ArticlePDF Available

Abstract and Figures

Wi-Fi Direct is a new technology defined by the Wi-Fi Alliance aimed at enhancing direct device to device communications in Wi-Fi. Thus, given the wide base of devices with Wi-Fi capabilities, and the fact that it can be entirely implemented in software over traditional Wi- Fi radios, this technology is expected to have a significant impact. In this paper we provide a thorough overview of the novel functionalities defined in Wi-Fi Direct, and present an experimental evaluation that portrays the performance to be expected in real scenarios. In particular, our results quantify the delays to be expected in practice when Wi-Fi Direct devices discover each other and establish a connection, and the performance of its novel power saving protocols. To the best of the authors’ knowledge this is the first paper in the state of the art that provides a wide overview and experimental evaluation of Wi-Fi Direct.
Content may be subject to copyright.
IEEE Wireless Communications • June 2013
96 1536-1284/13/$25.00 © 2013 IEEE
ACCEPTED FROM OPEN CALL
INTRODUCTION
More than a decade after its initial design, the
IEEE 802.11 standard [1], has become one of
the most common ways to access the Internet.
However, to continue with its striking success
the Wi-Fi technology needs to evolve and
embrace a larger set of use cases. Given the
wide adoption of Wi-Fi in many kinds of devices,
a natural way for the technology to progress is to
target device-to-device connectivity, i.e. without
requiring the presence of an Access Point (AP),
traditionally provided by other technologies [2].
This is the purpose of the Wi-Fi Direct technol-
ogy that has been recently developed by the Wi-
Fi Alliance [3].
Direct device to device connectivity was
already possible in the original IEEE 802.11
standard by means of the ad hoc mode of opera-
tion. However, this never became widely
deployed in the market and hence presents sev-
eral drawbacks when facing nowadays require-
ments, e.g. lack of efficient power saving support
or extended QoS capabilities. Another relevant
technology in the Wi-Fi device to device com-
munications space is 802.11z, also known as
Tunneled Direct Link Setup (TDLS) [4], which
enables direct device to device communication
but requires stations to be associated with the
same AP.
Unlike the previous technologies, the Wi-Fi
Direct technology takes a different approach to
enhance device to device connectivity. Instead of
leveraging the ad-hoc mode of operation, Wi-Fi
Direct builds upon the successful IEEE 802.11
infrastructure mode and lets devices negotiate
who will take over the AP-like functionalities.
Thus, legacy Wi-Fi devices may seamlessly con-
nect to Wi-Fi Direct devices (as explained in
detail below). By taking this decision, Wi-Fi
Direct immediately inherits all the enhanced
QoS, power saving, and security mechanisms
(e.g., [5, 6]) developed for the Wi-Fi infra-
structure mode in the past years.
Wi-Fi Direct being a recent specification, its
deployment is still on a very early stage. There is
however an initial open source implementation,
available in [7], which we have used to evaluate
experimentally the performance of this technolo-
gy in realistic scenarios. In the following, we
review the main contributions of the article.
Firstly, we provide an overview of the Wi-Fi
Direct specification, focusing on its novel func-
tionalities and illustrating three representative
group formation procedures. We then present an
experimental evaluation of the delay associated
with these group formations, based on a popular
open source implementation of Wi-Fi Direct [7],
and compare its performance against the expect-
ed figures from simulations, discussing the
observed discrepancies. Finally, we implement a
novel power saving protocol (the “Notice of
Absence”) and assess the resulting performance
trade-offs in various realistic environments.
This article is organized as follows. The sec-
tion below provides a detailed overview of Wi-Fi
Direct. The section following that presents an
experimental evaluation of Wi-Fi Direct that
analyzes the performance of its group formation
procedures and of its power saving protocols.
Finally, the last section concludes the article.
WI-FIDIRECT: A TECHNICAL OVERVIEW
In a typical Wi-Fi network, clients discover and
associate to WLANs, which are created and
announced by Access Points (APs). In this way,
DANIEL CAMPS-MUR, NEC NETWORK LABORATORIES
ANDRES GARCIA-SAAVEDRA AND PABLO SERRANO, UNIVERSIDAD CARLOS III DE MADRID
ABSTRACT
Wi-Fi Direct is a new technology defined by
the Wi-Fi Alliance aimed at enhancing direct
device to device communications in Wi-Fi. Thus,
given the wide base of devices with Wi-Fi capa-
bilities, and the fact that it can be entirely imple-
mented in software over traditional Wi-Fi radios,
this technology is expected to have a significant
impact. In this article we provide a thorough
overview of the novel functionalities defined in
Wi-Fi Direct, and present an experimental eval-
uation that portrays the performance to be
expected in real scenarios. In particular, our
results quantify the delays to be expected in
practice when Wi-Fi Direct devices discover
each other and establish a connection, and the
performance of its novel power saving protocols.
To the best of the authors’ knowledge this is the
first article in the state of the art that provides a
wide overview and experimental evaluation of
Wi-Fi Direct.
DEVICE-TO-DEVICE COMMUNICATIONS WITH WIFI
DIRECT: OVERVIEW AND EXPERIMENTATION
CAMPS-MUR LAYOUT_Layout 1 6/21/13 11:52 AM Page 96
IEEE Wireless Communications • June 2013 97
a device unambiguously behaves either as an AP
or as a client, each of these roles involving a dif-
ferent set of functionality. A major novelty of
Wi-Fi Direct is that these roles are specified as
dynamic, and hence a Wi-Fi Direct device has to
implement both the role of a client and the role
of an AP (sometimes referred to as Soft-AP).
These roles are therefore logical roles that could
even be executed simultaneously by the same
device, for instance by using different frequen-
cies (if the device has multiple physical radios)
or time-sharing the channel through virtualiza-
tion techniques (similar to [8] or [9]). In order to
establish a communication, then, P2P devices
have to agree on the role that each device will
assume. In the following we describe how this
communication is configured, by first introduc-
ing the general architecture and then summariz-
ing the main specified procedures, namely device
discovery, role negotiation, service discovery,
security provisioning and power saving.
ARCHITECTURE
Wi-Fi Direct devices, formally known as P2P
Devices, communicate by establishing P2P
Groups, which are functionally equivalent to tra-
ditional Wi-Fi infrastructure networks. The
device implementing AP-like functionality in the
P2P Group is referred to as the P2P Group
Owner (P2P GO), and devices acting as clients
are known as P2P Clients. Given that these roles
are not static, when two P2P devices discover
each other they negotiate1their roles (P2P Client
and P2P GO) to establish a P2P Group. Once
the P2P Group is established, other P2P Clients
can join the group as in a traditional Wi-Fi net-
work. Legacy clients can also communicate with
the P2P GO, as long as they are not 802.11b-
only devices and support the required security
mechanisms (discussed later). In this way, legacy
devices do not formally belong to the P2P Group
and do not support the enhanced functionalities
defined in Wi-Fi Direct, but they simply “see”
the P2P GO as a traditional AP.2
The logical nature of the P2P roles supports
different architectural deployments, two of them
illustrated in Fig. 1. The upper part of the figure
represents a scenario with two P2P groups. The
first group is created by a mobile phone sharing
its 3G connection with two laptops; for this first
group, the phone is acting as P2P GO while the
two laptops behave as P2P Clients. In order to
extend the network, one of the laptops establish-
es a second P2P Group with a printer; for this
second group, the laptop acts as P2P GO. In
order to act both as P2P Client and as P2P GO
the laptop will typically alternate between the
two roles by time-sharing the Wi-Fi interface;
later we introduce the NoA protocol that can be
used for this purpose. The lower part of Fig. 1
illustrates the case of a laptop accessing the
Internet through a legacy infrastructure AP
while at the same time streaming content to a
TV set by establishing a P2P Group, where the
laptop acts as P2P GO (see [10] for more illus-
trative examples).
