The Use of Orthogonal Frequency Code
The Use of Orthogonal Frequency Code The Use of Orthogonal Frequency Code
The Use of Orthogonal Frequency Code
Division (OFCD) Multiplexing in
Division (OFCD) Multiplexing in Division (OFCD) Multiplexing in
Division (OFCD) Multiplexing in
Wireless Mesh Network (WMN)
Wireless Mesh Network (WMN)Wireless Mesh Network (WMN)
Wireless Mesh Network (WMN)
Syed S. Rizvi
, Khaled M. Elleithy
, and Aasia Riasat
Department of Computer Science and Engineering, University of Bridgeport, Bridgeport, CT
Computer Science Department, Institute of Business Management, Karachi, Pakistan
In the present scenario, improvement in the data rate, network capacity, scalability, and the network
throughput are some of the most serious issues in wireless mesh networks (WMN). Specifically, a major
obstacle that hinders the widespread adoption of WMN is the severe limits on throughput and the network
capacity. This chapter presents a discussion on the potential use of a combined orthogonal-frequency
code-division (OFCD) multiple access scheme in a WMN. The OFCD is the combination of orthogonal
frequency division multiplexing (OFDM) and the code division multiple access (CDMA). Since ODFM is
one of the popular multi-access schemes that provide high data rates, combing the OFDM with the
CDMA may yield a significant improvement in a WMN in terms of a comparatively high network
throughput with the least error ration. However, these benefits demand for more sophisticated design of
transmitter and receiver for WMN that can use OFCD as an underlying multiple access scheme. In order
to demonstrate the potential use of OFCD scheme with the WMN, this chapter presents a new transmitter
and receiver model along with a comprehensive discussion on the performance of WMN under the new
OFCD multiple access scheme. The purpose of this analysis and experimental verification is to observe
the performance of new transceiver with the OFCD scheme in WMN with respect to the overall network
throughput, bit error rate (BER) performance, and network capacity. Moreover, in this chapter, we
provide an analysis and comparison of different multiple access schemes such as FDMA, TDMA,
CDMA, OFDM, and the new OFCD.
Keywords— Wireless mesh networks, orthogonal frequency code division (OFCD), Code division
multiple access (CDMA), Multiple access, Multiple access interference (MAI), Network throughput, Bit
error rate (BER), Intersymbol interference (ISI).
Wireless Mesh Networks (WMNs) have become a major paradigm for construing a user access network
that provides high speed network access to users in the context of enterprise and community networks
(Akyildiz, Wang, & Wang, 2005; Tse & Grossglauser, 2002; Xiang, Peng-Jun, Wen-Zhan, & Yanwei,
2008). From the robust and stable connectivity perspective, WMN is considered as one of the most data
efficient mesh topology which is constructed by, mesh routers, clients, and gateways (Kim & Bambos,
2002; Zhang, Honglin, & Chen, 2008). In WMN, mesh routers can play the roll of gateways whereas the
clients provide connectivity with the Internet. One of the main advantages of WMN is the self-healing
and self-configuring nature of mesh routers. For extending the geographical area of a network, mesh
routers are equipped to provide connectivity between different networking technologies such as Wi-Fi,
IEEE 802.11, mobile technology and wired Ethernets (Li, Qiu, Zhang, Mahajan, Zhong, Deshpande, &
Rozner, 2007; Yu, Mohapatra, & Liu, 2008).
WMN itself brings many challenging issues from physical layer to application layer. Problems like
network capacity, protocols used in different layers, network management and the network security are
just some of the problems to point out. Recent theoretical studies and experimental verifications (Gupta &
Kumar, 2001) have shown the current WMNs are severely limited in network capacity. This is due to the
fact that when all nodes communicate using a single channel in a high speed wireless LAN (e.g., IEEE
802.11a), the number of simultaneous transmissions from multiple users is limited by interference. Since
WMNs are multi-hop in their nature, interference causes a serious degradation in overall network capacity
when adjacent hops on the same path start interfering with the neighboring paths (Kyasanur & Vaidya,
As of now, the scalability issue in WMN has not been fully solved yet (Xiang, Peng-Jun, Wen-Zhan, &
Yanwei, 2008). Most of the existing multiple access schemes which are based on CSMA/CA solve only
partial problems of the overall issue for WMN (Zhou & Lai, 2005). The implementation of such schemes
raises other performance issues such as minimum network capacity, low end-to-end throughput, and
scalability (Lin & Rasool, 2007). Thus, how to fundamentally improve the scalability and maximize the
throughput performance in WMN is an interesting research issue now days. One of the efficient solutions
that can be used for not only improving the network scalability but also maximizing the end-to-end
network throughput is the use of a hybrid multiple access scheme for WMN (Sundaresan & Rangarajan,
2008). For networks based on techniques other than CSMA/CA, code division multiple access (CDMA)
can be applied with orthogonal frequency division multiplexing (OFDM) in WMN as an efficient
multiple access scheme to overcome some of the problems mentioned above. The combination of these
two multiple access schemes allow us to take advantage of both CDMA and OFDM. OFDM has become
widely adopted in many next generation cellular systems such as 3GPP Long Term Evolution (LTE) and
IEEE 802.16m advanced WiMAX (Chang, Tao, Zhang, & Kuo, 2009; Zhang, Honglin, & Chen, 2008).
Recently, researchers have identified two fundamental problems that degrade the throughput of WMNs
(Li, Qiu, Zhang, Mahajan, Zhong, Deshpande, & Rozner, 2007). First, they identified that if we do not
consider the amount of data that the nodes can transmit, the network throughput may degrade when nodes
start transmitting more than what the intermediate links can support. They highlighted that this is possible
due to the presence of interference (i.e., multi-access interference (MAI)) which may cause additional
traffic to reduce the capacity of bottleneck links. The second problem that they identified is that the
current protocols are unable to accurately estimate link and path quality for the purposes of path selection.
