Minimizing End-to-End Delay: A Novel Routing Metric for Multi-Radio Wireless Mesh Networks

Conference Paper (PDF Available)inProceedings - IEEE INFOCOM · April 2009with62 Reads
DOI: 10.1109/INFCOM.2009.5061905 · Source: DBLP
Conference: INFOCOM 2009. 28th IEEE International Conference on Computer Communications, Joint Conference of the IEEE Computer and Communications Societies, 19-25 April 2009, Rio de Janeiro, Brazil
Abstract
This paper studies how to select a path with the minimum cost in terms of expected end-to-end delay (EED) in a multi-radio wireless mesh network. Different from the previous efforts, the new EED metric takes the queuing delay into account, since the end-to-end delay consists of not only the transmission delay over the wireless links but also the queuing delay in the buffer. In addition to minimizing the end-to-end delay, the EED metric implies the concept of load balancing. We develop EED- based routing protocols for both single-channel and multi-channel wireless mesh networks. In particular for the multi-radio multichannel case, we develop a generic iterative approach to calculate a multi-radio achievable bandwidth (MRAB) for a path, taking the impacts of inter/intra-flow interference and space/channel diversity into account. The MRAB is then integrated with EED to form the metric of weighted end-to-end delay (WEED). As a byproduct of MRAB, a channel diversity coefficient can be defined to quantitatively represent the channel diversity along a given path. Both numerical analysis and simulation studies are presented to validate the performance of the routing protocol based on the EED/WEED metric, with comparison to some well- known routing metrics.

Full-text (PDF)

Available from: Weihua Zhuang, Sep 02, 2014
Minimizing End-to-End Delay: A Novel Routing
Metric for Multi-Radio Wireless Mesh Networks
Hongkun Li, Yu Cheng, Chi Zhou
Department of Electrical and
Computer Engineering
Illinois Institute of Technology
{hli55, cheng, zhou}@iit.edu
Weihua Zhuang
Department of Electrical and
Computer Engineering
University of Waterloo
wzhuang@uwaterloo.ca
Abstract—This paper studies how to select a path with the
minimum cost in terms of expected end-to-end delay (EED) in a
multi-radio wireless mesh network. Different from the previous
efforts, the new EED metric takes the queuing delay into account,
since the end-to-end delay consists of not only the transmission
delay over the wireless links but also the queuing delay in the
buffer. In addition to minimizing the end-to-end delay, the EED
metric implies the concept of load balancing. We develop EED-
based routing protocols for both single-channel and multi-channel
wireless mesh networks. In particular for the multi-radio multi-
channel case, we develop a generic iterative approach to calculate
a multi-radio achievable bandwidth (MRAB) for a path, taking
the impacts of inter/intra-flow interference and space/channel
diversity into account. The MRAB is then integrated with EED
to form the metric of weighted end-to-end delay (WEED). As
a byproduct of MRAB, a channel diversity coefficient can be
defined to quantitatively represent the channel diversity along a
given path. Both numerical analysis and simulation studies are
presented to validate the performance of the routing protocol
based on the EED/WEED metric, with comparison to some well-
known routing metrics.
I. INTRODUCTION
Routing in wireless mesh networks has been a hot re-
search area in recent years, with the objective to achieve
as high throughput as possible over the network. The main
methodology adopted by most of the existing work is selecting
path based on interference-aware or load-balancing routing
metrics to reduce network-wide channel contentions. It has
been revealed that the capacity of a single-radio multi-hop
wireless network can not scale up with the network size,
due to the co-channel interference [1]–[3]. The multi-radio
multi-channel connection has been widely considered as an
efficient approach to increase the wireless network capacity
[8]. Design of efficient routing schemes for multi-radio multi-
channel wireless mesh network is much more challenging
compared to the single-channel case.
Many popular multimedia applications, e.g., voice over IP,
IPTV, and on-line gaming, have strict delay requirement. In
this paper, we aim at designing a routing metric to minimize
the end-to-end delay, considering not only the transmission
delay at the medium access control (MAC) layer, but also
the queuing delay at the network layer. Most of the previous
studies focus only on the transmission delay of the packet
This work was supported in part by NSF grant CNS-0832093.