Like a traditional AP, a P2P GO announces
itself through beacons, and has to support power
saving services for its associated clients. The P2P
GO is also required to run a Dynamic Host Con-
figuration Protocol (DHCP) server to provide
P2P Clients with IP addresses. In addition, only
the P2P GO is allowed to cross-connect the
devices in its P2P Group to an external network
(e.g., a 3G network or an infrastructure WLAN
as shown in Fig. 1), and for this cross-connection
bridging is not allowed. Therefore, the connec-
tion must be done at the network layer, typically
Figure 1. Example of Wi-Fi direct supported topologies and use cases.
P2P client
P2P client
P2P clientLegacy IEEE 802.11 AP
P2P client P2P GO
3G interface P2P GO
Legacy client P2P GO
IEEE 802.11 WLAN
channel M
P2P group 3
channel N
P2P group 1
channel X
P2P group 2
channel Y
1This step might be
avoided in some cases as
explained below.
2Most P2P functionality
is deployed through the
P2P Information Element
included in management
frames and through novel
action frames. Legacy
devices ignore these infor-
mation elements and
action frames.
Like a traditional AP, a
P2P GO announces itself
through beacons, and
has to support power
saving services for its
associated clients. The
P2P GO is also required
to run a Dynamic Host
Configuration Protocol
(DHCP) server to pro-
vide P2P Clients with
IP addresses.
CAMPS-MUR LAYOUT_Layout 1 6/21/13 11:52 AM Page 97
IEEE Wireless Communications • June 2013
98
implemented using Network Address Transla-
tion (NAT).
Finally, Wi-Fi Direct does not allow transfer-
ring the role of P2P GO within a P2P Group. In
this way, if the P2P GO leaves the P2P Group
then the group is torn down, and has to be re-
established using some of the specified proce-
dures.
GROUP FORMATION
There are several ways in which two devices can
establish a P2P Group, depending on, e.g., if
they have to negotiate the role of P2P GO, or if
there is some pre-shared security information
available. Here we first describe the most com-
plex case, which we denote as the Standard case,
to afterwards highlight a couple of simplified
cases which we denote as the Autonomous and
Persistent cases. These three group formation
cases are illustrated in Fig. 2.
Standard
— In this case the P2P devices have
first to discover each other, and then negotiate
which device will act as P2P GO. Wi-Fi Direct
devices usually start by performing a traditional
Wi-Fi scan (active or passive), by means of
which they can discover existent P2P Groups3
and Wi-Fi networks. After this scan, a new Dis-
covery algorithm is executed, which we describe
next. First, a P2P Device selects one of the so-
called Social channels, namely channels 1, 6, or
11 in the 2.4 GHz band, as its Listen channel.
Then, it alternates between two states: a search
state, in which the device performs active scan-
ning by sending Probe Requests in each of the
social channels; and a listen state, in which the
device listens for Probe Requests in its listen
channel to respond with Probe Responses. The
amount of time that a P2P Device spends on
each state is randomly distributed, typically
between 100 ms and 300 ms, but it is up to the
each implementation to decide on the actual
mechanism to e.g. trade-off discovery time with
energy savings by interleaving sleeping cycles in
the discovery process. An example operation of
this discovery algorithm is illustrated in the first
part of Fig. 2.
Once the two P2P Devices have found each
Figure 2. Typical frame exchange sequences in the standard, autonomous and persistent group formation procedures.
*see below
Becomes group
owner (GO) on
selected
channel
(from all
supported)
Search
ch. 1,6,11
Search
ch. 1,6,11
IEEE 802.11
scan Phase 1 & 2
P2P standard
group formation
DHCP
discover
DHCP
request
DHCP
ACK
DHCP
offer
Beacon ...
GO
negotiation
response
GO
negotiation
confirm
Probe
response
GO
negotiation
request
Probe request
Probe request
Probe request
Probe request
Listen
ch. 11
Search
ch. 1,6,11
Search
ch. 1,6,11
Discovery
Listen
ch. 6
Listen
ch. 6
GO negotiation WPS provisioning Address config.
IEEE 802.11
scan
P2P autonomous
group formation
DHCP
request
DHCP
ACK
DHCP
offer
Address config.WPS provisioningDiscovery
Becomes group
owner (GO) on
selected
channel
(from all
supported)
Search
ch. 1,6,11
Search
ch. 1,6,11
IEEE 802.11
scan
Both devices have
already information
about a pre-created
persistent group
Phase 2
P2P persistent
group formation
DHCP
discover
DHCP
request
DHCP
ACK
DHCP
offer
Beacon ...
...
Probe
response
Invitation
request
Auth.
response
Auth.
request
Association
response
Association
request
EAP
request
ID
M1
EAP
init.
EAP
DONE
EAP
FAIL
Auth.
response
EAPOL
key (1/4)
Re-assoc.
response
EAPOL
key (3/4)
EAPOL
key (2/4)
Auth.
request
EAPOL
key (4/4)
Probe request
Probe request
Probe request
Probe request
Listen
ch. 11
Search
ch. 1,6,11
Search
ch. 1,6,11
Discovery
Listen
ch. 6
Listen
ch. 6
Phase 1 Phase 2
Invitation WPS provisioning
WPS
provisioning
Address config.
Creates a P2P
group and
becomes GO
Invitation
response
EAP
response
ID
EAPOL
start
EAP
request
ID
M2 M8
De-
auth.
Re-assoc.
request
Phase 1 & 2
DHCP
discover
3Passive scanning should
be used to discover a P2P
GO that is sleeping.
CAMPS-MUR LAYOUT_Layout 1 6/21/13 11:52 AM Page 98
IEEE Wireless Communications • June 2013 99
other, they start the GO Negotiation phase. This
is implemented using a three-way handshake,
namely GO Negotiation Request/Response/Confir-
mation, whereby the two devices agree on which
device will act as P2P GO and on the channel
where the group will operate, which can be in
the 2.4 GHz or 5 GHz bands. In order to agree
on the device that will act as P2P GO, P2P
devices send a numerical parameter, the GO
Intent value, within the three-way handshake,
and the device declaring the highest value
becomes the P2P GO. To prevent conflicts when
two devices declare the same GO Intent, a tie-
breaker bit is included in the GO Negotiation
Request, which is randomly set every time a GO
Negotiation Request is sent.