This inaccurate estimation is caused since the routing protocols do not consider the interference when
they are determining the quality of links. However, due to interference, the quality can change arbitrarily
with any change in the routing pattern. Based on this work (Li, Qiu, Zhang, Mahajan, Zhong, Deshpande,
& Rozner, 2007), we can say that the use of OFDM as one part of the multi-access scheme (i.e., the
OFCD) can solve the problem of interference since it provides strong resistance against both MAI and
inter-symbol interference (ISI). The use of CDMA allows the fast and stable transmission of data.
Combining the features of these schemes, we can significantly enhance the network throughput for the
The demand for high speed wireless applications and limited radio frequency (RF) signal bandwidth
has spurred the development of power and bandwidth efficient air interference schemes. Therefore, the
adoption of more sophisticated multiple access technologies such as OFDM coupled with the CDMA,
provides two key benefits to WMN in the form of scalability and throughput gain. However, leveraging
these benefits call for more sophisticated transmitter and receiver designs for WMNs that can provide
greater system capacity with a reasonable bit error rate (BER) performance.
WMN have emerged to be a cost effective and performance adaptive paradigm for the next generation
wireless Internet. The main reason why WMN is so attractive for several applications is its low cost of
deployment and maintenance due to the absence of a wired infrastructure. However, the absence of a
wired infrastructure causes communications among routers to suffer from noise and interference problem.
The interference in WMN can be alleviated by using the OFCD as a multicarrier/multi-access scheme that
provides strong resistance against noise and ISI which is especially occurred in wideband transmission
over multipath fading channels (Gupta & Kumar, 2001; Yu, Mohapatra, & Liu, 2008).
OFDM is a multi-carrier multi-access scheme which was originally designed and developed in late
1960’s (Ye & Gordon, 2006). In OFDM, the data is transmitted using a large number of sub-carriers that
are completely orthogonal (Chuang & Sollenberger, 2000). Each sub carrier is modulated with a
conventional modulation scheme at a low symbol rate. The transmitted data is typically divided into a
large number of parallel data streams with respect to the number of sub carriers. Even though, the
subcarriers used in OFDM are theoretically orthogonal, in practice, the orthogonality might not be
ensured that results ISI that is the most important effect of multi-path delay spread (Xiang, Peng-Jun,
Wen-Zhan, & Yanwei, 2008; Zhang, Honglin, & Chen, 2008). With the help of cyclic prefix, longer time
duration symbols are transmitted in which the length of each transmitted bit is typically longer than the
length of the impulse response of the channel (Ye & Gordon, 2006). This is one of the ways that can be
used to mitigate the effects of ISI.
One of the objectives of this chapter is to analyze the potential use of a combined OFCD multiple
access technique in a WMN. The OFCD is the hybrid of OFDM and CDMA. OFDM has become the
popular choice for air interface technology in future local and wide area wireless networks. For instance,
it has been applied to high speed wireless LAN (e.g., IEEE 802.11a) (Bing, Frank, James, Rainer,
Hermann, & Adam, 2001), and high performance radio LAN type 2 HIPERLAN/2 (Johnsson, 1999).
Code division multiple access (CDMA) is originally designed as a multiplexing technique with strong
spread spectrum characteristics. In CDMA, a pseudo random number generator (PRN) generates a
spreading code that has comparatively large frequency components than the frequency components of an
input narrow band signal. CDMA allows all users to transmit signals simultaneously by utilizing the
entire available spectrum. CDMA is typically used with the direct sequence (DS) and the frequency
hopping (FH) spread spectrums techniques. Most of the 3G cellular systems are proposing to use the
combination of direct sequence (DS CDMA) as their modulation technique. In DS-CDMA, transmitter
spreads the original data stream using a given spreading code in the time domain. The purpose of this
spreading of the input signal using the spreading code is to increase the frequency range of the resultant
signal. How effectively this technique can suppress MAI depends on the cross correlation between the
spreading codes. Theoretically, spreading codes are perfectly orthogonal; however, in practice, the
orthogonality can not be guaranteed.
The implementation of OFCD scheme in WMN requires the design of new sophisticated transmitter
and receiver models that should be capable of using OFCD as an underlying multichannel subcarrier. This
chapter provides a discussion on a new transmitter and a receiver model that uses OFCD as a multi-access
multi-carrier modulation scheme for transmitting and receiving the signals at the transmitting and
receiving ends, respectively. Moreover, the goal of this chapter is to provide an analysis of different
multi-access techniques such as CDMA, OFDM, and OFCD with respect to their utilization with the
WMNs. Our analysis could be used effectively to determine the performance differences between these
multi-access schemes when implement with a WMN. For the sake of the performance analysis and the
experimental verifications, this chapter adapted BER and the network throughput as performance
COMBINING DIFFERENT MULTIPLE ACCESS SCHEMES
Based on OFDM, several multiple access schemes have been designed such as OFDM-frequency
division multiple access (OFDM-FDMA), OFDM-time division multiple access (OFDM-TDMA), and
OFDM-code division multiple access (OFDM-CDMA). The first two multiple access schemes (OFDM-
FDMA and OFDM-TDMA) have been adopted by the IEEE 802.16 standard as two options for
transmissions at the 2.11 GHz band (Akyildiz, Wang, & Wang, 2005). The last one is also referred as
Multicarrier code division multiplexing (MC-CDMA) which has been drawing much attention as an
alternative to conventional direct-sequence CDMA (DS-CDMA).
Both FDMA and TDMA schemes are used as resource management schemes in a multiaccess
communication system where transmission resources are shared among multiple users (Wang & Xiang,
2006).The purpose of these techniques is to efficiently manage the resource sharing in a multiuser
environment based on the principle of timesharing (TDMA) and frequency-sharing (FDMA). In other
words, in a multicarrier system, the resources are shared among multiple users by allocating the subcarrier
to them across the time and frequency domains.
