Issues and Challenges of Energy-efficient Hybrid Routing Schemes: A Review
Ming Ann Ng,
Kok-Lim Alvin Yau
The increasing focus on green communications is bringing a
fresh attention towards the research works on energy-efficient hybrid routing
schemes. The energy-efficient hybrid routing schemes are applied in various
types of wireless multi-hop networks. This review discussed the characteristics
and challenges associated with energy-efficient hybrid routing schemes.
July 08, 2012; Accepted: September 18, 2012;
Published: November 01, 2012
Various researches on energy-efficient hybrid routing schemes have been carried
out to provide energy-efficient wireless communications in order to prolong
the lifetime of battery-powered devices in wireless multi-hop networks, such
as Wireless Mesh Networks (WMNs) (Wei et al., 2007),
wireless ad hoc networks (Alfawaer et al., 2007),
Mobile Ad hoc Networks (MANETs) (Viswacheda et al.,
2007), Wireless Sensor Networks (WSNs) (Thangammal et
al., 2012) and cognitive radio networks (Tingrui
et al., 2011). Routing enables a source to search and establish the
best possible route(s) to the destination (Al-Rawi and Yau,
2012). Each link within a route represents different types of dynamic costs
or rewards. For instance, the link cost is generally associated with energy
consumption in energy-efficient routing schemes (Nurul Huda
et al., 2007; Shi et al., 2010).
Although, hybrid routing has been investigated in the past (Cano
and Kim, 2002; Helmy, 2005), this topic has gained
fresh attention with the recently published IEEE 802.11s Wireless Mesh Networking
standard by the 11s Task Group in September, 2011. A hybrid routing scheme,
namely Hybrid Wireless Mesh Protocol (HWMP), has been proposed. CARrier grade
MEsh Networks (CARMEN) (Azcorra et al., 2009)
is another kind of wireless mesh network, which aims to provide high performance
triple play services. High performance triple play services ensure that voice,
video and data contents are received by the end users with the highest possible
quality. Studies have been made on CARMEN, such as cost analysis (Chieng
et al., 2010), monitoring and performance management (Loziak
et al., 2010), MAC layer analysis (Serrano et
al., 2009), self configuration design (Simsek et
al., 2009) and Architecture design (Banchs et
Figure 1 shows an IEEE 802.11s-based WMN. Three types of
nodes are mesh STAtion (STA), mesh gateway and Mesh Access Point (MAP). The
mesh STAs connect among themselves to form a mesh network. Based on its functions,
a STA may serve as a MAP or a gateway. A MAP provides access to the clients;
while a gateway provides connection to internet or other IEEE 802-based networks
(e.g., IEEE 802.16) (Hiertz et al., 2010). Packets
from a client are sent to a mesh MAP, and they traverse through the mesh network
in one or multiple hops to the mesh gateway (Lin et al.,
HWMP applies layer two (data-link layer) routing, which is based on MAC address,
instead of applying layer three (network layer) routing, as seen in the conventional
IEEE 802.11 standards. The upper layers see any destination node in a WMN as
its direct neighbour; but in fact, the destination node may be multiple hops
away. HWMP is a hybrid routing scheme comprised of proactive tree-based and
reactive routing approaches. Using a proactive tree-based routing approach,
a root node establishes and maintains routes connecting all nodes in the network.
A drawback is that all packets must be forwarded to the root node, which then
forwards them to their respective destinations.
|| A WMN based on the IEEE 802.11s standard
Thus, a source node must send its packets to the root node, which may be multiple
hops away, in order to send packets to its neighbour node. Using a reactive
routing approach, a source node can establish new route to destination node
directly without passing through the root node, and this addresses the aforementioned
Despite of the elegant routing design, this aspect of energy efficiency has
not been implemented in HWMP. This is important as WMNs are suitable to be deployed
in rural areas where there is lack of basic infrastructure (Kwong
et al., 2011; Ting et al., 2011).
Due to the similarity in characteristics and challenges faced by WSNs, especially
in terms of multi-hop transmissions and energy awareness, we include WSNs in
our review beside WMNs and MANETs. To the best of our knowledge, this is the
first comprehensive review on the aspect of energy-efficient for Hybrid Routing
Schemes in Wireless Multi-hop Networks.
CHARACTERISTICS OF ROUTING APPROACHES
Traditional characteristics (e.g., tree and flat) are briefly presented. Figure
2 shows the taxonomy of the characteristics. Generally speaking, there are
three types of characteristics, namely, network structure, routing approaches
and energy-efficient decision factors.
