and read
messages.
Forbidden message (hidden computing 2)
Compose e-mail on your PC while you are on
an airplane. When you land and are allowed
to switch on your mobile phone, the messages
are sent immediately.
Automatic synchronization (hidden
computing 3)
Automatically synchronize your desktop com-
puter, portable PC, notebook (PDA or HPC) and
mobile phone. As soon as you enter the office,
the address list and calendar in your notebook
automatically updates the files on your desk-
top computer or vice versa.
Instant postcard—send photos and video
clips instantly from any location
Connect a camera cordlessly to your mobile
phone or to any wire-bound connection. Add
comments from your mobile phone, a note-
book, or portable PC and send them instant-
ly to a recipient anywhere in the world. Suit-
able for professional and personal use.
Cordless desktop—connect all peripheral
tools to your PC or the LAN
Connect your desktop/laptop computer cord-
lessly to printers, scanners and the LAN.
Increase your sense of freedom through
cordless mouse and keyboard connections to
your PC.
Box C User scenarios
fkfk+1 fk+2
t
625 µst
Figure 2
Frequency-hop/time-division-duplex channel.
Ericsson Review No. 3, 1998 113
per interval/slot. Subsequent slots are alter-
nately used for transmitting and receiving,
which results in a TDD scheme.
Two or more units sharing the same chan-
nel form a piconet, where one unit acts as a
master, controlling traffic on the piconet,
and the other units act as slaves. The
frequency-hop channel is determined by the
frequency-hop sequence (the order in which
hops are visited) and by the phase in this se-
quence. In Bluetooth, the sequence is de-
termined by the identity of the piconet mas-
ter and phase is determined by the master
unit’s system clock (Figure 3). In order to
create the master clock in the slave unit, the
slave may add an offset to its own native
clock. The repetition interval of the
frequency-hop sequence, which is very long
(more than 23 hours), is determined by the
clock. If every participant on a given chan-
nel uses the same identity and clock as input
to the hop-selection box, then each unit will
consistently select the same hop carrier and
remain synchronized. Every piconet has a
unique set of master parameters which cre-
ate a unique channel.
The channel makes use of several, equal-
ly spaced, 1 MHz hops. With Gaussian-
shaped frequency shift keying (FSK) modu-
lation, a symbol rate of 1 Mbit/s can be
achieved. In countries where the open band
is 80 MHz or broader, 79 hop carriers have
been defined. In countries where the band
is narrower (Japan, France, and Spain), only
23 hop carriers have been defined (Table 1).
On average, the frequency-hop sequence vis-
its each carrier with equal probability.
Packet definition
In each slot, a packet can be exchanged be-
tween the master unit and one of the slaves.
Packets have a fixed format (Figure 4). Each
packet begins with a 72-bit access code that
is derived from the master identity and is
unique for the channel. Every packet ex-
changed on the channel is preceded by this
access code. Recipients on the piconet com-
pare incoming signals with the access code.
If the two do not match, the received pack-
et is not considered valid on the channel and
the rest of its contents are ignored. Besides
packet identification, the access code is also
used for synchronization and compensating
for offset. The access code is very robust and
resistant to interference. Correlation of the ac-
cess code by recipients provides similar pro-
cessing gains as direct-sequence spreading.
A header trails the access code. It contains
important control information, such as a
three-bit media-access-control (MAC) ad-
dress, packet type, flow control bits, bits for
the automatic-retransmission-query (ARQ)
scheme and a header-error-check (HEC)
field (Figure 5). The header, whose length
is fixed at 54 bits, is protected by a one-third
rate forward-error-correction (FEC) code.
Payload may or may not trail the header.
The length of the payload may vary from 0
to 2,745 bits.
To support high data rates, multi-slot
packets have been defined. A packet can
cover one slot, three slots, or five slots. Pack-
ets are always sent on a single-hop carrier.
For multi-slot packets, the hop carrier is
used as applied in the first slot. After the
multi-slot packet, the channel continues on
the hop as dictated by the master clock. For
example, let us consider four slots: k, k+1,
k+2and k+3. Ordinarily, these would be as-
sociated with hop frequencies fk, fk+1, fk+2
and fk+3. However, a three-slot packet that
starts in slot kuses fkfor the entire packet.
The next packet begins in slot k+3and uses
fk+3.