XY
S
ABC
D
M = 10 M = 10
M = 2 M = 2 M = 3
1
1
1
10.1
0.1
0.5
Fig. 1. The impact of queuing delay on path selection.
being served at the MAC layer [13], [15], while in many cases
the queuing delay takes a significant portion of the total delay
over a hop. The delay through a node, which has many packets
in queue but short transmission time, could be larger than
through the one, which has less packets in the queue but longer
transmission delay.
We here use an example, as illustrated in Fig. 1 to emphasize
the impact of network-layer queuing delay on routing. The
number annotating each link is the success probability for
a transmission over the link, denoted as psuc, which means
on average it takes 1/psuc transmission trails to successfully
deliver a packet. The number Mdenotes the number of packets
in the network-layer queue, waiting to be served by the MAC
layer. Suppose that the bandwidth of each link is 11Mbit/s and
the packet length is 1100bytes; it gives a transmission time of
0.8ms. If the queue delay is not included, routing based on the
expected transmission time (ETT) would prefer the path S-X-
Y-D (9.6ms) over the path S-A-B-C-D (11.2ms). Nevertheless,
the path S-A-B-C-D would be the better one with the queuing
delay taken into account. In this case, the end-to-end delay
over S-X-Y-D is 97.6ms, but only 24 ms over S-A-B-C-D. In
this example, the delay values ignore the backoff overhead,
which will be considered in our routing metric design.
The newly proposed routing metric of end-to-end delay
(EED) in fact exploits the cross-layer design: each node needs
to not only monitor the transmission failure probability at
the MAC layer to estimate the MAC transmission delay, but
also count the number of packets waiting in the network-
layer buffer to estimate the queuing delay. The EED metric
also implies the concept of load-balancing. The path with
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE INFOCOM 2009 proceedings.
978-1-4244-3513-5/09/$25.00 ©2009 IEEE
Advertisement:
minimum EED normally passes through the links with less
packets in the queue, and thus balances the traffic off those
congested links. Moreover, counting the number of packets in
the buffer is a convenient implementation; most of the existing
load-balancing routing schemes require the traffic information
available, which is usually not easy to obtain in practice [16].
In addition to the transmission delay and queuing delay
at each hop, the end-to-end delay over a multi-hop wireless
network is particularly impacted by the interferences among
different hops, which can be classified into inter-flow and intra-
flow interference [23]. In this paper, we further propose a path
metric called multi-radio achievable bandwidth (MRAB) to
accurately capture the impacts of inter/intra-flow interferences
and space/channel diversity along a path. We consider a
practical scenario that an end-to-end path may consist of
both multi-radio hops and single-radio hops, where different
channels do not interfere with each other but interferences exist
within the same channel. We particularly develop a sub-path
based iterative approach to model the complex interactions
among inter-flow interference, intra-flow interference, and
simultaneous transmission due to space and channel diversity.
The MRAB is then integrated with EED to form the metric
of weighted end-to-end delay (WEED). As a byproduct of
MRAB, a channel diversity coefficient can be defined to
quantitatively represent the channel diversity along a given
path. We evaluate the performance of the WEED based routing
protocol via numerical analysis and ns2 simulations, with
comparison to some popular metrics, under both single and
multiple channel cases. It is confirmed that the EED/WEED
metric consistently yields better performance.
The reminder of this paper is organized as follows: Sec-
tion II reviews more related work. Section III derives the
routing metric of EED. Section IV presents the algorithm to
compute the MRAB, which captures the interaction between
the inter- and intra-flow interferences. The MRAB metric is
integrated with the EED metric to form the WEED metric
for routing over the multi-radio mesh networks. The routing
protocol is described in Section V. Section VI presents some
numerical analysis and simulation studies to validate the
routing performance based on the EED/WEED metric, with
comparison to some well-known routing metrics. Section VII
gives the concluding remarks.
II. RELATED WORK
The routing metric plays a critical role in a routing protocol.