Once the devices have discovered each other
and agreed on the respective roles, the next
phase is the establishment of a secure communi-
cation using Wi-Fi Protected Setup, which we
denote as WPS Provisioning phase and describe
later, and finally a DHCP exchange to set up the
IP configuration (the Address config. phase in
the figure).
Autonomous
— A P2P Device may autonomously
create a P2P Group, where it immediately
becomes the P2P GO, by sitting on a channel
and starting to beacon. Other devices can discov-
er the established group using traditional scan-
ning mechanisms, and then directly proceed with
the WPS Provisioning and Address Configura-
tion phases. Compared to the previous case,
then, the Discovery phase is simplified in this
case as the device establishing the group does
not alternate between states, and indeed no GO
Negotiation phase is required. An exemplary
frame exchange for this case is illustrated in the
middle part of Fig. 2.
Persistent
— During the formation process, P2P
devices can declare a group as persistent, by
using a flag in the P2P Capabilities attribute pre-
sent in Beacon frames, Probe Responses and
GO negotiation frames. In this way, the devices
forming the group store network credentials and
the assigned P2P GO and Client roles for subse-
quent re-instantiations of the P2P group. Specifi-
cally, after the Discovery phase, if a P2P Device
recognizes to have formed a persistent group
with the corresponding peer in the past, any of
the two P2P devices can use the Invitation Proce-
dure (a two-way handshake) to quickly re-instan-
tiate the group. This is illustrated in the lower
part of Fig. 2, where the Standard case is
assumed as baseline, and the GO Negotiation
phase is replaced by the invitation exchange, and
the WPS Provisioning phase is significantly
reduced because the stored network credentials
can be reused.
LAYER TWO SERVICE DISCOVERY
A salient feature of Wi-Fi Direct is the ability
to support service discovery at the link layer.
In this way, prior to the establishment of a
P2P Group, P2P Devices can exchange queries
to discover the set of available services and,
based on this, decide whether to continue the
group formation or not. Notice that this repre-
sents a significant shift from traditional Wi-Fi
networks, where it is assumed that the only
service clients are interested in is Internet
connectivity.
In order to implement the above, service dis-
covery queries generated by a higher layer pro-
tocol, e.g., UPnP or Bonjour [11], are
transported at the link layer using the Generic
Advertisement Protocol (GAS) specified by
802.11u [12]. GAS is a layer two query/response
protocol implemented through the use of public
action frames, that allows two non-associated
802.11 devices to exchange queries belonging to
a higher layer protocol (e.g. a service discovery
protocol). GAS is implemented by means of a
generic container that provides fragmentation
and re-assembly, and allows the recipient device
to identify the higher layer protocol being trans-
ported. The interested reader is referred to [3]
for a detailed description.
SECURITY
Security provisioning starts after discovery has
taken place and, if required, the respective roles
have been negotiated. Wi-Fi Direct devices are
required to implement Wi-Fi Protected Setup
(WPS) [6] to support a secure connection with
minimal user intervention. In particular, WPS
allows to establish a secure connection by, e.g.,
introducing a PIN in the P2P Client, or pushing
a button in the two P2P Devices.
Following WPS terminology, the P2P GO is
required to implement an internal Registrar,
and the P2P Client is required to implement an
Enrollee [6]. The operation of WPS is com-
posed of two parts. In the first part, denoted as
“Phase 1” in the lower part of Fig. 2, the inter-
nal Registrar is in charge of generating and
issuing the network credentials, i.e., security
keys, to the Enrollee. WPS is based on WPA-2
security and uses Advanced Encryption Stan-
dard (AES)-CCMP as cypher, and a randomly
generated Pre-Shared Key (PSK) for mutual
authentication. In the second part, depicted as
“Phase 2” in Fig. 2, the Enrollee (P2P Client)
disassociates and reconnects using its new
authentication credentials. In this way, if two
devices already have the required network cre-
dentials (this is the case in the Persistent group
formation), there is no need to trigger the first
phase, and they can directly perform the
authentication.
POWER SAVING
Using Wi-Fi Direct, battery-constrained devices
may typically act as P2P GO (soft-AP), and
therefore energy efficiency is of capital impor-
tance. However, power saving mechanisms in
current Wi-Fi networks are not defined for APs
but only for clients. Notice that with Wi-Fi
Direct, a P2P Client can benefit from the exist-
ing Wi-Fi power saving protocols, i.e. legacy
power save mode [1] or U-APSD [5]. In order to
support energy savings for the AP, Wi-Fi Direct
defines two new power saving mechanisms: the
Opportunistic Power Save protocol and the Notice
of Absence (NoA) protocol.
Opportunistic Power Save
— The basic idea of
Opportunistic Power Save is to leverage the
sleeping periods of P2P Clients. The mecha-
Using Wi-Fi Direct, bat-
tery-constrained devices
may typically act as P2P
GO (soft-AP), and there-
fore energy efficiency is
of capital importance.
However, power saving
mechanisms in current
Wi-Fi networks are not
defined for APs but only
for clients.
CAMPS-MUR LAYOUT_Layout 1 6/21/13 11:52 AM Page 99
IEEE Wireless Communications • June 2013
100
nism assumes the existence of a legacy power
saving protocol, and works as follows. The P2P
GO advertises a time window, denoted as
CTWindow, within each Beacon and Probe
Response frames. This window specifies the
minimum amount of time after the reception of
a Beacon during which the P2P GO will stay
awake and therefore P2P Clients in power sav-
ing can send their frames. If after the CTWin-
dow the P2P GO determines that all connected
clients are in doze state, either because they
announced a switch to that state by sending a
frame with the Power Management (PM) bit
set to 1, or because they were already in the
doze state during the previous beacon interval,
the P2P GO can enter sleep mode until the
next Beacon is scheduled; otherwise, if a P2P
Client leaves the power saving mode mode
(which is announced by sending a frame with
the PM bit set to 0) the P2P GO is forced to
stay awake until all P2P Clients return to power
saving mode. Figure 3 provides an example of
the operation of the Opportunistic Power Save
protocol for a scenario consisting of one P2P
GO and one P2P Client.
Notice that, using this mechanism, a P2P GO
does not have the final decision on whether to
switch to sleep mode or not, as this depends on
the activity of the associated P2P Clients. To
give a P2P GO higher control on its own energy
consumption Wi-Fi Direct specifies the Notice of
Absence protocol, which is described next.
Notice of Absence
— The Notice of Absence (NoA)
protocol allows a P2P GO to announce time
intervals, referred to as absence periods, where
P2P Clients are not allowed to access the chan-
nel, regardless of whether they are in power save
or in active mode. In this way, a P2P GO can
autonomously decide to power down its radio to
save energy.
Like in the Opportunistic Power Save proto-
col, in the case of NoA the P2P GO defines
absence periods with a signaling element includ-
ed in Beacon frames and Probe Responses. In
particular, a P2P GO defines a NoA schedule
using four parameters:
Duration that specifies the length of each
absence period
Interval that specifies the time between consec-
utive absence periods
Start time that specifies the start time of the
first absence period after the current Beacon
frame
Count that specifies how many absence peri-
ods will be scheduled during the current NoA
schedule.4
A P2P GO can either cancel or update the cur-
rent NoA schedule at any time by respectively
omitting or modifying the signaling element. P2P
Clients always adhere to the most recently
received NoA schedule. Figure 3 depicts an
example operation of the NoA protocol.