Physical link definition
Two types of link have been defined for sup-
porting multimedia applications that mix
voice and data:
synchronous connection-oriented (SCO)
link;
Native CLK
Hop selection
Hop
Offset Master identity
Phase
Sequence
+
Modulation
Peak data rate
RF bandwidth
RF band
RF carriers
Carrier spacing
Peak TX power
G-FSK,
h
0.35
1 Mbit/s
220 kHz (–3dB), 1 MHz (–20 dB)
2.4 GHz, ISM band
23/79
1 MHz
20 dBm
Parameters Values
TypeM_ADDR Flow ARQN SEQN HEC
34
111 8
Figure 5
Header fields.
Figure 4
Fixed packet format.
Table 1
Radio parameters.
Figure 3
Hop selection scheme: In the selection box,
the master identity selects the sequence,
and the clock selects the phase.
Combined, they give the hop carrier to be
used.
asynchronous connectionless (ACL) link.
SCO links support symmetrical, circuit-
switched, point-to-point connections typi-
cally used for voice. These links are defined
on the channel by reserving two consecutive
slots (forward and return slots) at fixed in-
tervals.
ACL links support symmetrical or asym-
metrical, packet-switched, point-to-
multipoint connections typically used for
bursty data transmission. Master units use
a polling scheme to control ACL connec-
tions.
A set of packets has been defined for each
physical link.
For SCO links, three kinds of single-slot
voice packet have been defined, each of
which carries voice at a rate of 64 kbit/s.
Voice is sent unprotected, but if the SCO
interval is decreased, a forward-error-
correction rate of 2/3 or 1/3 can be se-
lected.
• For ACL links, 1-slot, 3-slot, and 5-slot
data packets have been defined. Data can
be sent either unprotected or protected by
a 2/3 forward-error-correction rate. The
maximum data rate—721 kbit/s in one
direction and 57.6 kbit/s in the reverse di-
rection—is obtained from an unprotect-
ed, 5-slot packet. Table 2 summarizes the
data rates that can be obtained from ACL
links. DMx represents x-slot, FEC-
encoded data packets; DHx represents un-
protected data packets.
Figure 6 depicts mixed SCO and ACL links
on a piconet with one master and two slaves.
Slave 1 supports an ACL link and an SCO
link with a six-slot SCO interval. Slave 2
only supports an ACL link. Note: slots may
be empty when no data is available.
Interference immunity
As mentioned above, the Bluetooth radio
must operate in an open band that is sub-
ject to considerable uncontrolled interfer-
ence. Thus, the air interface has been opti-
mized to deal with interference.
Frequency hopping techniques are ap-
plied with a high hopping rate and short
packet lengths (1,600 hops/s for single-
slot packets). If a packet is lost, only a
small portion of the message is lost.
Packets can be protected by forward error
control.
• Data packets are protected by an ARQ
scheme in which lost data packets are au-
tomatically retransmitted. The recipient
checks each received packet for errors. If
errors are detected, it indicates this in the
header of the return packet. This results
in a fast ARQ scheme—delays are only
one slot in duration, and only packets that
have been lost need to be retransmitted.
Voice is never retransmitted. Instead, a ro-
bust voice-encoding scheme is used. This
scheme, which is based on continuous
variable slope delta (CVSD) modulation,
follows the audio waveform (Figure 7) and
is very resistant to bit errors—errors are
perceived as background noise, which in-
tensifies as bit errors increase.
Networking
Piconets
Bluetooth units that are within range of each
other can set up ad hoc connections. In prin-
ciple, each unit is a peer with the same hard-
ware capabilities (unlike cellular systems,
there is no distinction between terminals
and base stations). Two or more Bluetooth
units that share a channel form a piconet. To
regulate traffic on the channel, one of the
participating units becomes a master of the
piconet. Any unit can become a master, but
by definition, the unit that establishes the
piconet assumes this role. All other partici-
pants are slaves. Participants may change
roles if a slave unit wants to take over the
master role. Nonetheless, only one master
may exist in a piconet at any time.
Every unit in the piconet uses the master
identity and clock to track the hopping
channel. Each unit also has its own (native),
free-running clock. When a connection is
established, a clock offset is added to syn-
chronize the slave clock with the master
clock. The native clock is never adjusted,
however, and offsets are solely valid for the
duration of the connection.