The studies in [8], [16], [17] design routing metrics for load-
balancing in a multi-hop wireless network. The routing metrics
however require the real-time traffic information. To exploit
the space diversity, the link conflict graph is normally applied
to model the interference among different hops [18], and
the interference clique transmission time is proposed as a
routing metric in [20]. However, the conflict graph based
approaches normally induce large computation overhead in
searching for the maximal independent sets or cliques, and
are not suitable for dynamic distributed routing protocols.
De Couto et al. propose the metric of expected transmission
count (ETX) [21] to describe the channel contention level
experienced by a wireless link, which works well in a ho-
mogeneous single-radio environment. However, ETX is not
capable of describing the complex scenarios in a multi-radio
wireless mesh network, normally involving inter-/intra-flow
interferences and different rate/intererence/topology profiles
over different channels. mETX and ENT [6] are proposed to
enhance ETX by considering the variable link reliability. The
ETOP metric enhances ETX by incorporating the impact of
link positions [5].
A bandwidth-aware routing with QoS requirement is pro-
posed in [26]. The link metric of expected transmission time
(ETT) and the associated path metric of weighted cumulative
ETT (WCETT) are proposed in [13] for multi-channel mesh
networks, which try to enhance the ETX by counting the
heterogeneous channel rate and intra-flow interference, but
the inter-flow interference is still not considered. Furthermore,
when calculating the intra-flow interference, WCETT always
takes all links into account and overlooks the situation that two
links far away enough can transmit packets simultaneously.
The metric of interference and channel switching (MIC)
[15] incorporates both inter-flow and intra-flow interference,
whereas it only contains the number of interfering nodes
rather than the total amount of interference on these nodes
for the inter-flow interference. In [25], we propose a metric
of multi-hop effective bandwidth (MHEB) to compute the
usable bandwidth when both inter- and intra-flow interferences
are present. However, the MHEB metric just uses a simple
weighted average to combine the inter- and intra- flow in-
terferences. In this paper, the MRAB is based on MHEB, but
use a more accurate approach to capture the complex interplay
between the two types of interferences.
III. END-TO-END DELAY METRIC
The end-to-end delay over a path is the summation of delays
experienced by all the hops along the path. For convenience,
we also use EED to denote the delay metric at each link.
The meaning of EED will be clear in the context. In order
to compute the EED metric over a wireless channel, each
node needs to monitor the number of packets buffered at
the network layer waiting for MAC layer service, as well
as measuring the transmission failure probability at the MAC
layer. The transmission failure probability is the probability
that a MAC-layer transmission fails due to either collisions or
bad channel quality. While counting the number of packets in
the queue is straightforward, we will discuss how to measure
the transmission failure probability over a link in Section V.
The EED over a link i, say between node niand ni+1, consists
of the queuing delay and transmission delay as
EEDi=E[queuing delay +transmission delay].(1)
The transmission delay can also be interpreted as the packet
service time, which is defined as the period from the instant
that a packet begins to be serviced by the MAC layer to the
instant that it is either successfully transmitted or dropped after
a predefined number of retransmissions.
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE INFOCOM 2009 proceedings.