There are four kinds of network structure, as follows:
||R1.1: Tree structure: A root node constantly broadcasts
routing packets to the entire network in order to connect all nodes in a
tree-like structure (Figueiredo et al., 2007)
||R1.2: Hierarchical/clustered structure: The entire network is divided
into lusters. Each cluster has a clusterhead that forwards its member nodes
packets to base station. There are two levels of routings: intra-cluster
routing connects member nodes to their respective clusterheads; while inter-cluster
routing connects clusterheads to base stations in single or multiple hops
(Zhu et al., 2010)
||R1.3: Zone-based structure: Each node forms its own routing zone
comprised of nodes within a certain number of hops away. Border nodes, which
are located at the border of a zone, establish routes between zones (Wang
et al., 2008)
||R1.4: Flat structure: Nodes are not organized into any form of
structure (Bernardos et al., 2006)
||Taxonomy of energy-efficient hybrid routing scheme characteristics
There are three kinds of routing approaches as follows:
||R2.1: Proactive routing: Routes are established and
maintained by a node through constant broadcast of routing information to
the entire network and so it may incur higher energy consumption
||R2.2: Reactive routing: Routes are established only when a source
node wants to send packets to its destination node. The source node broadcasts
Route Discovery, which is flooded throughout the network. When the destination
node receives Route Discovery, it sends Route Reply to the source node so
that it can establish the best route to the destination node
||R2.3: Hybrid routing: Hybrid routing is a combination of two or
more routing approaches. The routing approaches may be applied in two manners.
Firstly, the routing approaches are applied concurrently. For instance,
different types of routing approaches can be applied for communication among
intra-zone or inter-zone nodes in zone-based network structure (Wang
et al., 2008). Secondly, the routing approaches are applied in
an alternative manner. For instance, Zhao et al.
(2007) propose a routing scheme that switches among different routing
approaches based on the rate of changes in network topology
There are five kinds of decision factors for energy-efficient hybrid routing
schemes. A source node may take account of any of the following decision factors
while making routing decisions:
||R3.1: Transmission distance: Data transmission over
longer physical distance between a node and its next hop increases energy
||R3.2: Residual energy: Nodes with lower residual energy may cause
||R3.3: Traffic load: Different traffic load incurs different energy
consumption. As an example, a routing scheme may switch between proactive
and reactive routing approaches to reduce routing overhead and energy consumption
(Figueiredo et al., 2007). When traffic load
is low, constant route maintenance is not necessary, and so reactive routing
is applied. When traffic load is high, more Route Discovery packets may
be sent in reactive routing compared to route maintenance messages in proactive
routing, and so proactive routing is applied.
||R3.4: Location: Nodes are aware of their respective locations in
location-aware networks. For instance, nodes choose the shortest routes
to destination in order to reduce energy consumption (Priyankara
et al., 2010)
||R3.5: Node heterogeneity: Nodes may have different characteristics
in heterogeneous networks. For instance, a few high-end nodes with higher
transmission power and longer transmission range can relay large amount
of packets in order to reduce energy consumption of non-high-end nodes (Priyankara
et al., 2010).
The challenges associated with energy-efficient hybrid routing schemes are
divided into two categories. The node-level challenges are associated with issues
related to individual nodes, such as traffic load and energy consumption of
|| Taxonomy of the challenges of energy-efficient hybrid routing
The network-level challenges are associated with issues related to the entire
network such as end-to-end delay and routing overhead. Figure
3 shows the taxonomy of the challenges. The rest of this section presents
the two categories of challenges.
There are three types of challenges in the node-level category, as outlined
||C1.1: Longer inactive period: Nodes incur higher energy
consumption in active mode, as well as during transitions between active
and inactive modes. The challenge is that, nodes should be made to sleep
as long as possible without jeopardizing network performance as a result
of packet loss during sleep. As an example, Bernardos
et al. (2006) address this challenge using a scheduling scheme
that only allows a node to send packets during a certain time frame. When
a node is not permitted to send packets, the node switches to inactive mode
||C1.2: Different levels of residual energy among nodes: Nodes with
lower residual energy may cause link breakage. For instance, clusterheads
may incur higher energy consumption while handling higher traffic load in
a clustered network structure, and so it may have lower residual energy.
Tabibzadeh et al. (2009) address this challenge
by rotating the role of clusterhead among nodes within a cluster
||C1.3: Higher energy consumption of nodes closer to base station:
Nodes closer to base station may incur higher energy consumption while handling
higher traffic load. Priyankara et al. (2010)
address this challenge using flat routing among nodes closer to base station.
Flat routing can establish multiple routes towards the base station, and
this prevents hotspots, which tend to have lower residual energy.
There are five types of challenges in the network-level category as shown below:
||C2.1: Dynamic traffic load: The choice of routing approaches
can affect energy consumption of a network with dynamic levels of traffic
loads. For instance, when traffic load is low, reactive routing, which does
not require constant route maintenance, is applied (Figueiredo
et al., 2007). When traffic load is high, proactive routing is
applied by a base station to construct a routing tree throughout the network.