Master units control all traffic on a chan-
nel. They allocate capacity for SCO links by
reserving slots. For ACL links, they use a
114 Ericsson Review No. 3, 1998
DM1
DH1
DM3
DH3
DM5
DH5
Type Values
108.8
172.8
256.0
384.0
286.7
432.6
108.8
172.8
384.0
576.0
477.8
721.0
108.8
172.8
54.4
86.4
36.3
57.6
Symmetric
(kbit/s) Asymmetric
(kbit/s)
Table 2
Achievable data rates (in kbit/s) on the ACL
link.
Master
Slave 1
Slave 2
SCO ACL SCO ACLACL
Figure 6
SCO and ACL links in a piconet with one
master and two slaves.
Ericsson Review No. 3, 1998 115
polling scheme. A slave is only permitted to
send in the slave-to-master slot when it has
been addressed by its MAC address in the
preceding master-to-slave slot. A master-to-
slave packet implicitly polls the slave; that
is, an ordinary traffic packet addressed to a
slave polls the slave automatically. If no in-
formation to the slave is available, the mas-
ter can use a POLL packet to poll the slave
explicitly. POLL packets consist of an access
code and header only. This central polling
scheme eliminates collisions between slave
transmissions.
Establishing connection
When units are not participating in a pico-
net, they enter standby mode, from which
state they periodically listen for page mes-
sages. From the total set of 79 (23) hop car-
riers, a subset of 32 (16) wake-up carriers
has been defined. The subset, which is cho-
sen pseudo-randomly, is determined by the
unit identity. Over the wake-up carriers, a
wake-up sequence visits each hop carrier
once: the sequence length is 32 (16) hops.
Every 2,048 slots (1.28 s), standby units
move their wake-up hop carrier forward one
hop in the wake-up sequence. The native
clock of the unit determines the phase of the
wake-up sequence. During the listening in-
terval, which lasts for 18 slots or 11.25 ms,
the unit listens on a single wake-up hop car-
rier and correlates incoming signals with the
access code derived from its own identity. If
the correlator triggers—that is, if most of
the received bits match the access code—the
unit activates itself and invokes a
connection-setup procedure. Otherwise, the
unit returns to sleep until the next wake-up
event.
Units connecting to a unit in standby
mode must know the standby unit’s identi-
ty and preferably its native clock
to generate the required access code
(which constitutes the paging message);
to derive the wake-up sequence;
to predict the phase of this sequence.
Since paging units cannot accurately know
the native clock of a recipient, they must re-
solve the time-frequency uncertainty. They
do so by transmitting the access code con-
tinuously—not only in the hop in which
they expect the recipient to wake up, but
also in hops before and after. For a period of
10 ms, paging units transmit the access code
sequentially on several hop frequencies
around the expected hop carrier. Note: the
access code is only 72 bits long (72 ms).
Therefore, many codes can be sent in the
space of 10 ms. The 10 ms train of access
codes on different hop carriers is transmit-
ted repeatedly until the recipient responds
or a time-out is exceeded.
When a paging unit and recipient select
the same wake-up carrier, the recipient re-
ceives the access code and returns an ac-
knowledgement. The paging unit then
sends a packet containing its identity and
its current clock. After the recipient ac-
knowledges this packet, each unit uses the
paging unit’s parameters for hop selec-
tion—thereby establishing a piconet in
which the paging unit acts as the master.
To establish a connection, the paging unit
must obtain the identity of units within
transmission range. Therefore, it executes an
inquiry procedure: the paging unit trans-
mits an inquiry access code (which is com-
mon to all Bluetooth devices) on the inquiry
wake-up carriers. When a recipient receives
the inquiry, it returns a packet containing
its identity and clock—the very opposite of
the paging procedure. After having gath-
ered each response, the paging unit can then
select a specific unit to page (Figure 8).
Scatternet
Users of a channel must share capacity. Al-
though channels are 1 MHz wide, as more
and more users are added, throughput per
user quickly drops to less than a few tens of
kbit/s. Furthermore, although the medium
available bandwidth is 80 MHz in the US
and Europe (slightly less in Japan, France
and Spain), it cannot be used efficiently
when every unit must share the same 1 MHz
hop channel. Therefore, another solution
was adopted.
1 1 0 0 0 0 0 0 1 0 1 1 1 1 1 0 1 0 0 0 0 1 1 1 0 0 0 1 0 1 0 1 0 . . . . . . .
Figure 7
Continuous variable slope delta (CVSD) wave-
form coding.