In OFDM-FDMA scheme, OFDM can be implemented to allocate subcarriers to different users where
each subcarrier can be multiplexed using the FDMA technique. In other words, each subcarrier can be
considered as a narrowband communication channel that contains a portion of the entire bandwidth. Since
each user has its own subcarrier, this allows users to transmit their data simultaneously over parallel
channels. The principle advantage of OFDM-FDMA multiple access scheme is its ability to support
simultaneous downlink data transmissions to different terminals. Based on this basic multiple access
scheme, an OFDM-interleaved-FDMA scheme was proposed in which the subcarriers assign to users
need not be consecutive in their order (i.e., the subcarriers allocated to users can be interleaved). Even
though, the subcarriers are interleaved, they are fixed to users on time axis, and thus, making the recourse
allocation non-flexible (Bing, Frank, James, Rainer, Hermann, & Adam, 2001).
In OFDM-TDMA scheme, we allocate predetermined time slots to users, where each allocated time slot
contains ODFM symbols. In other words, the OFDM-TDMA scheme allocates the OFDM symbols while
OFDMA allocates subcarriers to users. In addition, since the time slots are predetermined and pre-
assigned, the subcarriers of one OFDM symbol can not be allocated to different user. In OFDM-TDMA
scheme, the OFDM can be used in two modes for frame allocations: static and dynamic. In static modes,
frames are allocated to users which are independent of their channel conditions whereas the dynamic
mode allocates frames to users with the best channel gain. Based on the OFDM-TDMA, (Wang & Xiang,
2006) proposed a new multiple access scheme, which is called OFDM-TDMA with subcarrier allocation
(OFDM-TDMA/SA). In this scheme, OFDM symbols are organized in a TDMA frame where each
subcarrier for an OFDM symbols is assigned to a different user. The main aim of this scheme is to
enhance the flexibility of recourse allocation for an OFDM based multiple access system. This work is
slightly different from the work of (Cheong, Cheng, Lataief, & Murch, 1999), where different frame
structures are used for large radio and small radios resources.
To improve the MAC performance of WMN, many multiaccess techniques are currently under
development. For instance, Multichannel MAC (MMAC) (Bahl, Chandra, & Dunagan, 2004; Bahl, 2007),
multiple radios (Adya, Bahl, Padhye, Wolman, & Zhou, 2004), directional and steerable antennas (Bahl,
2007) are example of some new multi-access techniques. Several efforts have been made by the
researchers to combine the OFDM with the CDMA for wireless networks (Neishaboori & Kesidis, 2008;
Xiang, Peng-Jun, Wen-Zhan, & Yanwei, 2008; Zhang & Tang, 2006). The principle reason for using the
combined multiple access scheme (i.e., the OFDM-CDMA) is its ability to exploit frequency diversity in
an explicit manner. This exploitation of frequency diversity is possible in this multiple access scheme due
to the fact that the energy of the symbol is spread over several subcarriers. The combination of OFDM
with the CDMA is refereed as multicarrier (MC) CDMA (MC-CDMA).
CDMA is a multiplexing technique with some strong spread spectrum characteristics. In this scheme, a
number of users simultaneously access a channel by spreading their narrowband signals (i.e., input
informational signal) with pre-assigned signature sequence. CDMA has become a prominent multiple
access technique in mobile wireless systems (Prasanna & Ravichandran, 2007), because it has the
capability to provide higher capacity over conventional techniques such as TDMA and FDMA, and to
combat the hostile channel frequency selectivity (Neishaboori & Kesidis, 2008). Direct sequence (DS-)
and frequency hopping (FH-) CDMA techniques have been subject to extensive research. When CDMA
operates with DS, it suffers from the frequency selective fading problem, particularly in the downlink
where orthogonal spreading codes are typically employed (Zhang, Tan, Chun, Laberteaux, & Bahai,
2008). The basic idea of CDMA is to maintain a sense of orthogonality among the signature waveforms
in order to minimize the MAI. However, in practice, the orthogonality among the signature waveforms
can not be guaranteed. In other words, in CDMA systems, the communication channels are defined by the
pseudo-random codewords, which are carefully designed to cancel each other out as far as possible (i.e.,
maximizing the orthogonality between the codewords or signature waveforms) (Xiang, Peng-Jun, Wen-
Zhan, & Yanwei, 2008). These codewords are then used to spread a narrowband signal (i.e., each bit of
input signal) into a wideband signal by spreading it over a unique codeword. The bandwidth components
of the resultant wideband signal are much wider than the minimum bandwidth required transmitting the
original input narrowband signal.
On the other hand, OFDM is a multicarrier modulation scheme which has drawn a lot of attention in the
field of radio communications. OFDM divides the available bandwidth into a large number of orthogonal
bands or subcarriers. Each subcarrier has a bandwidth which is typically smaller than the coherence
bandwidth and thus exposed only to frequency flat fading (Chang, Chien, & Kuo, 2007; Zhang, Tan,
Chun, Laberteaux, & Bahai, 2008). OFDM addresses the ISI problem arising in channel where the signal
bandwidth exceeds the coherence bandwidth of the fading process (Wang & Xiang, 2006; Ye & Gordon,
2006). The transmission of data using OFDM mitigates the problem of the frequency selectivity in multi-
path fading channels while at the same time the use of CDMA provides good spectral properties of the
transmitted data (Hottinen & Heikkinen, 2006; Zhang, Tan, Chun, Laberteaux, & Bahai, 2008). Another
scheme for combining the OFDM with the CDMA is proposed in (Zhang, He, & Chong, 2005). In this
scheme, different input symbols are transmitted using multiple subcarriers. This proposed technique
offers all the advantages of multicarrier CDMA (MC-CDMA) scheme. However, this technique mainly
depends in the derivations of subcarriers.
With the combination of these two multi-access schemes, the data can be transmitted over a large
number of subcarriers where each subcarrier is assumed to be conventionally modulated. The combined
multi-access scheme provides the spectral efficiencies, increases the network capacity, and minimizes the
end-to-end delay to support 4G wireless systems. The principle advantage of using OFDM-CDMA as a
multiple access scheme is its ability to satisfy some of the main requirements of 4G wireless systems such
as the minimization of MAI and ISI and thus provides a better BER performance to the end user.