In order to foster vendor differentiation, the
Wi-Fi Direct specification does not define any
mechanism to compute the CTWindow in the
Opportunistic Power Save protocol or the sched-
ule of absence periods in the Notice of Absence
protocol. Below we provide some experimental
results comparing the impact of different power
saving policies that can be used to configure the
NoA protocol.
Figure 3. P2P GO power saving protocols in Wi-Fi Direct: a) example of opportunistic power save; and b)
example of NoA.
FRAME
PM flag=0
FRAME
PM flag=1
P2P GO
P2P client
Beacon
CTWindow(n)
Beacon
CTWindow(n-1)
Beacon
CTWindow(n+1)
Sleeping
P2P GO
P2P client
SleepingSleepingSleeping
Beacon
NoA schedule 1
count = 2
SleepingSleepingSleeping
Sleeping
Start
time
SleepingSleeping SleepingSleepingSleeping
Sleeping
SleepingSleeping
(a)
(b)
CTWindow(n-1) CTWindow(n+1)CTWindow(n)
Absent period
Duration
Duration
Interval Interval
Duration
Start
time
Start
time
Beacon
NoA schedule 2
count = 3
Beacon
NoA schedule 3
count = 1
4If count is set to 255 the
current NoA schedule
runs permanently until it
is explicitly cancelled.
In order to foster vendor
differentiation, the Wi-Fi
Direct specification does
not define any mecha-
nism to compute the
CTWindow in the
Opportunistic Power
Save protocol or the
schedule of absence
periods in the Notice of
Absence protocol.
CAMPS-MUR LAYOUT_Layout 1 6/21/13 11:52 AM Page 100
IEEE Wireless Communications • June 2013 101
EXPERIMENTAL EVALUATION
In this section we report experimental results
using Wi-Fi Direct in a testbed composed of
commercial, off-the-shelf devices. Our goal is
two-fold: first, to analyze the time required to
form a P2P group for the three cases described
in the previous section; second, to quantify the
achievable performance trade-offs using Wi-Fi
Direct and compare them against legacy opera-
tion.
TESTBED SETUP
We deploy a testbed consisting of two nodes in
an office environment, where different WLANs
could be detected, and perform our experiments
during working hours to guarantee that our
numerical figures correspond to realistic settings.
Our nodes are two laptops equipped with an
802.11a/b/g D-Link PCMCIA card with an
Atheros chipset, running Linux and the
mac80211/ath5k driver.
We use an open source implementation of
Wi-Fi Direct [7], which builds on the widely
deployed wpa_supplicant software, and extend
it to implement the Notice of Absence (NoA)
protocol. We decided to implement NoA as this
mechanism provides the AP with the tightest
control and therefore should provide the maxi-
mum energy savings. Throughout our experi-
ments we always pre-provision the devices with a
WPS PIN, to support automatized execution.
GROUP FORMATION DELAYS
We first analyze the time needed to establish a
P2P Group for the three group formation cases
introduced earlier, namely Standard,
Autonomous, and Persistent. In order to gain
insight on the total time required to form a P2P
group, we measured two delays:
Discovery delay, this being the time required
for the two P2P devices to find each other
Formation delay, this being the time required
to agree on the roles, establish a secure com-
munication, and perform the DHCP exchange
With the above, the total delay is obtained by
adding the two previous delays. To compute
these delays, we process the log files provided by
wpa_supplicant to obtain the time instants
when a device
• Starts the discovery phase
• Discovers the other device
• Obtains an IP address from the P2P GO
We repeat each group establishment 250 times
and report the experimental cumulative distribu-
tion function (CDF) of the above delays in Fig.
4. In order to understand the behavior of the
experimental implementation, we also plot in the
figure using dashed lines the results from an
event-driven simulator which closely follows the
exchanges illustrated in Fig. 2.5
The CDF of the discovery delay (top subplot)
shows that the initial scans delay all procedures
by at least 3 seconds, and that the Standard and
Persistent cases behave very similarly, an expect-
ed result as they use the same discovery algo-
rithm. The figure also quantifies the randomness
introduced by the novel Wi-Fi Direct discovery
algorithm, which results in additional delays
between approximately one and seven seconds.
The Autonomous case, in contrast, shows an
approximately constant delay of three seconds,
which is caused by the implementation of its dis-
covery mechanism: the list of P2P GO candi-
dates is returned only after an active scanning in
all available channels has been completed (note
that our card operates in both the 2.4 and the 5
GHz band). The simulation results show, for the
Autonomous case, that there is room for signifi-
cant improvement if the procedure stops imme-
diately after the P2P GO of interest is
discovered; for the Standard and Persistent
cases, simulations provide the lower bound for
the discovery delay in case we were not experi-
encing interference and traffic from neighboring
WLANs (which is the reason for the differences
between the simulation and the experimental
results). Given the observed discovery times, we
conclude that continuously trying to discover
other Wi-Fi Direct devices in a battery con-
strained device may be a challenging task. For
this purpose, smart algorithms would be required
that interleave sleeping and discovery cycles in
order to trade-off discovery time and battery life.
We next analyze the CDF of the formation
delay (middle subplot), i.e., the time required to
negotiate roles (if needed), to establish a secure
communication and to provide the IP configura-
tion. It should be first noted that, for the case of
the Autonomous group formation, the open
source implementation we are using [7] triggers
an additional scan right before the WPS phase,
Figure 4. Evaluation of P2P Group Formation considering Standard, Aut-
nomous and Persistent group formation procedures.
Discovery
10
0.2
0
F(x)
0.4
0.6
0.8
1
2 3 4 5 6 7 8 910
Formation
0.50
0.2
0
F(x)
0.4
0.6
0.8
1
1 1.5 2 2.5 3 3.5 4 4.5 5
Total delay
Time (secs)
10
0.2
0
F(x)
0.4
0.6
0.8
1
2 3 4 5 6 7 8 910
P2P standard
P2P persistent
P2P autonomous
Simulation
5Note that our simulator
uses an optimized discov-
ery mechanism for the
autonomous group for-
mation, in which the pro-
cess stops as soon as the
GO is found.
CAMPS-MUR LAYOUT_Layout 1 6/21/13 11:52 AM Page 101
IEEE Wireless Communications • June 2013
102
something not specified in the standard. This
scan increased the delay by approximately three
seconds, thus we decided to remove its effect to
present a clear overview of the Wi-Fi Direct per-
formance. The figure shows that the three group
formations present similar behavior, with the dif-
ferences over simulations being caused by chan-
nel interference and the generation of
cryptographic information, and that the simpli-
fied signaling of the Persistent case reduces
delays in about half a second. It should also be
noted that despite the fact that the Autonomous
frame exchange is simpler than the Standard
one, in practice they result in almost the same
delay. This is because the GO Negotiation phase
takes very little time as compared to the WPS
provisioning phase, which requires a long frame
exchange that is affected by the interference
from other WLANs.