Inquiry Page Connection
Typical
Max.
5.12 s
15.36 s
0.64 s
7.68 s
0.1-300 minutes
Figure 8
Connection-establishment procedure and
maximum time associated with establishing
a connection.
Units that share the same area and that
are within range of one another can poten-
tially establish ad hoc connections between
themselves. However, only those units that
truly want to exchange information share
the same channel (piconet). This solution
permits several piconets to be created with
overlapping areas of coverage. Each piconet
adheres to its own hopping sequence
through the 80 MHz medium. The channel
in each piconet hops pseudo-randomly over
the carriers in the 80 MHz band. The users
in each piconet have only a 1 MHz hop chan-
nel at their disposal.
A group of piconets is called a scatternet.
Aggregate and individual throughput of
users in a scatternet is much greater than
when each user participates on the same pi-
conet with a 1 Mbit/s channel. Additional
gains are obtained by statistically multi-
plexing hop channels and by reusing chan-
nels. The 1 MHz hop channel in any given
piconet need only be shared by users of that
piconet. Because individual piconets hop
differently, different piconets can simulta-
neously use different hop channels. Conse-
quently, units in one piconet do not share
their 1 MHz channel with units in another
piconet. The aggregate throughput (the
total throughput accumulated over all pi-
conets) increases as more piconets are added.
Collisions do occur, however, when two pi-
conets use the same hop channel simultane-
ously. As the number of piconets increases,
performance in the frequency-hop system
degrades gracefully. Simulations of a scat-
ternet consisting of 10 piconets indicate that
reduction in throughput per piconet is less
than 10%. In the scatternet, the radio medi-
um is shared; in a piconet, the channel and
information are shared.
Since every piconet uses the same band-
width, each shares the 80 MHz in an aver-
age sense. Provided they select different hop
channels, however, no two piconets must si-
multaneously share the same 1 MHz chan-
nel.
Let us assume there are 100 users. If each
belonged to the same network, all 100 users
would have to share the same 1 MHz chan-
nel. Thus, average throughput per user
would be 10 kbit/s and aggregate through-
put would be 1 Mbit/s. However, if not
everyone wanted to talk to each other, we
could split the piconet into independent
piconets. For example, if the users separat-
ed themselves into groups of five, then we
could create 20 independent piconets. With
only five users sharing the 1 MHz hop chan-
nel, throughput per user increases to
200 kbit/s and aggregate throughput in-
creases to 20 Mbit/s. Obviously, this as-
sumes ideal conditions, where no two pico-
nets select the same hop channel at the same
time. In reality—because the piconets hop
independently—collisions will occur, re-
ducing effective throughput. Nonetheless,
the final throughput obtained from multi-
ple piconets exceeds that of a single piconet.
The maximum number of units that can
actively participate on a single piconet is
eight: one master and seven slaves. The
MAC address in the packet header, which is
used to distinguish each unit, is limited to
three bits. Figure 9 illustrates the scatternet
approach applied to the scenario shown in
Figure 1.
Inter-piconet communication
Different piconets adhere to different
frequency-hop sequences and are controlled
116 Ericsson Review No. 3, 1998
Printer
WORLD
11:33
CLR
123
456
789
*
0
#
e
ON/OFF
GH337
Mobile phone
Laptop
Laptop
Mouse
Headset
LAN Access point
Figure 9
A scatternet of four piconets applied to the
scenario described in Figure 1.
Native CLK
Hop selection
Hop
Piconet X
Phase
Sequence
+
Piconet Y
Piconet Z
Offset X Identity X
Offset Y Identity Y
Offset Z Identity Z
Figure 10
Hop selection in inter-piconet communica-
tion.
Ericsson Review No. 3, 1998 117
by different masters. If a hop channel is tem-
porarily shared by independent piconets,
packets can be distinguished by the access
codes that precede them—access codes are
unique for each piconet. Piconets are unco-
ordinated and hop independently: synchro-
nization of different piconets is not permit-
ted in the unlicensed ISM band. Nonethe-
less, units may participate in different pico-
nets on a time-division-multiplexing
(TDM) basis. That is, a unit can sequen-
tially participate in different piconets, pro-
vided it is active in only one piconet at a
time.
Inter-piconet communication is achieved
by selecting the proper master identity and
clock offset in order to synchronize with the
channel of the desired piconet (Figure 10).