This avoids Route Discovery in proactive routing, and so it reduces energy
||C2.2: Higher rate of topological change: Different routing approaches
may be used when different levels of topological change are detected in
order to reduce energy consumption incurred by the distribution of routing
information. For instance, when the rate of topological change is low, it
is unnecessary to provide constant updates of routing information because
established route can be reused, and so reactive routing is applied (Zhao
et al., 2007). On the other hand, when the rate of topological
change is high, proactive routing is applied.
||C2.3: Lower node density: Lower node density may lead to lesser
options for next-hop node selection, and hence the next-hop node may be
physically further away. As a consequence, energy consumption increases
due to longer transmission range. For instance, when the node density is
high, cluster-based routing is applied so that clusterheads can aggregate
redundant data from member nodes (He and Wei, 2008).
However, when the node density is low, there is less redundant data, and
so flat routing is applied
||C2.4: Reduction of routing overhead: Reducing unnecessary flooding
of routing packets can reduce energy consumption. For instance, proactive
routing is applied to maintain routes among two-hop neighbour nodes so that
flooding of routing packets is reduced (Bernardos et
||C2.5: Achieving a balance between end-to-end energy consumption and
residual energy of nodes: A route comprised of nodes with higher (lower)
residual energy may incur higher (lower) end-to-end energy consumption.
For instance, when all nodes of a candidate route have sufficient amount
of residual energy, it chooses a route with the lowest end-to-end energy
consumption; otherwise, it chooses a route that avoids nodes with comparatively
lower residual energy (Cano and Kim, 2002)
ENERGY-EFFICIENT HYBRID ROUTING SCHEMES
Table 1 shows the characteristics of, the challenges being
addressed by, as well as the performance enhancement achieved by the energy-efficient
hybrid routing schemes for wireless multi-hop networks. From Table
1, several energy-efficient hybrid routing schemes are categorized based
on network structures rather than routing approaches. As an example, hybrid
routing schemes by Zhu et al. (2010) and Tabibzadeh
et al. (2009) are comprised of cluster-based R1.2 routing and flat
routing R1.4. Additionally, the energy-efficient hybrid routing schemes may
be comprised of two similar routing approaches. For instance, hybrid routing
scheme by Cano and Kim (2002) is comprised of two reactive
routing approaches, each having their own energy-efficient features. The rest
of this section presents how the node-level and network-level challenges have
been addressed by the existing hybrid routing schemes.
||Summary of routing characteristics and challenges of energy-efficient
hybrid routing schemes
|R1.1: Tree, R1.2: Cluster-based, R1.3: Zone-based, R1.4: Flat,
R2.1: Proactive, R2.2: Reactive, R3.1: Transmission distance, R3.2: Residual
energy, R3.3: Traffic load, R3.4: Location,R3.5: Node heterogeneity, P1:
Lower energy consumption, P2: Lower control overhead, P3: Lower end-to-end
delay, P4: Higher No. of alive nodes, P5: Higher packet delivery rate
Challenge C1.1: Routing Schemes with Longer Inactive Period. This challenge
has been addressed by incorporating scheduling into routing in order to increase
inactive period. Bernardos et al. (2006) proposed
a hybrid routing scheme, which is comprised of proactive R2.1 and reactive R2.2
routing approaches. Each node applies proactive routing to establish routes
to its two-hop neighbour nodes in order to avoid flooding otherwise reactive
routing is applied. An enhanced scheduling scheme is applied to assign time
slots to next-hop node so that the node can safely switch to inactive mode.
Hence, nodes that are not involved in packet forwarding can stay inactive as
long as possible to conserve energy.
Challenge C1.2: Routing Schemes for Addressing Different Levels of Residual
Energy among Nodes. This challenge is commonplace in clustered-based networks
R1.2 and it has been addressed by forming backbones of clusterheads, and rotating
the role of clusterhead among nodes based on their respective residual energy
(Sun et al., 2011; Tabibzadeh
et al., 2009; Zhu et al., 2010; He
and Wei, 2008; Muruganathan and Fapojuwo, 2008).
Tabibzadeh et al. (2009) and Zhu
et al. (2010) proposed a hybrid routing scheme which is comprised
of cluster-based routing R1.2 and flat routing R1.4. Due to the high traffic
load at clusterheads, a backbone of clusterheads is formed to relay packets
from member nodes to base station. This decreases transmission power and hence
energy consumption, of clusterheads. Flat routing is applied to establish routes
among the clusterheads. Similar approach has also been proposed by Muruganathan
and Fapojuwo (2008) to establish a backbone of clusterheads. The hybrid
routing scheme is comprised of cluster-based routing R1.2 and tree-based routing
R1.1 approaches. The tree-based routing establishes routes to base station using
a minimum spanning tree approach. Since clusterheads handle higher traffic load
compared to member nodes, the role of clusterhead is rotated among the nodes
(Tabibzadeh et al., 2009; Muruganathan
and Fapojuwo, 2008). This helps to achieve a balance in energy consumption
among the nodes.
Challenge C1.3: Routing Schemes for Addressing Higher Energy Consumption
of Nodes Closer to Base Station/Sink. Nodes closer to the base station may tend
to exhaust its residual energy causing link breakage. This challenge has been
addressed using flat routing so that these nodes do not forward packets all
times (Sun et al., 2011; Abdulla
et al., 2012; Priyankara et al., 2010).