Broadband WMNs demand both high speed transmission rate and a more sophisticated multiple access
technique that can be used to minimize MAI and ISI and maximize the BER performance. The high speed
transmission rate requires higher frequency bands for transmission. However, due to sever frequency
selectivity in broadband communications, the number of resolvable multiple paths fading degrades BER
performance. The MC-CDMA is a multiple access scheme that is typically used with an OFDM based
system that allows system to support multiple users simultaneously. In MC-CDMA, each data symbol is
transmitted at multiple narrowband subcarriers where each subcarrier is typically encoded with a phase
offset of 0 or 180 degree (Hottinen & Heikkinen, 2006). It uses WALSH code which is an orthogonal
spreading code sequence in a frequency domain (Zhang, Tan, Chun, Laberteaux, & Bahai, 2008). In MC-
CDMA, all data symbols are not transmitted on each subcarrier. Instead, they can be transmitted on some
of the selected channels. Those few channels on which data symbols can be transmitted are chosen after
the channel assignment. This scheme not only transmits the data symbols but also resolves the problem of
flat fading by minimizing the bit lost. Each subcarrier can be used by all users presented in the system
whereas all the active users are differentiated by a WALSH code. In MC-CDMA, since each receiver uses
a fast Fourier Transform (FFT) circuit with a variable gains diversity combiner, signal can be easily
recovered (Lin & Rasool, 2007).
For more connectivity, mesh routers are equipped to provide connectivity between different networking
technologies such as Wi-Fi, IEEE 802.11, mobile technology and wired Ethernet. In WMN, each client
with same radio technology communicates via Ethernet links, for different clients communications are
first made with their base station which has Ethernet connections to the mesh routers. Sometimes client
nodes actually form network to perform routing and this kind of infrastructure is called client wireless
mesh networks. Mesh clients can perform mesh functions with other mesh clients as well as accessing the
network through routers yielding hybrid wireless mesh networks.
One of the main advantages of OFDM scheme is its robustness to frequency selective fading. However,
this scheme also requires synchronization in each subcarrier and sensitivity to frequency offset and
nonlinear amplification. This problem is caused in OFDM due to the fact that it is composed of several
subcarriers with their overlapping power spectra and exhibits a non constant nature in its envelop.
However, the combinations of CDMA with the ODFM scheme can significantly minimize the symbol
rate in each subcarrier to maximize the symbol duration which makes the synchronization process simple
among the multiple transmissions of users. Using OFDM with CDMA, symbols are transmitted on many
carriers. In addition, different spread input symbols are fed to the subcarriers, when OFDM is combined
with the CDMA scheme (Zhang & Tang, 2006). With OFDM technique, frequency selectivity in
multipath fading channel is resolved (Cheong, Cheng, Lataief, & Murch, 1999; Zhang, Tan, Chun,
Laberteaux, & Bahai, 2008).
FEATURES OF OFCD MULTIPLE ACCESS SCHEME FOR WMN
In this section of the chapter, we first give a logical reasoning of using OFCD in WMNs. Specifically,
we discuss that what flexibilities this new multiaccess technique provides to WMN and what changes we
may need to make in the framework of both WMN and OFCD in order to implement this new multiaccess
WMN’s operation is similar to the way that packets are routed over the wired Ethernet (i.e., data hops
from one device to another until it reaches its destination). This is possible only when each node shares its
dynamic routing algorithm with every single node to which it is connected. The routing algorithm
implemented in each node takes the fastest route to its destination. Since in WMNs there is no central
server, each node (client) transmits data to the next node. As a result, each node behaves like a repeater
that forms an externally big network which is analogous to the Internet. In today’s scenario, hybrid
WMN’s have taken place of basic WMNs. The basic advantage of Hybrid WMN is that the network can
be accessible either through mesh routers or through mesh clients. It supports all different kinds of
network technologies like wired Ethernet, mobile communication, Wi-Fi, Wi-MAX, IEEE 802.11 etc.
The access points form a wireless backbone, providing connectivity in places otherwise it is difficult to
access through traditional wired infrastructure. The wireless communication between the access points
can use different technologies such as IEEE 802.11a/b/g or IEEE 802.16 and different hardware
(directional or Omni-directional antennas).The use of multi channels in wireless network leads to
throughput and reduced delay. One class of such protocols divides the available channels in two classes,
control and data channels. Control channels are used to exchange network control information, while data
channels are used for data transfer (Li, Qiu, Zhang, Mahajan, Zhong, Deshpande, & Rozner, 2007;
Rappaport, 2002). The use of multi transceiver allows a node to scan all available channels concurrently,
hence solves many complex problems.
Currently, in WMNs, most of the systems implement distributed multiple access schemes such as
CSMA/CA as their multiaccess scheme. The principle advantage of using this scheme in WMNs is that it
does not require the accurate timing synchronization within the global network. For WMN, as the size of
the network grows, CSMA/CA systems suffer from scalability issues (i.e., when the size of the network
increases, the network performance degrades significantly due to significant throughput reduction. For
instance, current IEEE 802.11 MAC protocol and its variants cannot achieve a reasonable throughput as
the number of hops increases to 4 or higher. This low scalability is due to the fact that the end-to-end
reliability sharply drops as the scale of the network increases. This implies that in a large network, CSMA
systems may suffer from high packet queuing delays.
Moreover, the current multiaccess scheme (CSMA/CA) has very low frequency spatial-reuse efficiency
(Acharya, Misra, & Bansal, 2003), which significantly limits the scalability of CSMA/CA-based multi-
hop networks. In order to fundamentally resolve the issue of low end-to-end throughput in a WMN,
innovative solutions are necessary. Determined by their poor scalability in WMN, random access
protocols such as CSMA/CA are not an efficient solution (Kim & Bambos, 2002; Yu, Mohapatra, & Liu,
2008). Thus, revisiting the design of MAC protocols based on OFDM and CDMA is an important
research topic (Acharya, Misra, & Bansal, 2003). As of now, only few TDMA or CDMA based MAC
protocols have been proposed for WMNs. This is mainly because of two reasons. Firstly, it is relatively
expensive from cost point of view to design and implement a distributed and cooperative MAC protocol
with CDMA scheme. Secondly, a framework is needed to provide compatibility between CDMA and
other existing MAC protocols.