A corresponding set of identity and clock
offsets is available for each piconet. When-
ever a unit enters a piconet, it adjusts the
clock offset to account for minor drifts be-
tween the master clock and the unit’s native
clock. A unit can thus act as a slave in sev-
eral piconets. When leaving one piconet for
another, a slave informs the current master
that it will be unavailable for a predeter-
mined period. During its absence, traffic on
the piconet between the master and other
slaves continues as usual.
A master unit can also periodically jump
to another piconet and act as a slave (were it
to act as a master in the new piconet, that
piconet would have the same channel para-
meters as the “old” piconet—therefore, by de-
finition the two would be indiscernible).
When a master unit leaves a piconet, all traf-
fic on the piconet is suspended until it
returns.
Figure 11 shows a slave participating in
two piconets. Piconet Xconsists of master
Xand slaves Axand Bx. Piconet Yconsists
of master Yand slaves Ay,By,and Dy.Slave
Cxy participates in piconets Xand Y. The
clock of each unit is also shown. Positive
offset (indicated by blue) or negative offset
(indicated by red) has been added for syn-
chronization with the master clock. Slave
Cxy contains a native clock and two offsets
for the master units Xand Yrespectively.
Authentication and
encryption
To ensure user protection and information
secrecy, the system must provide security
measures that are appropriate for a peer en-
vironment—that is, each unit in Bluetooth
must implement authentication and en-
cryption algorithms in the same way.
A base level encryption has been specified
that is well suited to silicon implementa-
tion, and an authentication algorithm has
been specified that provides a level of secu-
rity for devices lacking in processing capa-
bilities. Future support for ciphering algo-
rithms will be backward-compatible.
The main security features are:
a challenge-response routine—for au-
thentication;
stream cipher—for encryption;
session key generation—session keys can
be changed at any time during a connec-
tion.
Three entities are used in the security algo-
rithms: the Bluetooth unit address, which
is a public entity; a private user key, which
is a secret entity; and a random number,
which is different for each new transaction.
As described above, the Bluetooth address
can be obtained through an inquiry proce-
dure. The private key is derived during ini-
tialization and is never disclosed. The ran-
dom number is derived from a pseudo-
random process in the Bluetooth unit.
Conclusion
Bluetooth is a system for providing local
wireless connectivity between portable de-
vices. It is particularly suitable for ad hoc net-
working. The air interface has been opti-
mized to provide maximum immunity
against interference in the 2.45 GHz band.
The system is supported by several leading
manufacturers of personal computers and
telecommunications equipment. The first
consumer products to support Bluetooth are
expected to appear on the market around
year-end 1999.
Slave Cxy
Native
Offset X
Offset Y
Slave Ax
Master X
Slave Bx
Slave AySlave By
Slave Dy
Master Y
Piconet X Piconet Y
Figure 11
Participation of slave
Cxy
in two piconets.
... After testing our fall detector on a variety of seniors, we discovered a significant limitation of our device: Bluetooth connectivity [14]. This limitation affects the accuracy and practicality of our approach, as it relies on Bluetooth connectivity to a mobile phone for timely notifications. ...
Conference Paper
In today's society, elderly adults living alone face various challenges. From this we can feel the indifference of modern society and the danger of falling for elderly people living alone. By recounting my grandfather's personal experience of a fall, the urgent need for assistance among elderly adults and the potential dangers of isolation are highlighted. The core question is: How can technology be used to improve the safety and overall quality of life of elderly adults living alone [1]? To solve this serious problem, we developed a fall detection device. When elderly adults wear fall detection devices, it can address these challenges by providing a mechanism to send instant alerts to family members through a connected app in the event of a fall or emergency. The device aims to improve the safety and well-being of the elderly, provide timely intervention, reduce social burdens, and solve potential social problems [2]. In addition to direct benefits, fall detectors have significant commercial value and can have a positive impact on the quality of life of older adults. The device enhances physical and mental health, increasing independence and safety in outdoor activities with fall detection and emergency alert features [3]. Fall detectors not only have commercial value, positive impact and wider social contribution, they can also play a role in addressing social neglect and inspiring elderly adults to make positive changes.