The addition of the above delays results in
the total delay depicted in Fig. 4, bottom. The
figure shows that 80 percent of the time the
Autonomous group formation took less than five
seconds to form the group, while for the Persis-
tent and Standard case this value is eight and
nine seconds, respectively, an order that follows
the relative complexity of the signaling proce-
dures illustrated in Fig. 2. It is also worth
remarking that, on average, there are a few sec-
onds of difference between simulations and the
use of our real-life testbed, caused by interfer-
ences from neighboring WLANs.
The above non-negligible delays prevent the
use of opportunistic P2P Group establishments
to support energy-efficient operation, e.g., set-
ting up a group when traffic is detected, and
powering down wireless interfaces when there is
no traffic. Instead, in order to amortize the over-
head of creating a P2P Group, we expect P2P
Groups to have a considerable life time. Thus, in
that case, energy efficient operation within a
P2P Group, for both P2P Client and P2P GO,
becomes a critical feature, and motivates the use
of the power saving mechanisms introduced ear-
lier.
ENERGY-EFFICIENT OPERATION
Many Wi-Fi Direct devices will run on batteries,
therefore energy efficiency is of paramount
importance. In order to quantify the achievable
energy savings using Wi-Fi Direct we consider
the case of tethering, i.e., a mobile device acting
as a soft-AP and sharing its 3G connection with
a client (Fig. 1). Note that in legacy Wi-Fi net-
works the AP is forced to remain active at all
times. In contrast, the Notice of Absence (NoA)
protocol lets a P2P GO announce sleep periods
and turn its interface(s) to doze in order to save
power. Thus, a key challenge with the use of
NoA is how to compute the length of the
absence periods in order to trade-off the energy
saving achieved in the P2P GO with the perfor-
mance of the traffic in the P2P Group. We ana-
lyze this question next.
In order to quantify the energy savings sup-
ported by Wi-Fi Direct, we fix the Beacon peri-
od to 100 ms and analyze three representative
policies to implement the NoA protocol:
•An Active policy where the P2P GO remains
always active, thus mimicking the legacy oper-
ation. Notice that this policy is optimal in
terms of traffic performance, but should result
in a worst case in terms of energy efficiency.
•A Static policy, where the P2P GO advertises a
fixed presence window of 25 ms right after
each Beacon frame.
•A Dynamic policy based on the Adaptive Sin-
gle Presence Period (ASPP) algorithm, which
adjusts the presence window based on the
estimated traffic activity (utilization) in the
channel. Specifically, ASPP estimates the time
the channel is busy using an exponentially
weighted moving average, and based on this
estimation uses a proportional controller to
drive the WLAN to the desired point of oper-
ation (for more details we refer the interested
reader to [13]).
To understand the impact of these policies,
we first equip one laptop with a 3G USB dongle
and configure it as P2P GO, running NoA with
the Dynamic policy; the other laptop is config-
ured as P2P Client and mimics a web-browsing
session during ten minutes. We measure the
resulting channel activity and the presence win-
dow advertised by the P2P GO, and depict both
in Fig. 5. We also mark in the figure the 25 ms
window that would be announced with the Static
policy; note that the Active operation would cor-
respond to a fixed setting of 100 ms. The figure
illustrates how an Active policy would disregard
any chance to save energy, despite the frequent
periods of time when there is little traffic. In
contrast, the use of a Static setting of 25 ms
would help to significantly reduce the energy
wastage; however, not only this policy requires
to manually tune the announced window for
each application, but also it would suffer from
performance issues due to the accidental traffic
bursts. Finally, the figure illustrates how the use
of the Dynamic policy is effective in adapting the
presence window to the traffic demand in the
network.
In order to quantitatively compare these poli-
cies in a fair manner, we substitute the 3G don-
gle by an NS3-based emulator [14], thus ensuring
Figure 5. ASPP presence window adaptation with Web traffic.
Time
1000
25
0
Presence window
Occupation [%]
50
75
100
25
0
50
75
100
200 300
ASPP
Occupation
Static-25
CAMPS-MUR LAYOUT_Layout 1 6/21/13 11:52 AM Page 102
IEEE Wireless Communications • June 2013 103
control over the 3G channel conditions and the
repeatability of the experiments. We implement-
ed three HSDPA models defined by the Eurane
project [15], namely a pedestrian channel with
good radio conditions, an urban channel with
average conditions, and a vehicular channel with
poor conditions. For each scenario and policy,
we measure the throughput obtained by a TCP
download using iperf, and the time that the
Wi-Fi interface of the P2P GO spends in the
active state and in the sleep state. To provide
absolute figures in terms of energy consumption,
we translate these times into Joules using the
power consumption characteristics of the
AR6001 Atheros chipset. In order to have fig-
ures with good statistical meaning, each experi-
ment is run for 100 seconds and repeated 5
times. The obtained results for throughput and
energy cost are depicted in Fig. 6, in which we
also add the results for the case of an intra-
group communication (i.e., without 3G channel),
namely, a TCP download from the P2P GO to
the P2P Client.
Figure 6 confirms the results observed in Fig.
5. For the case of 3G channels, the Active policy
achieves the highest throughput in all cases, but
at the highest energy cost, increasing this energy
cost as the available bandwidth decreases (e.g.,
Vehicular channel). In contrast, the Static policy
is able to achieve a good performance both in
terms of throughput and energy cost when the
bandwidth is relatively small, but the 25 ms win-
dow becomes a bottleneck when the achievable
bandwidth is large. This poor performance is
exacerbated in the intra-group communication,
where the throughput is less than one fourth of
the other mechanisms, this being caused by the
overly small presence window, while the energy
cost per unit of transferred information is the
highest, due to the non-negligible energy con-
sumption of the device during the sleep state.
Finally, the Dynamic policy is able to achieve
throughput figures very close to those of the
Active case, while maintaining a small energy
cost across all considered scenarios. For the case
of 3G channels, then, the energy costs of the
communication are reduced by up to 66 percent
as compared to legacy operation (Active case),
with almost no price paid in terms of perfor-
mance. Therefore we conclude that the use of
NoA together with an adaptive policy would sig-
nificantly extend the lifetime of battery-powered
devices using Wi-Fi Direct at no significant per-
formance penalty.
CONCLUSIONS
After a tremendous success whereby Wi-Fi has
become a predominant way to access the Inter-
net wirelessly, it is now embracing the challenge
of becoming pervasive also in direct device to
device communications. In this respect, the Wi-
Fi Alliance has recently developed the Wi-Fi
Direct technology that builds upon the Wi-Fi
infrastructure mode to enable direct device to
device connectivity.
In this article we presented a thorough
overview of the novel technical features specified
in Wi-Fi Direct, following by an experimental
evaluation that quantifies the group formation
delays to be expected in real-life scenarios.
