01-01-2013, 02:06 PM
Achieving capacity fairness for wireless mesh networks
Achieving capacity fairness for wireless.pdf (Size: 984.52 KB / Downloads: 22)
Introduction
In this paper, we consider a two-tiered wireless mesh
network that consists of a number of user nodes (UNs)
and multiple mesh routers (MRs). UNs are deployed
at strategic positions, transmitting and receiving information
to/from the nearby MR via a single hop (see
Figure 1). Such a two-tiered architecture has been used
in many wireless communication systems. In disaster
area, the transceivers carried by the first responders
play the role of UNs while the vehicles with powerful
antennas near those UNs work as MRs to facilitate
communication between first responders and the command
centre. As another example, powerful receivers
are set near humidity sensors on farms to gather information
for controlling irrigation systems. As can be
seen, both examples share a similar scenario: UNs are
placed at fixed positions in the sensing field and sending
information to nearby MRs. In our proposed network
scenarios, UNs are distributed on various spots to
collect or receive information, thus their positions cannot
be altered in the sensing field. MRs, a higher-level
nodes forming the backbone network, are responsible
for communicating with UNs.
Multi-tiered Wireless Networks
Many researchers have proposed their network models
with a multi-tiered architecture, most of which set the
limit of tiers as 2. The second tier of network nodes are
introduced due to various reasons. For instance, In [1],
additional relay nodes, which are identical to a regular
node but without sensor unit, are disseminated into a
dense wireless sensor network. Joint optimization of
relay deployment and power control is performed for
lifetime elongation. A heuristic algorithm is presented
in [2] for energy provisioning and relay node placement
in wireless sensor networks to elongate the network
lifetime. In [3], relay points with access to power supply
are strategically placed to improve the throughput
of wireless LAN.Aniterative procedure is developed to
compute the best placement of a fixed number of relay
points. Simulation results for the 802.11-like system
model showsignificant performance gain through optimal
placement. For our work, MRs can connect to each
other without distance and interference constraints.
At least one of the MRs can access to the backbone
network.
Channel Assignment
A number of channel assignment schemes have been
proposed in recent years. In [11], Chin et al. address the
problem of dynamically assigning channels in ad hoc
wireless networks via power control in order to satisfy
their minimum QoS requirements. The objective
then is to maximize the number of co-channel links
subject to some stability conditions. In [12], a clusterbased
multipath topology control and channel assignment
scheme is proposed, which explicitly creates a
separation between the channel assignment and topology
control functions, thus minimizes flowdisruptions.
In [13], Raniwala et al. propose a greedy load-aware
channel assignment scheme when network nodes are
with multiple radios. The goal of channel assignment
is to bind each network interface to a radio channel
such that the available bandwidth on each link is proportional
to its expected load. In [14], Alicherry et al.
mathematically formulate the joint channel assignment
and routing problem, taking into account the interference
constraints, the number of channels in the network
and the number of radios available at each MR. A centralized
algorithm is developed to solve the problem
to yield the optimized network throughput. The channel
assignment algorithm is used to adjust the flow on
the flow graph to keep the increase of interference for
each channel to a minimum.
Max–min Fairness
In the area of Operations Research, the max–min
problem has been extensively studied. In [16], Tang
develops a nonsimplex-based algorithm that finds an
optimal solution to a max–min allocation problem
with nonnegative integer variables. Fairness has been
well studied in both network layer and MAC layer.
Recently, in [17], Hou et al. develop an elegant
polynomial time algorithm to calculate the rate
allocation under a network lifetime constraint with
respect to a two-tiered wireless sensor network.
In [6], a Linear Programming (LP) formulation is
provided to solve the max–min guaranteed maximum
throughput bandwidth allocation problem. In
addition, the Lexicographical Max–Min Bandwidth
Allocation (LMMBA) problem is solved by a polynomial
time optimal algorithm. In [18], Liang et al.
investigate resource allocation for fading relay channels
under separate power constraints, which falls
into max–min problems. However, it is studied within
a context of three-terminal networks. Few research
efforts truly address the max–min fairness problem
of link throughput via channel assignment and power
control.
Simulation Results
In this section, we present simulation results on max–
min capacity values yielded by BIPA and the corresponding
upper bound. In particular, we compare the
fairness performance yielded by BIPA and the upper
bound by proposing a new evaluation metric named
performance ratio.We produce performance ratios under
different parameter sets to investigate the impact
of different factors on the fairness performance, such
as the number of channels, the number of power levels
and the interference threshold.