CDMA offers several advantages to WMN when it works as a multiaccess scheme. First, nodes in
CDMA networks can interfere each other but they do not damage each others’ data as long as the degree
of interference is relatively low. This implies that the CDMA does not impose any hard limit on the user
capacity since we can continue adding users in a CDMA network as long as the level of interference can
tolerate. This is one of the features that make CDMA unique when compared to the other conventional
multiplexing schemes such as TDMA and FDMA. Another advantage that CDMA provides is that the
data transmission rate for wireless networks can be increased by using the additional power control. This
is especially true for 802.11 WLAN where the data rate is restricted due to the absence of a proper power
control. The implementation of power control with CDMA allows us to increase the transmission rate of
certain traffic flow which consequently increases the transmission range and decreases the number of
hops between the transmitter and receiver. The end result of this additional power controller module in
CDMA provides several advantages such as lower BER, less end-to-end delays, and higher throughput.
Based on the above discussion, one may conclude that the combination of CDMA and OFDM (we refer
it as OFCD) may provide several advantageous to WMN as a multiaccess scheme such as minimizing the
end-to-end delay, improving the BER performance, and increasing the network capacity. In order to
increase the capacity of WMN, OFDM has significantly increased the speed of IEEE 802.11 from 11
Mbps to 54 Mbps (Chang, Chien, & Kuo, 2007; Neishaboori & Kesidis, 2008).
IMPLEMENTATION OF OFCD MULTIPLE ACCESS SCHEME IN WMN
In this section, we present a discussion on the implementation of OFCD multiaccess scheme in WMN.
Specifically, we present the system model and the framework required to implement the OFCD with the
WMN. To support the framework of OFCD, we also provide an analytical model to exhibit some of the
strong characteristics of OFCD that it offers for WMN. All system variables, along with their definitions,
are listed in Table I. Before we present the implementation of OFCD in WMN, it is worth mentioning
some of our key assumptions:
Assumptions and System Model
• There are n available channels where each channel has equal bandwidth.
• The access points or base station can receive data on multiple channels simultaneously. This is a
reasonable assumption since the access points can be more specialized in higher end device as
compared to a simple client that it serves.
• The channels are orthogonal and CDMA scheme is used (i.e., transmission on a channel does not
interfere with transmission on any other channels). Here a channel may represent a code or a
frequency band (OFDM-CDMA multi-access technique).
• Each network node including the base station is equipped with a multi radio multi transceiver,
which is capable of performing in full duplex mode. Hence each node can either transmit or
receive a signal on channel at any point in time. The nodes can however switch to different
• The network is assumed to be fully synchronized (Prasanna & Ravichandran, 2007).
Basic Framework of OFCD in WMN
For multi hop mesh network, we consider the hybrid architecture as it is the most widely deployed
architecture. This architecture is characterized by the fact that mesh clients do not need direct connection
to a mesh route, but can connect multi hop over mesh clients to a route. The advantages are improved
connectivity and coverage. And the disadvantages are that mesh clients need more resources because they
also need to have routing capability.
Table I. Notation Used In Analysis
Notation Related Quantities
Represents number of channels for signal
N Represents number of nodes in unit area (A)
Represents frame length (bits)
Number of contention slots
Number of data slots
(interval length) Contention duration
(length of data interval) Data duration
Let’s assume there are N sensor nodes distributed over an area A. Sensors are assumed to be
independently and uniformly scattered over a region of interest (Yu, Mohapatra, & Liu, 2008). Number of
nodes (n) is assumed to be distributed independently and uniformly over an area of Πr
. Each node can
communicate with every other node within the radius of r. Taking these factors into account, one can
derive the following expression:
r n c n n
= + ⁄
The networking is connected with unit probability if and only if the following expression exists:
lim ( )
. From this expression, one can choose the transmission range. After flooding the network
traffic, network is organized into a tree with the observer at each possible route. In the first step of
flooding, the observer first broadcasts a wakeup signal. All sensor nodes within the direct communication
range receive this signal and reply to the observer. Once the observer receives one or more messages from
sensor nodes, it registers them as first level nodes in the node tree and instructs them to repeat the process
of broadcasting in a time shared manner to avoid collisions.
All sensor nodes that were not previously registered would register themselves as first level nodes when
receives these broadcast messages from observer. Those sensor nodes that were at first level move to the
second level. More over, the second level nodes continue to broadcast and repeat the same procedure until
all nodes have been registered. Whenever a node broadcasts a wakeup signal, it also attaches its unique
address and chain of nodes which leads to it from the observer. Nodes that are wakened up by the
broadcast are designated as children of broadcasting nodes. These children nodes obtain the chain of
nodes leading from observer to them by concatenating the last link with the chain, leading to there parent.
Each node obtains a route from observer to reverse this chain.
The same reasoning suggests that a fewest-hop route should be optimal even in a mobile observer
network. Since the observer does not stay at a fixed position, the fewest-hop route is time variant in
nature, and so is the number of hops. Quite obviously, the best solution is to choose the route which
consists of the fewest hops at any time. In other words, if S = Ns
, ... is the set of all the nodes
that come within a direct communication range of the observer at any time, then the fewest-hop route to
the observer is the shortest of the fewest-hop routes to any of the nodes in S joined with the link between
the corresponding node in S and the observer.
Finding the shortest route in practice involves a procedure very similar to the flooding procedure
described above except one difference. The 1
level nodes (those belonging to the set S just described) are
discovered by moving the observer on its path while transmitting a wake- up signal to all nodes that are
within the range. These nodes are registered as 1st level nodes. The remaining process of flooding
proceeds exactly as described in this section. At the end of this registration procedure, each node in the
network knows its route to one of the nodes in S that communicate directly with the observer.
OFCD System Model
As we briefly mentioned in Introduction Section, one of the objectives of this chapter is to describe a
hybrid multiple access scheme for WMNs in which different users are transmitting signals simultaneously
over multiple subcarriers where each subcarrier is the sum of orthogonal subcarriers. Such a hybrid
multiple access technique is indeed very important in order to over come some of the deficiencies caused
by the use of distributed multiaccess technique such as CSMA/CA. To show this, we present a discussion
on the implementation of OFCD in WMN.