Finally we have analyzed the performance-ener-
gy trade-offs of the novel NoA power saving
protocol defined in Wi-Fi Direct through exten-
sive experimentation. To the best of the authors’
knowledge this is the first article in the state of
the art that provides a wide overview and experi-
mental evaluation of Wi-Fi Direct.
Regarding future research directions. First,
the NoA protocol could also be re-used to virtu-
alize the roles of P2P GO/Client over multiple
concurrent P2P Groups. Second, concurrent
operation together with the dynamic nature of
the P2P GO/Client roles could be used to
improve performance in dense environments, for
instance by means of dynamic relays. Finally, if
Figure 6. Perfomance of the Dynamic, Active and Static power saving policies for different scenarios.
3G channel
Pedestrian Urban Vehicular
Intra-group
1
0
Throughput
(Mbps)
2
3
4
5
6
7
0
20
15
10
5
0
Energy cost
(Joule/MB)
1.5
2
1
0.5
0
0.25
0.2
0.3
0.15
0.1
0.05
Active
Dynamic
Static
We conclude that the
use of NoA together
with an adaptive policy
would significantly
extend the lifetime of
battery-powered devices
using Wi-Fi Direct
at no significant
performance penalty.
CAMPS-MUR LAYOUT_Layout 1 6/21/13 11:52 AM Page 103
IEEE Wireless Communications • June 2013
104
Wi-Fi Direct becomes a widespread technology
as expected, it faces the challenge of improving
coexistence and reducing interference with other
unlicensed devices.
ACKNOWLEDGMENT
The authors are grateful to the anonymous ref-
erees for their valuable comments which greatly
helped in improving the article.
REFERENCES
[1] IEEE 802.11-2007 Standard, Wireless LAN Medium
Access Control (MAC) and Physical Layer (PHY) Specifi-
cations, 2007.
[2] J.-S. Lee, Y.-W. Su, and C.-C. Shen, A Comparative
Study of Wireless Protocols: Bluetooth, UWB, ZigBee,
and Wi-Fi, Industrial Electronics Society, 2007, IECON
2007.
[3] Wi-Fi Alliance, P2P Technical Group, Wi-Fi Peer-to-Peer
(P2P) Technical Specification v1.0, Dec. 2009.
[4] IEEE 802.11z-2010 — Wireless LAN Medium Access
Control (MAC) and Physical Layer (PHY) specifications
Amendment 7: Extensions to Direct-Link Setup (DLS).
[5] Wi-Fi Alliance, Quality of Service (QoS) Task Group, Wi-
Fi Multimedia (including WMM PowerSave) Specifica-
tion v1.1, 2005.
[6] Wi-Fi Alliance, Wi-Fi Protected Setup Specification
v1.0h, Dec. 2006.
[7] Wi-Fi Direct in Linux, http://linuxwireless.org/en/develop-
ers/p2p/
[8] A. J. Nicholson et al., “Virtual Networks for Fun and
Profit,” IEEE Trans. Mobile Computing, vol. 9, no. 1,
Jan. 2010, pp. 31–43.
[9] Microsoft’s Virtual WiFi Homepage,
http://research.microsoft.com/en-us/projects/virtualwifi/
[10] L. Verma and S. S. Lee, “Proliferation of Wi-Fi: Oppor-
tunities in CE Ecosystem,” Proc. PerNets 2011, Las
Vegas, USA, Jan 9–12, 2011.
[11] W. K. Edwards, “Discovery Systems in Ubiquitous
Computing,” IEEE Pervasive Computing, vol. 5, no. 2,
Apr.–June 2006, pp. 70–77.
[12] 802.11u-2011 — Wireless LAN Medium Access Control
(MAC) and Physical Layer (PHY) specifications Amend-
ment 9: Interworking with External Networks.
[13] D. Camps-Mur, X. Perez-Costa, and S. Sallent-Ribes,
“Designing Energy Efficient Access Points with Wi-Fi
Direct,” Computer Networks, Sept. 2011.
[14] ns-3 simulator, http://www.nsnam.org.
[15] Enhanced UMTS Radio Access Network Extensions for
NS2 (EURANE), http://eurane.ti-wmc.nl/eurane.
BIOGRAPHIES
DANIEL CAMPS-MUR (Daniel.CampsMur@neclab.eu) is a
senior researcher in the Mobile and Wireless Networks
group at NEC Network Laboratories in Heidelberg, Ger-
many. He holds since 2004 a Masters degree and since
2012 a Ph.D. from the Polytechnic University of Catalonia
(UPC). His current research interests include QoS and ener-
gy efficiency in wireless networks. Daniel is also an active
contributor to the IEEE 802.11 group and to the Wi-Fi
Alliance. In addition, his master thesis work received the
Mobile Internet and 3G Mobile Solutions award from the
Spanish Association of Telecommunication Engineers
(COIT), and in the field of network simulations he also
received the OPNET Significant Technical Challenge Solved
Award.
ANDRES GARCIA-SAAVEDRA received his B.Sc. in Telecommuni-
cation Engineering from Universidad de Cantabria in 2009,
and his M.Sc. in Telematics Engineering from Universidad
Carlos III de Madrid (UC3M) in 2010. Currently he holds
the position of Teaching Assistant and pursues his Ph.D. in
the Telematics Department of UC3M. His work focuses on
performance evaluation and energy efficiency of wireless
networks.
PABLO SERRANO got his Telecommunication Engineering
degree and his PhD from the Universidad Carlos III de
Madrid (UC3M) in 2002 and 2006, respectively. He has
been with the Telematics Department of UC3M since 2002,
where he currently holds the position of Associate Profes-
sor. In 2007 he was a Visiting Researcher at the Computer
Network Research Group at Univ. of Massachusetts
Amherst partially supported by the Spanish Ministry of
Education under a José Castillejo grant. He has over 40 sci-
entific papers in peer-reviewed international journal and
conferences. He serves on the Editorial Board of IEEE Com-
munications Letters, and has served on the TPC of a num-
ber of conferences and workshops including IEEE Globecom
and IEEE INFOCOM. His current work focuses on perfor-
mance evaluation of wireless networks.
if Wi-Fi Direct
becomes a widespread
technology as expected,
it faces the challenge
of improving
coexistence and
reducing interference
with other unlicensed
devices.
CAMPS-MUR LAYOUT_Layout 1 6/21/13 11:52 AM Page 104
... whenever there is no data to be transmitted [2], [3]. Although the original NoA method has some success in reducing the energy consumption, the various studies have attempted to modify the NoA method in the different ways to enhance the energy efficiency because the existing power saving methods do not provide sufficient functions to dynamically cope with varying traffic load [4]. ...
... Wi-Fi Direct technology was designed by Wi-Fi Alliance to support the D2D communication without an AP [2]. Wi-Fi Direct devices discover each other by performing a conventional Wi-Fi scan to negotiate which device will be selected as a group owner (GO). ...