... The wearable sensor also demonstrates excellent robustness over 10 repeat cycles of multidirectional stretching and bending and a minimal repeatability error less than 4% (Figure 1c), and subsequent experiments show that it also has good fatigue resistance. For the functional human-machine interface, the sensor is combined with gloves and a wireless module [38][39][40] to measure finger bending angle very sensitively (Figure 1d,e), and a series of applications has been implemented such as controlling the Bluetooth car with two−finger gestures with multidirectional stretching and bending (Figure 1f) [41]. The results demonstrate the multifunctional sensing abilities of the developed liquid metal−based sensor and verify its potential in wearable applications as an interaction interface between humans and machines. ...
Article
Full-text available
Rigid sensors are a mature type of sensor, but their poor deformation and flexibility limit their application range. The appearance and development of flexible sensors provide an opportunity to solve this problem. In this paper, a resistive flexible sensor utilizes gallium−based liquid metal (eutectic gallium indium alloy, EGaIn) and poly(dimethylsiloxane) (PDMS) and is fabricated using an injecting thin−line patterning technique based on soft lithography. Combining the scalable fabrication process and unique wire−shaped liquid metal design enables sensitive multifunctional measurement under stretching and bending loads. Furthermore, the flexible sensor is combined with the glove to demonstrate the application of the wearable sensor glove in the detection of finger joint angle and gesture control, which offers the ability of integration and multifunctional sensing of all−soft wearable physical microsystems for human–machine interfaces. It shows its application potential in medical rehabilitation, intelligent control, and so on.
... The 1-hop interference model is also referred to in the literature as the primary interference model or the node-exclusive interference model. Applications where this interference model arises include Bluetooth networks, where one channel in a piconet is shared by all the links between a master and its slaves [22], and CDMA systems when each node is equipped with only one transceiver [23] [24]. For the 1-hop interference model, the problem of finding a maximum independent set in the conflict graph is equivalent to finding a maximum matching in the network graph, which is solvable in polynomial time [42]. ...
Preprint
Full-text available
div> If each node in a wireless network has information about only its 11-hop neighborhood, then what are the limits to performance? This problem is considered for wireless networks where each communication link has a minimum bandwidth quality-of-service (QoS) requirement. Links in the same vicinity contend for the shared wireless medium. The conflict graph captures which pairs of links interfere with each other and depends on the MAC protocol. In IEEE 802.11 MAC protocol-based networks, when communication between nodes ii and jj takes place, the neighbors of both ii and jj remain silent. This model of interference is called the 22-hop interference model because the distance in the network graph between any two links that can be simultaneously active is at least 22. In the admission control problem studied in the present paper, the objective is to determine, using only localized information, whether a given set of flow rates is feasible. While distance-dd distributed algorithms have been analyzed for the 11-hop interference model, an open problem in the literature is to extend these results to the KK-hop interference model, and the present work initiates the generalization to the KK-hop interference model. We show that the centralized version of the problem is NP-hard and then investigate distributed, low-complexity solutions for this problem. We propose a distributed algorithm for this problem where each node has information about only its 11-hop neighborhood. The worst-case performance of the distributed algorithm, i.e. the largest factor by which the performance of this distributed algorithm is away from that of an optimal, centralized algorithm, is analyzed. Lower and upper bounds on the suboptimality of the distributed algorithm are obtained, and both bounds are shown to be tight. The exact worst-case performance is obtained for some ring topologies. The performance of the distance-11 distributed algorithm is compared with that of the row constraints, and these two distributed algorithms are shown to be incomparable. </div
... The 1-hop interference model is also referred to in the literature as the primary interference model or the node-exclusive interference model. Applications where this interference model arises include Bluetooth networks, where one channel in a piconet is shared by all the links between a master and its slaves [22], and CDMA systems when each node is equipped with only one transceiver [23] [24]. For the 1-hop interference model, the problem of finding a maximum independent set in the conflict graph is equivalent to finding a maximum matching in the network graph, which is solvable in polynomial time [42]. ...