Priyankara et al. (2010) proposed a location-aware
R3.4 hybrid routing scheme, which is comprised of flat routing R1.4 and cluster-based
routing R1.2. The network is comprised of high-end nodes and non-high-end nodes.
The high-end nodes have higher transmission energy for longer transmission range
and they can be selected as clusterheads to perform more tasks; while the non-high-end
nodes connect themselves to the high-end nodes in one or multiple hops in order
to reduce energy consumption. Nodes closer to the base station apply flat routing,
which relays packets in shorter transmission distance through multiple hops,
to reduce energy consumption. Nodes further from the base station perform cluster-based
routing. Abdulla et al. (2012) applied the similar
approach (Priyankara et al., 2010) in which flat
routing is applied to reduce energy consumption. Sun et
al. (2011) also applied the similar approach (Priyankara
et al., 2010) in which flat routing is applied to establish multiple
disjoint routes between clusterheads and base station in order to avoid only
certain links are used to send packets to base station.
Challenge C2.1: Routing Schemes for Addressing Dynamic Traffic Load.
This challenge has been addressed by using different routing approaches at different
traffic load levels (Abdulla et al., 2012; Figueiredo
et al., 2007). Figueiredo et al. (2007)
and Arabi (2010) proposed a hybrid routing scheme, which
is comprised of proactive R2.1 and reactive R2.2 routing approaches. The objective
of the routing scheme is to switch between reactive and proactive routing approaches
based on network traffic level. Whenever network traffic load is low, reactive
routing is used because of its low energy consumption nature (i.e., constant
broadcast of routing packets is not necessary). Whenever network traffic load
is high, the base station switches from reactive routing to proactive routing
to build a tree R1.1 throughout the entire network. By switching to proactive
routing, the base station avoids nodes from further initiating reactive routing,
and so this may reduce overall routing overheads.
Challenge C2.2: Routing Schemes for Addressing Higher Rate of Topological
Change. This challenge has been addressed by switching to proactive routing
when the rate of topological change is low (Zhao et al.,
2007). Zhao et al. (2007) proposed a zone-based
routing scheme R1.3, which is comprised of proactive R2.1 and reactive R2.2
routing approaches. Proactive and reactive routing approaches are applied to
establish routes among intra-zone nodes and inter-zone nodes, respectively.
There are two techniques to improve energy efficiency. Firstly, in inter-zone
routing, boundary nodes ensure that there is only a single Route Discovery message
being distributed to other zones. Secondly, in intra-zone routing, zones are
combined based on traffic condition and traffic pattern to form a larger proactive
routing area, which is more energy efficient. When the rate of topological change
is low, it is unnecessary to provide constant updates of routing information
because an established route can be reused, and so reactive routing R2.2 is
applied. On the other hand, when the rate of topological change is high, proactive
routing R2.1 is applied.
Challenge C2.3: Routing Schemes for Addressing Networks with Lower Node
Density. This challenge has been addressed by using a more appropriate routing
approach at different node density levels (He and Wei, 2008).
Additionally, a routing scheme can also avoid low-density area in order to avoid
high transmission power (Wang et al., 2006).
He and Wei (2008) proposed a hybrid routing scheme that
switches from cluster-based routing R1.2 to flat routing R1.4 as node density
decreases in a network in WSNs. In cluster-based routing, clusterheads aggregate
redundant data, however, when nodes are sparsely distributed, there is lesser
redundant data, and so the efficiency of data aggregation becomes lower. Due
to this reason, flat routing outperforms cluster-based routing when node density
Challenge C2.4: Routing Schemes for Reducing Routing Overhead. This
challenge has been addressed using two methods. Firstly, a more appropriate
routing approach is chosen to reduce routing overhead (Sato
et al., 2010; Kamboj and Sharma, 2009; Wang
et al., 2008; Chen et al., 2007; Helmy,
2005). Secondly, multiple routes are established from a source node to its
destination node to reduce routing overhead being incurred to re-establish a
new link when a current link is broken (Ren et al.,
2011). Sato et al. (2010), Kamboj
and Sharma (2009), Wang et al. (2008), Zhou
and Han (2007) and Helmy (2005) proposed a hybrid
routing scheme, which is comprised of proactive routing R2.1 and reactive routing
R2.2, to minimize flooding in order to reduce routing overhead in a zone-based
R1.3 network structure. Similar routing scheme has also been applied by Chen
et al. (2007) in cluster-based network structure R1.2. Proactive
routing is applied among intra-zone nodes, hence frequent updates about a nodes
changes (i.e., node joining and leaving), which incur routing overhead, is limited
within the zones. Reactive routing is applied among inter-zone nodes to establish
routes across different zones. A node sends unicast queries to a set of boundary
nodes of its zone rather than all boundary nodes of its zone in order to reduce
routing overhead. The boundary nodes, possess routing information of other zones,
relay the queries their respective boundary nodes. Therefore, the queries are
relayed among boundary nodes from one zone to another until the destination
node is found. Wang et al. (2008) applied an
ant-based technique. There are two types of ant packets based on their respective
traversing direction between a source and destination nodes, namely forward
ant and return ant. To establish a route, a source sends forward ants to its
destination, which subsequently replies with return ants. The ants deposit pheromone
(i.e., an average metric) on each intermediate node; and a route with higher
pheromone indicates a better route. In (Kamboj and Sharma,
2009) and (Zhou and Hou, 2007), location-aware R3.4
technique is applied in low- and medium-mobility networks. The location information
is applied to determine next-hop nodes towards the direction of destination
nodes. With reduced number of candidate next-hop nodes, the broadcast area of
reactive routing overhead is limited. The reactive routing also considers residual
energy of candidate nodes as one of the routing metrics.