Article
Full-text available
Recent studies on energy efficiency and scheduling of power-saving mode have been considered as key technologies for reducing the energy consumption of device-to-device (D2D) communication. Wi-Fi Direct (P2P), one of the key protocols for D2D communication, defines the on-off power saving mechanic called the notice of absence (NoA) power-saving mode that can be applied to the multimedia video traffic. The on-off power saving mechanic enables the user to transmit or receive the real-time video frame during the awake interval in which the video frame rate should meet the requirement. When the user can wholly transmit one video frame before the end time of a required inter-frame interval, it can switch to the sleep mode to save the power consumption. However, the challenge remaining for the NoA method is the fixed length of awake/sleep interval, even if the traffic load is varied. Therefore, in this study, we proposed a reinforcement learning-based power saving (RLPS) method to enhance the performance of the notice of absence (NoA) power-saving mode in Wi-Fi direct with taking the multimedia video transmission and the network delay jitter into consideration. The proposed RLPS method enables the Wi-Fi direct device to dynamically estimate the length of awake interval for transmitting the future video frame in real-time. In addition, the Wi-Fi direct device may wake up too early before the arrival of the video frame, which is caused by the network delay jitter. Thus, the client device has to wait for receiving the video frame. To tackle this challenge, the proposed RLPS method enables the device to predict the start time of awake interval for the purpose of reducing the delay time for receiving the upcoming video frame. Results show that the proposed RLPS method outperforms the existing NoA power-saving mode in terms of the outage probability, energy consumption, and transmission delay of Wi-Fi Direct devices.
... The Bluetooth technology is known as the cable replacement technology which is widely used in headphones, mouse, and keyboards while Bluetooth Low Energy (BLE) is used in accessories for smartphone apps and internet-connected devices [85]- [87]. Wi-Fi is the most common communication technology in which used to build smart sensor ad-hoc networks [88], [89]. LTE (Long-Term Evolution) is a standard wireless communication technology for high-speed data transfer between mobile phones based on GSM/UMTS network technologies [90], [91]. ...
Article
Full-text available
Water distribution systems are one of the critical infrastructures and major assets of the water utility in a nation. The infrastructure of the distribution systems consists of resources, treatment plants, reservoirs, distribution lines, and consumers. A sustainable water distribution network management has to take care of accessibility, quality, quantity, and reliability of water. As water is becoming a depleting resource for the coming decades, the regulation and accounting of the water in terms of the above four parameters is a critical task. There have been many efforts towards the establishment of a monitoring and controlling framework, capable of automating various stages of the water distribution processes. The current trending technologies such as Information and Communication Technologies (ICT), Internet of Things (IoT), and Artificial Intelligence (AI) have the potential to track this spatially varying network to collect, process, and analyze the water distribution network attributes and events. In this work, we investigate the role and scope of the IoT technologies in different stages of the water distribution systems. Our survey covers the state-of-the-art monitoring and control systems for the water distribution networks, and the status of IoT architectures for water distribution networks. We explore the existing water distribution systems, providing the necessary background information on the current status. This work also presents an IoT Architecture for Intelligent Water Networks - IoTA4IWNet, for real-time monitoring and control of water distribution networks. We believe that to build a robust water distribution network, these components need to be designed and implemented effectively.
... There are several works done in order to improve the training procedure groups, especially at step (3) (Menegato et al. 2014;Chaki et al. 2015;Cherif et al. 2017;Jahed et al. 2016;Khan et al. 2016;Liu et al. 2016;Li et al. 2020;Shahin and Younis 2015). However, several empirical studies carried out to evaluate the Wi-Fi Direct technology on Android, show that the discovery of devices step takes more time in the group formation process presented above (Camps-Mur et al. 2013;Conti et al. 2013;Garcia-Saavedra and Serrano 2013). A long time to discovery and high latency have an impact on the quality of service, specifically those that have time transmission constraints. ...
Article
Full-text available
The main objective of this work is to propose concrete time reduction strategies for discovery of Wi-Fi Direct in Android. To achieve our goals, we perform a fairly general mathematical modeling of the discovery of devices using Poisson processes. Subsequently, under asymptotic invariance hypotheses of certain distributions, we derive formulas for the expected time to discovery. We provide sufficient condition for fast convergence to an invariant distribution and determine key decision parameters (jumps intensities) that minimize the average time to discovery. We also propose a predictive model for rapid evaluation of these optimal discovery parameters. Experimental tests in an emulator are also conducted to validate the theoretical results obtained. A comparative performance study is done with some optimization approaches from literature. Compared with existing methods, the improvement of the average time discovery we obtained with the proposed method is above 98.34%\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$98.34\%$$\end{document}.
... This implies specific procedures to ensure both the compliance with the user's privacy management and the secure communication of personal information. Specifically, the user's identifiers can be encrypted by the middleware and shared with other devices in proximity by using the available D2D technologies, such as Wi-Fi Direct (WFD) and Bluetooth LE (BLE) Services [31,32]. Upper layer applications do not need the access to this information, and the middleware can maintain only the encrypted identifier in order to guarantee pseudo-anonymization. ...
Article
Full-text available
Context modeling and recognition represent complex tasks that allow mobile and ubiquitous computing applications to adapt to the user’s situation. The real advantage of context-awareness in mobile environments mainly relies on the prompt system’s and applications’ reaction to context changes. Current solutions mainly focus on limited context information generally processed on centralized architectures, potentially exposing users’ personal data to privacy leakage, and missing personalization features. For these reasons on-device context modeling and recognition represent the current research trend in this area. Among the different information characterizing the user’s context in mobile environments, social interactions and visited locations remarkably contribute to the characterization of daily life scenarios. In this paper we propose a novel, unsupervised and lightweight approach to model the user’s social context and her locations based on ego networks directly on the user mobile device. Relying on this model, the system is able to extract high-level and semantic-rich context features from smartphone-embedded sensors data. Specifically, for the social context it exploits data related to both physical and cyber social interactions among users and their devices. As far as location context is concerned, we assume that it is more relevant to model the familiarity degree of a specific location for the user’s context than the raw location data, both in terms of GPS coordinates and proximity devices. We demonstrate the effectiveness of the proposed approach with 3 different sets of experiments by using 5 real-world datasets collected from a total of 956 personal mobile devices. Specifically, we assess the structure of the social and location ego networks, we provide a semantic evaluation of the proposed models and a complexity evaluation in terms of mobile computing performance. Finally, we demonstrate the relevance of the extracted features by showing the performance of 3 different machine learning algorithms to recognize daily-life situations, obtaining an improvement of 3% of AUROC, 9% of Precision, and 5% in terms of Recall with respect to use only features related to physical context.
... This implies specific procedures to ensure both the compliance with the user's privacy management and the secure communication of personal information. Specifically, the user's identifiers can be encrypted by the middleware and shared with other devices in proximity by using the available D2D technologies, such as Wi-Fi Direct (WFD) and Bluetooth LE (BLE) Services [31,32]. Upper layer applications do not need the access to this information, and the middleware can maintain only the encrypted identifier in order to guarantee pseudo-anonymization. ...