OFCD is the combination of OFDM and CDMA spread spectrum multiple access scheme. Next, we
present an analytical model for both transmitter and receiver for a WMN. The design of transmitter and
receiver help understanding the implementation of OFCD multiple access scheme for WMN. All system
variables, along with their definition, are presented in Table II.
Transmitter Design for OFCD in WMN
As shown in Fig. 1, x(k) is the discrete digital data that transmitter receives from a digital data source
where k represents discrete time. As mentioned above, in OFCD scheme, the transmitter spreads the
original data stream (i.e., x(k)) over different subcarriers using the assigned spreading code. The
assignment of spreading code to each user is done by CDMA technique.
The serial to parallel converter will convert x(k) into n number of parallel data symbols with a symbol
rate of 1/T. The parallel symbols are x
(k) to x
(k). Quadrature phase-shift keying (QPSK) block is used
which have the carrier frequency in the range of f
as its other inputs. The QPSK block will split the
input bit stream into in-phase and quadrature phase components. The quadrature components and the in-
phase components will be modulated with f
carrier frequencies. Based on these factors, the output of
a QPSK block would be approximated as:
fkx +)( .
This output block is then supplied to the mixer as shown in Fig. 1. At mixer, PRN code C
the incoming signal
fkx +)( . Taking these factors into account, we can derive the following
Fig. 1 IMAGE GOES HERE
Figure 1. Transmitter Model for WMNs with Orthogonal Frequency Code Division (OFCD) Multiple
Table II. Notation
s Used In Proposed Transmitter a
Notation Related Quantities
x(k) Input from digital data source
Parallel data symbols
Symbol mixed with PRN codes
Output of the mixer
Output of IDFT block
x(t) Analog received signal
w(k) Non negative weight function
Output of DFT
Output of mixer at receivers end
n(t) White additive Gaussian noise
G Guard time matrix
CfkxX ].)([ += (2)
The Inverse Discrete Fourier Transform (IDFT) would be used for modulation and it is described as:
where k = 0,…, N-1.
The last symbol coming out of IDFT block is taken and added on the beginning of source code block to
provide guard time which in fact provides orthogonality (this is analogous of an ideal case of OFDM).
The proper choice of the number of subcarriers and the guard time is important in order to increase the
robustness against the frequency selective fading.
If the last symbol coming out of IDFT block is X
, then both (2) and (3) can be combined together.
where w(k) in (4) is the non-negative weight function of OFCD and
orthogonal if and only if the following expression exists:
As the input bit stream can be of infinite length, the infinite integral can be defined as follows:
After digital to analog conversion, discrete digital signal would be converted into analog signal
and then would be transmitted over the medium.
Receiver Design for OFCD in WMN
In Fig. 2, x(t) is the received signal which can be shown as:
Fig. 2 IMAGE GOES HERE
Figure 2. Transmitter Model for WMNs with Orthogonal Frequency Code Division (OFCD) Multiple
where n(t) in (7) represent noise which gets added during the transmission. The signal has been passed
through the guard time removal block to remove the guard time G. the guard time can be expressed as:
where G is a matrix of 0’s and 1’s.
Receiver performs Discrete Fourier Transform (DFT) to demodulate the signal. DFT can be represented
where k = 0,…,N-1.
is then fed into the mixer which will mix it with the PRN code C
assumed to be deterministic to the receiver.
As we are deploying the CDMA’s feature, the receiver should know the seeds of the PRN codes which
were used at the transmitter. The output of mixer is represented by X(k) and can be further expressed as:
XCkX =)( (10)
The DS-CDMA is used to detect X(k). DS-CDMA scheme simply performs the XOR operation between
the PRN codes and the signal coming out of the mixer X(k). The unique chip sequence is used by DS-
CDMA depends on the number of seeds used by the transmitter. Band Pass Filter (BPF) will pass only a
certain range of frequency by considering the maximum and minimum frequency components of the
carrier frequency. If f
is the highest frequency component and f
is the smallest frequency component then
and those detected signals will be passed which lie in the frequency range of f
With the parallel to serial converter original transmitted signal x(k) will be fully recovered. Given the
property of a frequency selective fast multipath fading channel, there exists the optimal value to minimize
the BER in the number of subcarriers and the length of the guard interval. This is one of the features that
one can only achieve when a hybrid of CDMA and OFDM is implemented as a multiple access scheme in
The OFCD scheme for WMN was implemented using NS2. We use frequency selective slow Rayleigh
fading channels as channel model. The frame length of the OFCD frame is assumed to be 3 ms which is
close to that of IEEE 802.11. A total of 8 time slots are used in which 5 time slots are dedicated for
downlink frequency traffic where as rest of them are used for uplink traffic transmission. We assign
exactly one OFDM symbol to each time slot whereas each ODFM symbol contains 128 subcarriers for
the users/nodes. As mentioned in the previous section, QAM techniques was employed in each subcarrier
Fig. 3 shows the throughput of a WMN with the use of conventional distributed MAC protocol (i.e.,
CSMA/CS). This simulated computer network for WMN is congested with 10 nodes transmitting
randomly to generate the throughput rate.
The peak end-to-end throughput that can be achieved without using the OFCD is upper-bounded by
40080 bytes per seconds as shown in Fig. 3. Fig. 4 shows the end-to-end throughput for the case where
we implement OFCD as a multiple access scheme for WMN. As we can see in Fig. 4 that the throughput
increases significantly from 40080 bytes per seconds to 70440 bytes per second.
The BER performance comparison between the OFCD and CSMA/CA for WMN is presented in Fig. 5.
The theoretical-exact 0 (i.e., it represents the BER performance of conventional distributed MAC
protocol) is computed based on the implementation of CSMA/CA for WMN whereas the theoretical-
exact 1 is computed based on the implementation of OFCD for WMN. We consider the same parameters
as we discussed above for measuring the BER performance for implementing both CSMA/CA and OFCD
for WMN. As one can see in Fig. 5 that the OFCD outperforms the conventional MAC protocol for all
values of E
This significant gain in BER performance for OFCD is achieved due to the fact that
both OFDM and CDMA exhibit strong characteristics that can effectively eliminate both MAI and ISI
even in the presence of large subcarriers and users.