Preprint
Full-text available
If each node in a wireless network has information about only its 11-hop neighborhood, then what are the limits to performance? This problem is considered for wireless networks where each communication link has a minimum bandwidth quality-of-service (QoS) requirement. Links in the same vicinity contend for the shared wireless medium. The conflict graph captures which pairs of links interfere with each other and depends on the MAC protocol. In IEEE 802.11 MAC protocol-based networks, when communication between nodes ii and jj takes place, the neighbors of both ii and jj remain silent. This model of interference is called the 22-hop interference model because the distance in the network graph between any two links that can be simultaneously active is at least 22. In the admission control problem studied in the present paper, the objective is to determine, using only localized information, whether a given set of flow rates is feasible. While distance-dd distributed algorithms have been analyzed for the 11-hop interference model, an open problem in the literature is to extend these results to the KK-hop interference model, and the present work initiates the generalization to the KK-hop interference model. We show that the centralized version of the problem is NP-hard and then investigate distributed, low-complexity solutions for this problem. We propose a distributed algorithm for this problem where each node has information about only its 11-hop neighborhood. The worst-case performance of the distributed algorithm, i.e. the largest factor by which the performance of this distributed algorithm is away from that of an optimal, centralized algorithm, is analyzed. Lower and upper bounds on the suboptimality of the distributed algorithm are obtained, and both bounds are shown to be tight. The exact worst-case performance is obtained for some ring topologies. The performance of the distance-11 distributed algorithm is compared with that of the row constraints, and these two distributed algorithms are shown to be incomparable.
... Previous studies also examined how long it takes to detect a Bluetooth device from stationary observers and moving vehicles. They found that Bluetooth devices are often detected in less time than the Bluetooth standard suggests [39,43,24]. This work supports our findings that skimmers are often discovered within the first few seconds of passing by a gas station. ...
Article
This paper presents estimation and analysis of the higher order harmonics, power features, and real performance of flip-flop and master-slave latch topologies. This research article outlines the impact of transistor model quality and input signal selection on the estimate of higher order harmonic contents of switching waveform emitted by the digital integrated circuits. Highly integrated systems require accurate estimation of higher order harmonics to control noise. This work presents simulations of 12 kinds of flip-flop and latch topologies on different process technologies i.e., 28 nm, 45 nm, 65 nm, and 130 nm. It is implied that the steeper the spectrum roll-off, the fewer the contents of higher order harmonics. Results of 28 nm process design kit are improved compared to 45 nm, 65 nm, and 130 nm process design kits. Furthermore, the results of the comparison of representative flip-flop and latch topologies illustrate the advantage of the approach and the suitability for high performance and low power consumption.
Thesis
Full-text available
The number of devices in wireless Internet of Things (IoT) networks is now rapidly increasing and is expected to continue growing in the coming years. To support this massive connectivity, a number of new technologies, collectively known as Low Power Wide Area Network (LPWAN), have been developed. Many devices in LPWANs limit their transmissions by duty cycle constraints; i.e., the proportion of time allocated for transmission. For nearby wireless networks using the same time-frequency resources, the increasing number of devices leads to a high level of unintended signals, known as interference. In this thesis, we characterize the statistics of interference arising from LPWANs, with a focus on protocols related to Narrowband IoT (NB-IoT) and emerging approaches such as Sparse Code Multiple Access (SCMA). Such a characterization is critical to improve signal processing at the receiver in order to mitigate the interference. We approach the characterization of the interference statistics by exploiting a mathematical model of device locations, signal attenuation, and the access protocols of individual interfering devices. While there has been recent work developing empirical models for the interference statistics, this has been limited to studies of the interference power, which has limited utility in receiver design. The approach adopted in this thesis has the dual benefits of providing a model for the amplitude and phase statistics and while also yielding insights into the impact of key network parameters. The first contribution in this work is to revisit interference in a single subcarrier system, which is widely used in current implementations of IoT networks. A basic model in this scenario distributes interfering devices according to a homogeneous Poisson point process. It has been long known that the resulting interference is well approximated via an α-stable model, rather than a Gaussian model. In this work, the α-stable model is shown via theoretical and simulation results to be valid in a wider range of models, including the presence of guard zones, finite network radii, and non-Poisson point processes governing device locations. The second contribution in this thesis is the study, for the first time, of interference statistics in multi-carrier IoT networks, including those that exploit NB-IoT and SCMA. Motivated by the results in the single subcarrier setting, a multivariate model based on α-stable marginals and copula theory is developed. This model is verified by extensive simulations and further justified via a new, near-optimal, parameter estimation algorithm, which has very low complexity.The third part of this thesis applies the characterizations of the interference statistics to receiver design. A new design for nonlinear receivers is proposed that can significantly outperform the state-of-the-art in multicarrier IoT systems. When receivers are restricted to be linear, the optimal structure is identified and the bit error rate characterized. Numerical results also illustrate how the average quantity of data interfering devices are required to transmit affects the receiver performance.
ResearchGate has not been able to resolve any references for this publication.