Assisted network discovery for next generation wireless networks

Abstract

IEEE 802.11 networks are the most popular option to have wireless access to the Internet. The popularity of these networks have raised a costly topology discovery and connection process, in which any device has to pass through an expensive scanning process of available Access Points. In order to improve the connection process, we propose a novel architecture for asynchronous assistance for topology discovery. We discuss the role of a Topology Manager that uses computational intelligence for generating optimal scanning sequences. Preliminary results show that this approach results in 30% to 70% improvements on AP discovery rate in chaotic deployments.

Full text

Assisted network discovery for next generation
wireless networks
Andr´
es Arcia-Moret∗, Arjuna Sathiaseelan∗, Antonio Araujo†, Jos´
e Aguilar‡, Laudin Molina§
∗Computer Laboratory, University of Cambridge, Cambridge, UK
†Centro Nacional de Desarrollo e Investigaci´
on en Tecnolog´
ıas Libres, M´
erida, Venezuela
‡Universidad de Los Andes, M´
erida, Venezuela
§Institut Telecom / Telecom Bretagne, Universit´
e Europ´
eene de Bretagne, Rennes, France
Email: {andres.arcia, arjuna.sathiaseelan}@cl.cam.ac.uk, aaraujo@cenditel.gob.ve, aguilar@ula.ve, laudin.molina@telecom-bretagne.eu
Abstract—IEEE 802.11 networks are the most popular option
to have wireless access to the Internet. The popularity of these
networks have raised a costly topology discovery and connection
process, in which any device has to pass through an expensive
scanning process of available Access Points. In order to improve
the connection process, we propose a novel architecture for
asynchronous assistance for topology discovery. We discuss the
role of a Topology Manager that uses computational intelligence
for generating optimal scanning sequences. Preliminary results
show that this approach results in 30% to 70% improvements
on AP discovery rate in chaotic deployments.
I. INTRODUCTION
Today, IEEE 802.11 networks are the first option to have
ubiquitous and low-cost wireless access to the Internet [1].
We count on millions of Wi-Fi devices in many different
new and challenging contexts: Community Networks, Sensor
Networks, Internet of Things, and personal computers in Home
Networks. In any of these scenarios, there has been a dramatic
increase on the number of devices requiring to connect to Wi-
Fi Access Points (AP). In order to connect to the wireless
access network, a device must scan its surrounding and find
an appropriate AP. However, the scanning process generally
embedded in mobile and desktop devices follow simple and
greedy sequential scanning.
In Wi-Fi spontaneous deployments, people chaotically de-
ploy APs. Nomadic users may also be able to access thousands
of community APs belonging to the same direct provider, or
have access to virtual service providers such as PAWS1. In
both cases a user has to pass through a scanning process,
being the most expensive sub-process within the discovery of
the wireless topology. In this respect, the discovery process is
becoming iterative and time consuming in dense deployments.
Recently, we have observed in [2] that in a regular discovery
process, the device has to scan multiple times to discover
available APs in a densely-covered urban area. World-wide
trends show that this is likely to be a regular case in the near
future2.
The current suboptimal behaviour of the scanning algo-
rithm is present in the vast majority of devices, significantly
consuming energy and impacting network performance [3].
It is also well-known that the full scanning is a default
scanning strategy implemented on mobile devices, and that
1http://www.cl.cam.ac.uk/ as2330/paws.html
2https://wigle.net/
the role of the scanning process represents about 80% of the
handover time. So, an efficient scanning will not only represent
an improvement on aggregated control traffic reduction for
public Wi-Fi access [3], but also could be an alternative for
load-balancing. The scanning traffic is becoming a potential
problem lowering the speed of the network and frequently
interrupting regular transmissions. This is mostly because of
the increasing number of devices using WiFi and congestion
induced by non-adapted scanning process [2].
Scanning process in 802.11 networks. The scanning is
the first sub-process for a client willing to attach to an IEEE
802.11 network, in which the interface looks for available APs
for later associate to them. Although the ultimate goal of a
scanning is to find all available APs to which the station might
be able to join, it is very costly in terms of aggregated number
of beacons and energy consumption at the client. To discover
all APs, i.e., the topology within an area, the device should be
properly adjusted with a pair of timers, that mainly determine
the efficiency of the discovery process [4]. As there are 11
channels and 2 timers per channel, there is a high number
of possibilities for timer configuration, i.e., 22 configurable
timers in total. Adapting these timers could depend not only on
the topology, but also on the application requirements. As the
computational complexity of finding an appropriate scanning
sequence increases, we propose the use of a computational
intelligent technique for improving the scanning process.
Using computational intelligence for assisting Wi-Fi
scanning. We propose a Topology Manager that uses a com-
putational intelligence (CI) technique to calculate efficient
scanning sequences that improve the Wi-Fi discovery process.
As suggested in [2], in future community Wi-Fi deployments,
these central entities or topology managers can help wireless
clients in the decision process for efficiently connecting to a
dense Wi-Fi network, and they could also serve for coarse
content or service indication by means of the 802.11u access
network query protocol.
II. A TOPOLOGY MANAGE R FO R IEEE 802.11
NE TWORKS
A Topology Manager (TM) conveniently hosted by the
wireless service provider, could opportunistically assist mobile
users to better discover and control a crowded wireless network
topology. It could also help mobile users to determine link
quality and best possible connection allowing improved access
to the network. As we have observed, within a single scanning,