Fig. 3 IMAGE GOES HERE
Figure 3. An Illustration of End-to-End Throughput without using OFCD for WMNs
Fig. 4 IMAGE GOES HERE
Figure 4. Implementation of OFCD for WMNs to Maximize the End-to-End Throughput
Fig. 5 IMAGE GOES HERE
Figure 5. BER Performance of OFCD for WMNs
Future Research in WMN
Several efforts have been made to implement OFDM based technologies in WMN such as OFDM-
FDMA (Cheong, Cheng, Lataief, & Murch, 1999), OFDM-TDMA (Bing, Frank, James, Rainer,
Hermann, & Adam, 2001), and OFDM-CDMA (Prasanna & Ravichandran, 2007; Xiang, Peng-Jun, Wen-
Zhan, & Yanwei, 2008; Zhang & Tang, 2006). The implementation of these multi-access techniques for
the WMN brings many research challenges. For instance, one of the major issues in the design and
engineering of OFDM based WMNs would be cross layer optimization involving OFDM based
transmission, radio link level queue management, and network layer admission control management.
Recently researchers have investigated the QoS performances in a solar-powered mesh network using
OFDM-based radio transmission (Niyato, Hossain, & Fallahi, 2006). Specifically, they developed a
queuing analytical model to analyze both the connection-level and the packet-level performances at a
mesh node considering constrained power supply at that node. Their analysis demonstrates that their
proposed analytical framework can be used to achieve the optimal connection admission control (CAC)
threshold. The CAC threshold value can be further used to guarantee packet-level QoS.
As mentioned earlier, a WMN puts a hard limit on the number of users (i.e., the network capacity) that
it can support at one time due to the presence of interference. The network capacity also degrades sharply
when WMN is being used in multi-hop mode. One of the reasons of limited capacity in WMN is the
presence of interference within a limited band of radio frequencies. Recently, researchers have identified
that not only the channels but also the transmission rates of the links have to be properly selected to make
a given set of flow rates schedulable (Kashyap, Ganguly, & Das, 2007). Specifically, they developed a
forwarding paradigm to achieve a resulting set of flow rates while using a standard MAV protocol. Their
work presented a measurement-based model that can capture the effect of interference in 802.11 based
WMNs (Kashyap, Ganguly, & Das, 2007). In general, their proposed measurement-based model can
characterize and model the impact of interference caused by active traffic from multiple surrounding
nodes on the link capacity. Specifically, the main objective of their work is to model the network capacity
of any given link in the presences of any given number of interferers in a deployed network, carrying any
specified amount of traffic load.
Extensive research has done for improving the network capacity for WMN (Bahl, Chandra, & Dunagan,
2004; Raniwala & Chiueh, 2005). One scheme is to improve the network capacity is to utilize multiple
channels that are available in the IEEE 802.11 a/b/g standards (Yu, Mohapatra, & Liu, 2008). In order to
fully utilize the available multi-channels, the best solution is to equip each mesh node with multiple
radios that are tuned to different frequencies. For instance, (Bahl, Chandra, & Dunagan, 2004) identified
that the channel switching time could be decreased. This implies that it is possible for each node to use a
different channel in each time slot. Recently, (Yu, Mohapatra, & Liu, 2008) proposed a new scheme in
which they determine the highest gain by increasing the number of channels and radios with certain traffic
demands under both dynamic and static channel assignments. However, the channel switching requires
fine-grained synchronization among nodes in order to avoid the deafness problem (i.e., the transmitter and
the receiver may be on different channels at one time) (Kashyap, Ganguly, & Das, 2007). Also, the time
for channel switching which can be in the range of few milliseconds to a few hundred microseconds may
be unacceptable for most real time multimedia applications (Kyasanur & Vaidya, 2005).
In this chapter, we have studied a multi radio multi channel WMN using OFCD which amalgamates the
advantages of both OFDM and CDMA. A transmitter and a receiver model are described with the help of
useful mathematical expressions. This is a simple and a realistic multi-access technique that has good
performance properties. In very noisy multipath channel, the hybrid OFCD multiaccess scheme is
expected to work efficiently and provides better BER performance and end-to-end throughput as compare
to the conventional multiaccess schemes such as CSMA/CA. We have also highlighted some key
advantages of multi-radio wireless mesh networks along with primary technical challenges that must be
addressed for widespread deployment of such networks such as scalability, interference, and network
throughput. WMNs present a promising solution by extending network coverage based on mixture of
above wireless technologies through multi-hop communications. Finally, we also discussed several access
schemes that can be developed and implemented based on the OFDM systems for WMN such as OFDM-
FDMA, OFDM-TDMA, and OFDM-CDMA (also referred as MC-CDMA).
Acharya, A., Misra, A., & Bansal, S. (2003). High-performance architectures for IP-based multi-hop
802.11 networks. IEEE Wireless Communications, 10 (5), 22–28.
Adya, A., Bahl, P., Padhye, J., Wolman, A., & Zhou, L. (2004). A multi-radio unification protocol for
IEEE 802.11 wireless networks. In Proc. of the IEEE BroadNets, (pp. 344-354).
Akyildiz, F., Wang, X., & Wang, W. (2005, March). Wireless mesh networks: a survey. Computer
Networks, 47 (4), 445-487.
Avallone, S., Akyildiz, I., & Ventre, G. (2009). A channel and rate assignment algorithm and a layer-2.5
forwarding paradigm for multi-radio wireless mesh networks. IEEE/ACM Transactions on Networking,
17 (1), 267-280.
Bahl, P., Chandra, R., & Dunagan, J. (2004). SSCH: Slotted channel hopping for capacity improvement
in IEEE 802.11 ad hoc wireless networks. In Proc. of IEEE 802.11 Ad-Hoc Wireless Networks,
Bahl, V. (2007). Wireless mesh networks. IEEE INFOCOM Tutorial.