Ren et al. (2011) proposed a hybrid routing
scheme, which is comprised of proactive routing R2.1 and reactive routing R2.2,
to minimize flooding in order to reduce routing overhead in a zone-based R1.3
network structure. The reactive routing approach applies an ant-based technique
to establish multiple candidate routes; while the proactive routing approach
maintains the dynamic routes. When there is a link breakage, alternative route
can be chosen immediately in order to reduce energy consumption caused by retransmission
and routing overhead. A node determines its next hop based on its residual energy
level, link quality and congestion; therefore, there is lesser possibility of
choosing next-hop nodes with lower residual energy.
Challenge C2.5: Routing Schemes for Achieving a Balance between End-to-end
Energy Consumption and Residual Energy of Nodes. This challenge has been addressed
by achieving a balance between end-to-end energy consumption and residual energy
of each node (Cano and Kim, 2002). Cano
and Kim (2002) proposed a routing scheme which is comprised of two different
kinds of reactive routing approaches: node-level and route-level. The route-level
reactive routing approach establishes route with minimum energy consumption
throughout the entire route; while the node-level reactive routing approach
chooses nodes with sufficient residual energy level throughout the entire route.
When the residual energy level of all the nodes along a candidate route is greater
than a threshold, the route-level approach is used to establish a route; and
when the residual energy level of any node in all candidate routes is lower
than the threshold, the node-level approach used to establish a route to avoid
node with a comparatively low residual energy.
PERFORMANCE ENHANCEMENT OF ENERGY-EFFICIENT HYBRID ROUTING SCHEMES
Table 1 shows the performance enhancement brought about by
the routing schemes compared to traditional or existing schemes. The performance
metrics applied in Table 1 are explained below:
||P.1: Lower energy consumption: Lower overall network
energy consumption prolongs network lifetime. There are three kinds of network
lifetime. Specifically, the time taken for: (1) the first node, (2) a certain
number of nodes and (3) all nodes to exhaust its (their) residual energy
||P.2: Lower control overhead: Lower amount of routing overhead
indicates lower energy consumption incurred by exchanging routing control
||P.3: Lower end-to-end delay: End-to-end delay is the time duration
taken for a packet to traverse from a source node to its destination node.
Lower end-to-end delay indicates higher successful transmission rate. This
also indicates lower number of retransmissions and hence lower energy consumption
||P.4: Higher number of alive nodes: A higher number of alive node
indicates lower number of nodes failure (i.e., exhausted their residual
energy) and so it indicates lower energy consumption
||P.5: Higher successful packet delivery rate: A higher packet delivery
success rate indicates lower number of retransmissions and hence lower energy
Threshold determination: Firstly, further research could be pursued
to determine a threshold, which may be dynamic in nature, based on network conditions
so that a hybrid routing scheme switches to a suitable routing approach at the
right time in order to reduce energy consumption. Each switch may be triggered
if some conditions (e.g., rate of topological change) are less than or greater
than a certain threshold. The threshold level may be dependent on multiple criteria,
particularly energy consumption. An accurate calculation of the threshold enables
a hybrid routing scheme to switch to the right choice of routing approach (i.e.,
proactive and reactive) at the right time. As an example, when the rate of topological
change is lower than a certain threshold, which indicates that an established
route is likely to be available, reactive routing is applied because route maintenance
may not be necessary, otherwise proactive routing is applied (Zhao
et al., 2007).
Multi-channel environment: Secondly, further research could be pursued
to investigate energy-efficient hybrid routing scheme in multi-channel environment.
To the best of our knowledge, most energy-efficient hybrid routing scheme has
been investigated in a single-channel environment. Multi-channel environment
reduces transmission interference and channel contention among neighbour nodes,
and so it reduces number of packet retransmissions. Higher interference and
channel contention may increase link failure, and so rerouting is necessary.