a subset of the available APs are discovered, and usually, a
client does not have time to scan multiple times [2]. However,
an intelligent TM compute up-to-date and customized efficient
sequences, thus giving a more precise view of the topology.
The main advantage of this approach is that it allow clients
saving time during the expensive discovery process.
Topology
Manager
raw scan
Topology
aproximate topology
feeder
partial vision
regular client
Intelligent
Algorithm
efficient scan
required topology
local interface
external interface
yet another partial vision
optional message
intelligent sequence
F
C
Smart
Sequences
DB
raw
topology
model
Figure 1. Architecture for the Next Generation Wireless
In order to update the vision of the topology on the TM, we
rely upon two roles for participants. Firstly, new users would
act as feeders, as they push their vision of the topology from
time to time with a simple posts of their partial vision of the
network on the TM (through regular suboptimal scannings).
This update could be pushed through the IEEE 802.11u amend-
ment also known as Access Network Query Protocol (ANQP).
ANQP allows clients to query or to pass information to the TM
behind designated APs and before authentication. Moreover,
a client could use specific messages for finding out about a
specific mobile operator whose network is accessible through
the designated AP. Secondly, regular clients (RC) correspond
to those feeders that have already contributed enough to the
vision of the topology and that the TM has promoted to
RC. Hence, RCs looking for a candidate AP could retrieve
efficient sequences with special queries to the TM who, in
turn, asynchronously interacts with an intelligent algorithm.
Specific use-cases in which an RC benefits from this scheme,
correspond to handovers or when looking for a good candidate
AP. As shown in Fig. 1, a simple round trip between the RC
and the TM, or even by retrieving the appropriate scanning
sequence through an alternative interface (e.g., 3G), could
save several hundred of milliseconds through the increased
efficiency of the scanning.
Finally, Fig. 2 shows the comparison of various improved
sequences obtained from the TM. The discovery performance
was obtained from an emulated 802.11 spontaneously deployed
topology of about 2.5km and 1600 APs. Dots over the baseline
(within the dark grey area) show improved discovery rates
Ch1 Ch2 Ch3 Ch4 Ch5 Ch6 Ch7 Ch8 Ch9 Ch10 Ch11
Channels
0
0.1
0.2
0.3
rate (AP/s)
Baseline Sequence Rates
Efficient Sequence Rates
Figure 2. Comparison rates for the Cultural Algorithm efficient sequences
in APs per time unit, announced by the TM, and based on
data collected by a mobile client. Originally, the algorithm
residing at the TM, was fed with suboptimal sequences and
afterwards, an algorithm using computational intelligence de-
rived a variety of sequences. All sequences have considerably
higher discovery rate compared to the reference. Moreover,
delay improvements range from 30% to 70% with respect to
the reference.
III. CONCLUSION AND FUTURE WORK
In this work we have shown that the scanning process can
be significantly improved in chaotic Wi-Fi deployments by
using assisted network discovery. We have presented the design
of a Topology Manager and discussed the interaction with
mobile users, and we have shown the benefits for improving the
discovery process. Firstly, we propose different roles for users
contributing to build a vision of the chaotic Wi-Fi deployment.
Secondly, we propose the use of computational intelligence
to asynchronously calculate scanning sequences for improving
the overall discovery process.
ACKNOWLEDGMENT
The research leading to these results has received funding
from the European Union's (EU) Horizon 2020 research and
innovation programme under grant agreement No. 644663.
Action full title: architectuRe for an Internet For Everybody,
Action Acronym: RIFE. We would also like to thank Juan M.
Tirado for comments on an earlier draft.
REFERENCES
[1] J. Saldana, A. Arcia-Moret, B. Braem, E. Pietrosemoli, A. Sathiaseelan,
and M. Zennaro, “Alternative Network Deployments. Taxonomy, charac-
terization, technologies and architectures.” Internet Draft, GAIA - IETF,
July 2015.
[2] A. Arcia-Moret, L. Molina, N. Montavont, G. Castignani, and A. Blanc,
“Access Point Discovery in 802.11 Networks,” in IEEE WD, 2014.
[3] X. Hu, L. Song, D. V. Bruggen, and A. Striegel, “Is there wifi
yet? how aggressive wifi probe requests deteriorate energy and
throughput,” CoRR, vol. abs/1502.01222, 2015. [Online]. Available:
http://arxiv.org/abs/1502.01222
[4] G. Castignani, A. Arcia, and N. Montavont, “A study of the discovery
process in 802.11 networks,” SIGMOBILE Mob. Comput. Commun.
Rev., vol. 15, no. 1, pp. 25–36, Mar. 2011. [Online]. Available:
http://doi.acm.org/10.1145/1978622.1978626