Bing, C., Frank, F., James, G., Rainer, G., Hermann, R., & Adam, W. (2001). Framework for combined
optimization of DLC and physical layer in mobile OFDM systems. In Proc. of the 6
OFDM workshop, Hamburg, (pp. 32-36).
Chang, Y., Chien, F., & Kuo, J. (May 2007). Cross-layer QoS analysis of opportunistic OFDM-TDMA
and OFDMA networks. IEEE J. Select. Areas Commun., 25(4), 657 – 666.
Chang, Y., Tao, Z., Zhang, J., & Kuo, J. (2009, June). A graph approach to dynamic fractional frequency
reuse (FFR) in multi-cell OFDMA networks. Accepted for publication in IEEE ICC'09, June 2009, from
Cheong, Y., Cheng, R., Lataief, K., & Murch, R. (1999). Multiuser OFDM with adaptive subcarrier, bit,
and power allocation. IEEE J. Select. Areas Commun. 17 (10), 1747-1757.
Chuang, J., & Sollenberger, N. (2000). Beyond 3G: wideband wireless data access based on OFDM and
dynamic packet assignment. IEEE Commun., Magazine, 38 (7), (78 – 87).
Gupta, P., & Kumar, R. (2001, January). The capacity of wireless networks. IEEE Transactions on
Information Theory, 46 (2), 388 – 404.
Hottinen, A., & Heikkinen, T. (2006, March). Subchannel assignment in OFDM relay nodes. In Proc. of
40th Annual Conference on Information Sciences and Systems, (pp. 1314-1317), Princeton, NJ.
Johnsson, M. (1999). HiperLAN/2- The broadband radio transmission technology operating in the 5GHz
frequency band, technical specification. V 1.0, HiperLAN/2 Global Forum.
Kashyap, A., Ganguly, S., & Das, S. (2007). A measurement-based approach to modeling link capacity in
802.11-based wireless networks. In Proc. of the 13th annual ACM international conference on Mobile
computing and networking, (pp. 242-253).
Kim, J., & Bambos, N. (2002). Power efficient MAC scheme using channel probing in multi-rate wireless
ad hoc networks. In IEEE Vehicular Technology Conference, (pp. 2380–2384).
Kyasanur, P., & Vaidya, N. (2005). Routing and interface assignment in multi-channel multi-interface
wireless networks. In Proc. of IEEE Wireless Communications and Networking Conference, Vol. 4. (pp.
Lin, X., & Rasool, S. (2007, May). Distributed joint channel assignment, scheduling and routing
algorithm for multi-channel ad hoc wireless networks. In Proc. of 26th IEEE International Conference on
Computer Communications, (pp. 1118-1126).
Li, Y., Qiu, L., Zhang, Y., Mahajan, R., Zhong, Z., Deshpande, G., & Rozner, E. (2007). Effects of
interference on throughput of wireless mesh networks: Pathologies and a Preliminary Solution. In Proc.
of HotNets-VI, Atlanta, GA, USA.
Neishaboori, A., & Kesidis, G. (2008). Wireless mesh networks based on CDMA. Computer
Communications, 31 (8), 1513 – 1528.
Niyato, D., Hossain, E., & Fallahi, A. (2006). Solar-powered OFDM wireless mesh networks with sleep
management and connection admission control. In Proc. of the 2006 international conference on Wireless
communications and mobile computing, (pp. 653-658).
Prasanna, G., & Ravichandran, V. (2007). Performance analysis of MC-CDMA for wide band channels.
Information Technology Journal, 6(2), 267-270.
Raniwala, A., & Chiueh, C. (2005, March). Architecture and algorithms for an IEEE 802.11 based multi-
channel wireless mesh network. In Proc. of the IEEE Infocom, Miami, (pp. 13-17).
Rappaport, T. (2002). Wireless communications: principles and practice. 2
edition, Prentice Hall.
Sundaresan, K., & Rangarajan, S. (2008). On exploiting diversity and spatial reuse in relay enabled
wireless networks. In Proc. of the 9th ACM international symposium on Mobile ad hoc networking and
computing, (pp. 13-22).
Tse, D., & Grossglauser, M. (2002, August). Mobility increases the capacity of ad hoc networks.
IEEE/ACM Trans. Net, 10 (4), 477 – 486.
Wang, X., & Xiang, W. (2006, March). An OFDM-TDMA/SA MAC protocol with QoS constraints for
broadband wireless LANs. Wireless Networks, 12 (2), 159–170.
Xiang, Y., Peng-Jun, W., Wen-Zhan, S., & Yanwei Wu. (2008). Efficient throughput for wireless mesh
networks by CDMA/OVSF code assignment. Ad Hoc and Sensor Wireless Networks, Vol. 5, No. 3-4, (pp.
Ye, L., & Gordon, L. (2006, February). Orthogonal frequency division multiplexing for wireless
Yu, H., Mohapatra, P., & Liu, X. (2008, April). Channel assignment and link scheduling in multi-radio
multi-channel wireless mesh networks. Mobile Networks and Applications, Vol. 13, (pp. 169-185).
Zhang, Y., Honglin, H., & Chen, H. (2008, April). QoS differentiation for IEEE 802.16 WiMAX mesh
networking. Mobile Networks and Applications, 13 (1-2), 19-37.
Zhang, Z., He, Y., & Chong, K. (2005, March). Opportunistic downlink scheduling for multiuser OFDM
systems. Wireless Communications and Networking. Vol. 2, No. 13-17, (pp. 1206 – 1212).
Zhang, X., & Tang, J. (2006, July). QoS-driven asynchronous uplink sub-channel allocation algorithms
for space-time OFDM-CDMA systems in wireless networks. Wireless Networks, 12 (4), 411–425.
Zhang, Y., Tan, I., Chun, C., Laberteaux, K., & Bahai, A. (2008). A differential OFDM approach to
coherence time mitigation in DSRC. In Proc. of the fifth ACM international workshop on Vehicular Inter-
Networking, (pp. 1-6).
Zhou, D., & Lai, H. (2005). A compatible and scalable clock synchronization protocol. Synchronization
scheme reference IEEE working group. A wireless LAN MAC and physical layer (PHY) specification,