International Journal of Computer Applications (0975 8887)
Volume 94 - No. 19, May 2014
TDMA MAC Protocols for WiFi-based Long Distance
Networks: A Survey
I. Hussain
Department of IT
North-Eastern Hill University
Shillong-793022, India
N. Sarma
Department of CSE
Tezpur University
Tezpur-784028, India
D. K. Saikia
Department of CSE
National Institute of Technology
Shillong-793003, India
ABSTRACT
WiFi is widely being researched to use it beyond its original pur-
pose of short-range communication. These kinds of efforts are stim-
ulated due to its mass production and cost-effectiveness. WiFi-
based Long Distance (WiLD) networks envisage changing the ru-
ral communication scenario particularly in the undeveloped na-
tions. However, CSMA/CA, the de-facto MAC protocol of WiFi,
is found to be ill-suited for WiLD networks. As a result, various
TDMA-based MAC protocols are proposed in the literature as an
alternative to CSMA/CA. The proposed protocols have shown sig-
nificant performance improvement in such scenarios. In this pa-
per, we propose a classification framework of WiLD MAC proto-
cols based on their major characteristics. Using the classification
framework, a survey on the MAC protocols proposed for WiLD
networks is presented. The distinctive features of the TDMA MAC
protocols are critically examined by pointing out their strengths
and weaknesses in WiLD environment. In addition, this paper puts
forward some state of art open research challenges in this area.
General Terms:
Rural Network, MAC Layer, Wireless Communication
Keywords:
CSMA/CA, MAC Protocol, TDMA, WiFi, WiLD Network
1. INTRODUCTION
In recent times, WiFi has become very popular in providing Inter-
net connectivity to the remote underserved areas using long dis-
tance links[14]. However, WiFi was originally designed to support
Wireless Local Area Networks (WLANs) for short-range commu-
nication. The widespread standard for WLANs, IEEE 802.11b/g
operating in the 2.4 GHz frequency band, is one of the most popular
wireless standards. The license-free operations in the ISM band and
a variety of available low cost IEEE 802.11 hardware commodities
make WiFi an attractive and economically feasible communication
alternative for rural use[21]. Various research outcomes have also
established the viability of WiFi as a practical solution for long dis-
tance communication.
WiLD networks are usually comprised of long distance WiFi links
enabled by high-gain directional antenna covering up to several
tens of kilometers[3]. A typical WiLD network architecture for ru-
ral region is explained in the following section. Many WiLD net-
works including research test beds are deployed in different cor-
ners of the world. Few notable real life WiLD deployments in-
clude Digital Gangetic Plains[1] in Uttar Pradesh, India, Aravind
Network for Telemedicine[22] in Tamil Nadu, India, Long Dis-
tance Network in Amazonian Jungle of Peru[18] for telemedicine
and telephony, and Akshaya Network[13] for e-governance in Ker-
ala, India. Several research test beds have also been set up in re-
cent times. MIT’s Roofnet[4], QuRiNet[24] at the Quail Ridge Re-
serve in Napa County, California, FRACTEL[5] at IIT Bombay,
India, and VillageNet[7] are some of the important WiLD net-
work research test beds which are working towards network per-
formance enhancement, providing support for various envisaged
applications particularly real-time applications such as e-learning,
e-governance, telemedicine, and telephony.
The feasibility of WiLD networks comes with a bunch of chal-
lenges. Long distance wireless links are highly unreliable due to
the factors such as signal fading, and interference, which limits the
overall network performance. Furthermore, the multi-hop nature of
WiLD networks greatly impacts the end-to-end throughput and de-
lay. Some of the potential objectives of WiLD network research are
to further improve network performance in long distance communi-
cation, specifically to increase network throughput, reduce latency
by improving spectrum usage.
Medium Access Control (MAC) protocol plays a key role in opti-
mally utilizing the shared transmission medium which directly im-
pacts on overall network performance. It solves the main sources of
energy waste problems such as collision, idle listening, overhear-
ing, and control packet overhead. Schedule-based protocols like
Time Division Multiple Access (TDMA) provide good solutions
to traffic correlation, heavy collision, contention and overhearing
problems which usually occurs in contention-based channel access
protocols. Many TDMA-based MAC protocols for WiLD networks
have been proposed in the literature addressing issues pertaining to
performance enhancement in long distance communication contin-
uing the use of commodity 802.11 hardware. However, issues like
single point of failure, interference-aware scheduling, congestion
avoidance, QoS provisioning, multi-hop support, network reliabil-
ity, etc., still remain unaddressed or partially addressed.
In this paper, we first define a classification framework to catego-
rize different protocols based on factors like network control, syn-
1