Hence, routing in multi-channel environment may reduce routing overhead and
Optimal cluster/zone size: Thirdly, further investigation could be pursued
to investigate the optimal cluster or zone size. Both cluster-based R1.2 and
zone-based R1.3 routings divide the entire network into clusters and zones,
respectively. Due to the similar concept, we refer to cluster and zone as group
henceforth. Different routing approaches may be applied to inter-group and intra-group
routings. The open issue is to determine an optimum group size in order to increase
energy efficiency of the entire network. For instance, Helmy
(2005) applied proactive routing R2.1 in intra-group routing, and reactive
routing R2.2 in inter-group routing in a zone-based routing. When a group is
too small, a node has limited knowledge about the entire network; and hence,
reactive routing, which may incur more routing packets, higher energy consumption
and routing delay, is applied to establish links. When a group is too large,
a node must maintain a larger routing table, which increases routing overhead
and energy consumption throughout the entire network.
Real world implementation: Fourthly, there has been lack of real-world
implementation of energy-efficient hybrid routing schemes, and so further research
could be pursued in this area. Most of the existing energy-efficient hybrid
routing schemes have been evaluated through simulation. Real-world implementations
are important to analyze the open issues and performance enhancements achieved
by the proposed schemes in practice. Environmental factors, such as hardware
limitation, may affect network performance of the proposed scheme. To the best
of our knowledge, there is only a single implementation of energy-efficient
hybrid routing scheme (Zhou and Hou, 2007). The hybrid
routing scheme is developed and tested on Atmel and Philips evaluation boards
running in an embedded workbench.
Fifthly, further investigation could be pursued to investigate the characteristics
of energy-efficient hybrid routing schemes. Table 1 shows
a wide range of potential routing characteristics (i.e., network structure,
routing approaches and routing decision factors) that can be further investigated
to address various challenges associated with energy-efficient hybrid routing
schemes. For example, the application of tree-based (Krishna
and Doja, 2012) and zone-based network structure have not been investigated
with respect to challenge C2.3 (or lower node density); while zone-based network
structure has not been investigated with respect to challenge C1.2 (or longer
inactive period). Further investigation can be performed to explore the applications
of various approaches in order to address the challenges, for example the scalability
of existing energy-efficient hybrid routing schemes (Nazir
et al., 2006).
Finally, further investigation could be pursued to investigate the performance
enhancement achieved by the energy-efficient hybrid routing scheme. Table
1 shows a wide range of potential performance enhancements that can be further
investigated to improve energy efficiency. For example, further investigation
can be performed to understand the effects of the hybrid routing schemes on
the amount of control overhead P.2 (Dan-Yang et al.,
2009) and end-to-end delay P.3 while addressing the challenge C2.3 (or lower
node density). Further investigation can also be performed to improve energy
efficiency in other kinds of networks, particularly low-power personal area
networks (Oliveira et al., 2011) and heterogeneous
networks (Cui et al., 2011). Improving security
of energy-efficient hybrid routing schemes can also be investigated to improve
the vulnerability of wireless networks (Sharma et al.,
2007; Alsaade, 2011).
Energy-efficient hybrid routing schemes have been shown to improve energy efficiency
of wireless multi-hop networks through combinations of several routing approaches,
particularly proactive and reactive approaches. With the recent development
of IEEE 802.11s, in which hybrid routing schemes have been incorporated into
the standard, energy-efficient hybrid routing scheme is expected to draw significant
research interests in the near future. This article provides an extensive review
on energy-efficient hybrid routing schemes, their characteristics, challenges
and performance enhancements. This article also discusses open issues associated
with the energy-efficient hybrid routing schemes. Certainly, there is a great
deal of future work in addressing the open issues raised in this article.
Abdulla, A.E.A.A., H. Nishiyama and N. Kato, 2012. Extending the lifetime of wireless sensor networks: A hybrid routing algorithm. Comput. Commun., 35: 1056-1063.
Al-Rawi, H.A.A. and K.A. Yau, 2012. Routing in distributed cognitive radio networks: A survey. Wireless Pers. Commun. 10.1007/s11277-012-0674-7
Alfawaer, Z.M., G.W. Hua, M.Y. Abdullah and I.D. Mamady, 2007. Power minimization algorithm in wireless ad-hoc networks based on PSO. J. Applied Sci., 7: 2523-2526.
CrossRef | Direct Link |
Alsaade, F., 2011. Proposing a secure and reliable system for critical pipeline infrastructure based on wireless sensor network. J. Software Eng., 5: 145-153.
CrossRef | Direct Link |
Arabi, Z., 2010. HERF: A hybrid energy efficient routing using a fuzzy method in wireless sensor networks. Proceedings of the International Conference on Intelligent and Advanced Systems, June 15-17, 2010, Kuala Lumpur, Malaysia, pp: 1-6.
Azcorra, A., T. Banniza, D. Chieng, J. Fitzpatrick and D. Von-Hugo et al., 2009. Supporting carrier grade services over wireless mesh networks: The approach of the European FP-7 STREP Carmen. IEEE Commun. Mag., 47: 14-16.
Banchs, A., N. Bayer, D. Chieng, A. de la Oliva and B. Gloss et al., 2008. CARMEN: delivering carrier grade services over wireless mesh networks. Proceedings of the IEEE 19th International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), September 15-18, 2008, France, pp: 1-6.