Preprint
Full-text available
Context modeling and recognition represent complex tasks that allow mobile and ubiquitous computing applications to adapt to the user's situation. Current solutions mainly focus on limited context information generally processed on centralized architectures, potentially exposing users' personal data to privacy leakage, and missing personalization features. For these reasons on-device context modeling and recognition represent the current research trend in this area. Among the different information characterizing the user's context in mobile environments, social interactions and visited locations remarkably contribute to the characterization of daily life scenarios. In this paper we propose a novel, unsupervised and lightweight approach to model the user's social context and her locations based on ego networks directly on the user mobile device. Relying on this model, the system is able to extract high-level and semantic-rich context features from smartphone-embedded sensors data. Specifically, for the social context it exploits data related to both physical and cyber social interactions among users and their devices. As far as location context is concerned, we assume that it is more relevant to model the familiarity degree of a specific location for the user's context than the raw location data, both in terms of GPS coordinates and proximity devices. By using 5 real-world datasets, we assess the structure of the social and location ego networks, we provide a semantic evaluation of the proposed models and a complexity evaluation in terms of mobile computing performance. Finally, we demonstrate the relevance of the extracted features by showing the performance of 3 machine learning algorithms to recognize daily-life situations, obtaining an improvement of 3% of AUROC, 9% of Precision, and 5% in terms of Recall with respect to use only features related to physical context.
Chapter
An actual problem on an IoT adaptive systems is to manage user preferences and local actuators specifications. This paper uses a multi agent system to achieve a Smart Environment System, that supports interaction between persons and physical spaces, that users smartly adapt to their preferences in a transparent way. This work proposes also a set of security customization’s to secure the actuators and users on space, that has been developed using a multi agent system architecture with different features to achieve a solution that achieve all the proposed objectives.
Preprint
Full-text available
This paper describes a decentralized low-cost system designed to reinforce personal security in big events in case of emergency. The proposal consists of using smart contracts supported by blockchain in the management of events. An alternative communication channel that does not require any cloud service is also provided with the aim of improving the coordination of emergency services. Peers may use this emergency support tool to interact with each other through a chat when additional support is required. Since information security is mandatory in this scenario, Identity-Based Signcryption schemes are here used in order to guarantee communication confidentiality, authenticity and integrity. Depending on the communication mode (peer-to-peer or broadcast), different signcryption methods are used. A first implementation of the proposal has produced promising results.
Chapter
Changes and additions to IEEE Std. 802.11-1999 are provided to support the new highrate physical layer (PHY) for operation in the 5 GHz band.
Article
There are many situations in which an additional network interface-or two-can provide benefits to a mobile user. Additional interfaces can support parallelism in network flows, improve handoff times, and provide sideband communication with nearby peers. Unfortunately, such benefits are outweighed by the added costs of an additional physical interface. Instead, virtual interfaces have been proposed as the solution, multiplexing a single physical interface across more than one communication endpoint. However, the switching time of existing implementations is too high for some potential applications, and the benefits of this approach to real applications are not yet clear. This paper directly addresses these two shortcomings. It describes a link-layer implementation of a virtual 802.11 networking layer, called Juggler, that achieves switching times of approximately 3 ms, and less than 400 mus in certain conditions. We demonstrate the performance of this implementation on three application scenarios. By devoting 10 percent of the duty cycle to background tasks, Juggler can provide nearly instantaneous handoff between base stations or support a modest sideband channel with peer nodes, without adversely affecting foreground throughput. Furthermore, when the client issues concurrent network flows, Juggler is able to assign these flows across more than one AP, providing significant speedup when wired-side bandwidth from the AP constrains end-to-end performance.
Conference Paper
Bluetooth (over IEEE 802.15.1), ultra-wideband (UWB, over IEEE 802.15.3), ZigBee (over IEEE 802.15.4), and Wi-Fi (over IEEE 802.11) are four protocol standards for short- range wireless communications with low power consumption. From an application point of view, bluetooth is intended for a cordless mouse, keyboard, and hands-free headset, UWB is oriented to high-bandwidth multimedia links, ZigBee is designed for reliable wirelessly networked monitoring and control networks, while Wi-Fi is directed at computer-to-computer connections as an extension or substitution of cabled networks. In this paper, we provide a study of these popular wireless communication standards, evaluating their main features and behaviors in terms of various metrics, including the transmission time, data coding efficiency, complexity, and power consumption. It is believed that the comparison presented in this paper would benefit application engineers in selecting an appropriate protocol.
Conference Paper
With the rapid proliferation of bandwidth intensive content, a newer technology is essential as 802.11 a/b/g cannot satisfy the requirements, like higher bandwidth. The latest generation, 802.11n, is the answer and helps accelerate the uptake of Wi-Fi integration in CE (Consumer Electronic) devices. Nowadays, most of CE devices are Wi-Fi-enabled and the unprecedented growth causes a dearth of bandwidth for Wi-Fi operation. We present the various standardization activities in progress by industry alliances and international standard organizations to assuage the bandwidth paucity problem by defining protocols for Wi-Fi operation in the available license-exempt spectrum in VHF/UHF (Very High Frequency/Ultra High Frequency) and 60 GHz band. The expansion of Wi-Fi to CE coupled with the new standardization initiatives creates new usage models like P2P (Peer-to-Peer) communication, Wireless Display, Smart Grid applications, etc., that are also articulated in this work.
Article
Wi-Fi Direct is a new technology defined by the Wi-Fi Alliance in order to enable efficient device-to-device communication. Portable devices are a main target of this technology and hence, power saving is a key objective. In this paper we analyze the power saving protocols defined in Wi-Fi Direct and design two algorithms to efficiently use them: Adaptive Single Presence Period (ASPP) and Adaptive Multiple Presence Periods (AMPP). Operating only with layer two information, our proposed algorithms allow Wi-Fi Direct mobile devices to address the trade-off between performance and energy consumption in a configurable manner when offering access to an external network. The two algorithms presented, ASPP and AMPP, are evaluated by OPNET simulations.
Article
Ubiquitous computing introduces unique requirements for discovery technologies, which let services and devices become aware of each other without explicit human administration. Discovery lets services and devices spontaneously become aware of the availability and capability of peers on the network without explicit administration. In practice, this means that a client can discover and potentially use a device without prior knowledge of it. Although discovery is a necessary component of ubiquitous computing, the wide range of discovery systems in use today reflects the varied needs of the communities from which they originated. Some of these features are appropriate for ubiquitous computing, but others require further research and development.
P2P Technical Group, Wi-Fi Peer-to-Peer (P2P) Technical Specification v1
  • Wi-Fi
  • Alliance
Wi-Fi Alliance, P2P Technical Group, Wi-Fi Peer-to-Peer (P2P) Technical Specification v1.0, Dec. 2009.
Quality of Service (QoS) Task Group, Wi-Fi Multimedia (including WMM PowerSave) Specifica-tion v1
  • Wi-Fi
  • Alliance
Wi-Fi Alliance, Quality of Service (QoS) Task Group, Wi-Fi Multimedia (including WMM PowerSave) Specifica-tion v1.1, 2005.
Wi-Fi Protected Setup Specification v1.0h
  • Wi-Fi Alliance
Wi-Fi Alliance, Wi-Fi Protected Setup Specification v1.0h, Dec. 2006.
Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifi-cations
  • Standard