Bernardos, A.M., J.R. Casar and P. Tarrio, 2006. Efficient social routing in sensor fusion networks. Proceedings of the 9th International Conference on Information Fusion, July 10-13, 2006, Florence Italy, pp: 1-8.
Cano, J.C. and D. Kim, 2002. Investigating performance of power-aware routing protocols for mobile ad-hoc networks. Proceedings of the International Mobility and Wireless Access Workshop, October 12, 2002, Texas, USA., pp: 80-86.
Chen, J.L., H.W. Tzeng and C.P. Lai, 2007. Routing mechanism for reliable sensor network applications. Proceedings of the 2nd International Conference on Systems and Networks Communications, August 29-31, 2007, Cap Esterel, pp: 34-.
Chieng, D., D.V. Hugo and A. Banchs, 2010. A cost sensitiviy analysis for carrier grade wireless mesh networks with Tabu optimization. Proceedings of the INFOCOM IEEE Conference on Computer Communications Workshops, March 15-19, 2010, San Diego, USA., pp: 1-6.
Cui, Y., Y. Xu, R. Xu and X. Sha, 2011. A multi-radio packet scheduling algorithm for real-time traffic in a heterogeneous wireless network environment. Inform. Technol. J., 10: 182-188.
CrossRef | Direct Link |
Dan-Yang, Q., M. Lin, S. Xue-Jun and X. Yu-Bin, 2009. SRRG: An effective self recovery routing game for mobile ad hoc network. Inform. Technol. J., 8: 1006-1012.
CrossRef | Direct Link |
Figueiredo, C.M.S., E.F. Nakamura, A.A.F. Loureiro and L.B. Ruiz, 2007. An event-detection estimation model for hybrid adaptive routing in WSNs. Proceedings of the IEEE International Conference of Communications, August 24-28, 2007, Glasgow, Scotland, pp: 3887-3894.
He, G. and Z. Wei, 2008. A hybrid routing scheme for wireless sensor network. Proceedings of the IEEE International Conference on Networking, Sensing and Control, April 6-8, 2008, Sanya, China, pp: 961-965.
Helmy, A., 2005. Contact-extended zone-based transactions routing for energy-constrained wireless ad hoc networks. IEEE Trans. Veh. Technol., 54: 307-319.
Hiertz, G.R., D. Denteneer, S. Max, R. Taori, J. Cardona, L. Berlemann and B. Walke, 2010. IEEE 802.11s: The WLAN mesh standard. IEEE Wireless Commun., 17: 104-111.
Kamboj, P. and A.K. Sharma, 2009. Location aware reduced diffusion hybrid routing algorithm (LARDHR). Proceedings of the 2nd International Conference on Emerging Trends in Engineering and Technology, December 16-18, 2009, Nagpur, India, pp: 1156-1161.
Krishna, M.B. and M.N. Doja, 2012. Analysis of tree-based multicast routing in wireless sensor networks with varying network metrics. Int. J. Commun. Syst., (In Press). 10.1002/dac.1400
Kwong, K.H., A.T.K. Ngoh, D.H.T. Chieng and M. Abbas, 2011. WiFied up Malaysia villages: A case study of using WiFi technology to increase internet penetration rate in Malaysia rural areas. Proceedings of the IEEE 10th Malaysia International Conference on Communications, October 2-5, 2011, Kota Kinabalu, Sabah, pp: 7-12.
Lin, Y.D., S.L. Tsao, S.L. Chang, S.Y. Cheng and C.Y. Ku, 2010. Design issues and experimental studies of wireless LAN mesh. IEEE Wireless Commun., 17: 32-40.
Loziak, K., M. Natkaniec, J. Gozdecki, E. Chin, D. Chieng and V. Teh, 2010. Monitoring system for carrier grade MESH networks. Proceedings of the Future Network and Mobile Summit, June 16-18, 2010, Florence, Italy, pp: 1 -8.
Muruganathan, S.D. and A.O. Fapojuwo, 2008. A hybrid routing protocol for wireless sensor networks based on a two-level clustering hierarchy with enhanced energy efficiency. Proceedings of the IEEE Wireless Communications and Networking Conference, March 31-April 3, 2008, Las Vegas, NV., USA., pp: 2051-2056.
Nazir, M.B., M.S.H. Khiyal and T.U. Rahman, 2006. Scalability of zone routing protocol extensions for mobile Ad-hoc networks. Inform. Technol. J., 5: 373-385.
CrossRef | Direct Link |
Nurul Huda, M., F. Yasmeen, E. Kamioka and S. Yamada, 2007. Optimal path selection in MANET considering network stability and power cost. Inform. Technol. J., 6: 1021-1028.
CrossRef | Direct Link |
Oliveira, L.M.L., A.F. de Sousa and J.J.P.C. Rodrigues, 2011. Routing and mobility approaches in IPv6 over LoWPAN mesh networks. Int. J. Commun. Syst., 24: 1445-1466.
Priyankara, S., K. Kinoshita, H. Tode and K. Murakami, 2010. A clustering/multi-hop hybrid routing method for wireless sensor networks with heterogeneous node types. Proceedings of the IEEE GLOBECOM Workshops, December 6-10, 2010, Miami, FL., USA., pp: 207-212.
Ren, J., Y. Tu, M. Zhang and Y. Jiang, 2011. An ANT-based energy-aware routing protocol for ad hoc network. Proceedings of the International Conference on Computer Science and Service System, June 27-29, 2011, Nanjing, China, pp: 3844-3849.
Sato, Y., A. Koyama and L. Barolli, 2010. A zone based routing protocol for ad hoc networks and its performance improvement by reduction of control packets. Proceedings of the International Conference on Broadband, Wireless Computing, Communication and Applications, November 4-6, 2010, Fukuoka, Japan, pp: 17-24.
Serrano, P., P. Patras, X. Perez-Costa, B. Gloss and D. Chieng, 2009. A MAC layer abstraction for heterogeneous carrier grade mesh networks. Proceedings of the ICT-Mobile Summit, June 10-12, 2009, Santander, Spain -.
Sharma, S., R.C. Jain and S.S. Bhadauria, 2007. SBEVA: A secured bandwidth efficient variance adaptive routing protocol for mobile Ad hoc network. Asian J. Inform. Manage., 1: 1-10.
CrossRef | Direct Link |
Shi, Z., Z. Pu and Z.Q. Yu, 2010. A routing protocol based on energy aware in ad hoc networks. Inform. Technol. J., 9: 797-803.
CrossRef | Direct Link |
Simsek, B., J.Y. Chen, D. Chieng, M. Natkaniec and K. Loziak, 2009. Self Configuration Architecture for Carrier Grade Mesh Network (CARMEN). Proceedings of the ICT-Mobile Summit, June 10-12, 2009, Santander, Spain -.
Sun, N., Y.B. Cho and S.H. Lee, 2011. A distributed energy efficient and reliable routing protocol for wireless sensor networks. Proceedings of the IEEE 14th International Conference on Computational Science and Engineering, August 24-26, 2011, Dalian, China, pp: 273-278.
Tabibzadeh, M., M. Sarram and F. Adibnia, 2009. Hybrid routing protocol for prolonged network lifetime in large scale wireless sensor network. Proceedings of the International Conference on Information and Multimedia Technology, December 16-18, 2009, Jeju Island, South Korea, pp: 179-183.
Thangammal, C.B., P. Rangarajan and J.R.P. Perinbam, 2012. Maximization of wireless sensor network's lifetime using losningen cross-layer approach. Asian J. Sci. Res., 5: 133-142.
CrossRef | Direct Link |
Ting, A., D. Chieng and K.H. Kwong, 2011. Capacity and coverage analysis of rural multi-radio multi-hop network deployment using IEEE802.11n Radios. Proceedings of the IEEE 10th Malaysia International Conference on Communications (MICC), October 2-5, 2011, Kota Kinabalu, Sabah, pp: 77-82.
Tingrui, P., Z. Zhi, Z. Wenli and Z. Zhaoxia, 2011. A cognitive improved hierarchical AODV routing protocol for cognitive wireless mesh network. Inform. Technol. J., 10: 376-384.
CrossRef | Direct Link |
Viswacheda, D.V., L. Barukang, M.Y. Hamid and M.S. Arifianto, 2007. Performance evaluation of mobile ad hoc network based communications for future mobile tele-emergency system. J. Applied Sci., 7: 2111-2119.
CrossRef | Direct Link |
Wang, J., E. Osagie, P. Thulasiraman and R.P. Thulasiram, 2008. HOPNET: A hybrid ant colony optimization routing algorithm for mobile ad hoc network. Ad Hoc Networks, 7: 690-705.
Wang, T., S. Hao, P. Wang and G. Peng, 2006. Efficient and density-aware routing for wireless sensor networks. Proceedings of the 15th International Conference on Computer Communications and Networks, October 9-11, 2006, Arlington, USA., pp: 207-212.
Wei, D., H.A. Chan and M.B. Rawoot, 2007. Hybrid routing protocol to decrease delay and to extend lifetime for mesh networks. Inform. Technol. J., 6: 518-525.
CrossRef | Direct Link |
Zhao, Q., L. Tong and D. Counsil, 2007. Energy-aware adaptive routing for large-scale Ad Hoc networks: Protocol and performance analysis. IEEE Trans. Mobile Comput., 6: 1048-1059.
Zhou, H. and K. Hou, 2007. CIVIC: An power- and context-aware routing protocol for wireless sensor networks. Proceedings of the International Conference on Wireless Communications, Networking and Mobile Computing, September 21-25, 2007, Shanghai, China, pp: 2771-2774.
Zhu, A.D., W.M. Zhao and J.G. Xing, 2010. A novel hybrid routing model for wireless grain depot surveillance system. Proceedings of the IEEE/ASME International Conference on Mechatronics and Embedded Systems and Applications, July 15-17, 2010, Qingdao, China, pp: 263-268.