Suppose that the 802.11 distributed coordination function
(DCF) MAC protocol is used, each transmission or retrans-
mission includes protocol overhead due to the binary backoff
mechanism [28]. Let pidenote the transmission failure prob-
ability over link i, and assume it is stable through all the
retransmissions of the packet. Also, let Tidenote the packet
service time over link i, and Kthe maximum number of
retransmissions. The average transmission delay
E[Ti]=
K+1
k=1
pk1
i(1 pi)I{k<K+1}
k
j=1 E[Wj]+ L
B
(2)
where Wjdenotes the contention window at the jth backoff
stage, and L,Bdenote the packet length and link bandwidth,
respectively. According to the 802.11 standard [7], [28],
Wj=2
j1Wmin if ignoring the constraint of backoff stage,
and E[Wj]=Wj1
2. In (2), the indicator I(A)is equal to 1 if
Ais true, which is incurred to include the case that a packet is
dropped when the retransmission limit is reached. In addition,
the MAC overhead due to acknowledgement is incorporated
into the packet length Lfor convenience. After applying some
manipulations over (2), we can get
E[Ti]= L
B1pK
i
1pi+E[backoff time](3)
with
E[backoff time]=
K+1
k=1
pk1
i(1 pi)I{k<K+1}
k
j=1
E[Wj]
=Wmin 1(2pi)K+1
2(1 2pi)1pK
i
2(1 pi).(4)
If there are Mipackets in the queue when a new packet
reaches node ni, the EED metric can be defined as
EEDi=(Mi+1)E[Ti](5)
which means the total delay passing through the hop equals
the MAC service time of those packets queuing ahead of the
new packet plus the MAC service time of the new packet
itself. Note that the EED value in (5) implies the memoryless
property of the packet service time, as the head-of-line packet
may only need to finish a residue packet service time when
the new packet comes in. It is well-known that only an
exponentially distributed services time has the memoryless
property. It has been demonstrated in [30] that the MAC packet
service time over a 802.11 DCF can indeed be approximated
by an exponential random variable.
Consider an end-to-end path including Hhops, the EED
metric for the path is defined as
EED =
H
i=1
EEDi.(6)
We would like to emphasize that the EED given in (6) does
not consider the effect of co-channel interference in the multi-
hop wireless networks, which assumes that all the packets can
continuously go through the path hop-by-hop. However, in
a multi-hop wireless network, if two links working over the
same channel are located close, when one link is in trans-
mission, the MAC protocol will freeze the other link. Such
channel freezing can be due to either intra-flow transmissions
or inter-flow transmissions, which results in extra delays in
addition to the basic EED as shown in (6). In the following
section, we will discuss how to extend the basic EED with the
co-channel interferences taken into account.
IV. ACHIEVABLE BANDWIDTH OVER A MULTI-RADIO
MULTI-CHANNEL PATH
In this section, we will develop an algorithm to compute the
achievable bandwidth along a multi-radio multi-channel path,
termed as multi-radio achievable bandwidth, by capturing the
complex interplay between inter-flow and intra-flow interfer-
ences. The end-to-end delay over a multi-radio multi-channel
path can be described more accurately by incorporating the
MRAB metric into the EED computation to form a new metric
WEED. A side-effect benefit of MRAB analysis is that a
channel diversity coefficient can be defined to quantify the
resource consumption along a multi-radio multi-channel path.
A. Multi-Radio System
We consider a wireless mesh network, where each node is
equipped with one or more radio interfaces. The interfaces
assigned with different channels, located either in the same
node or in different nodes, can be active simultaneously.
Thus, the network throughput could be significantly improved
compared with single-radio system [8]. The interfaces work-
ing on different channels would form distinct interference
topologies. Channel assignment [10] plays an critical role
in determining the interference topology, and then impacts
on system performance. The channel assignment itself is a
challenging research topic, which is out of the scope of this
paper. We assume that the channel assignment for each node
is given. All the nodes are stationary, and any one can be used
as a router. We define the transmission range of a node as one
hop, while the interference range is r(2) hops. We consider
that the WMN operates over the IEEE 802.11 based MAC, and
assume that the routing control information exchanged among
neighbor nodes is error free.
We adopt the physical model presented in [18] to describe
the interference among different hops. Such an interference
model indicates that a transmission from node uto vis
successful if the signal to interference and noise ratio (SINR)
at receiver vis above the pre-determined threshold γ, i.e.,
Pv(u)
N+kvPv(k)γ(7)
where Ndenotes the background noise, Pv(u)the received
power at node vfrom node u,vthe set of nodes located in
interference range of v, and Pv(k)the interference power from
an interfering node k.
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE INFOCOM 2009 proceedings.
B. Multi-Radio Achievable Bandwidth
1) Inter-flow interference: we first compute the achievable
bandwidth under the inter-flow interference (ABITF). Every
node can monitor the received power to infer the magnitude
of the inter-flow interference around its neighborhood. Based
on the interference model (7), the SINR threshold implicitly
denotes the maximum interference a node could tolerate to
process a successful communication. We define the interfer-
ence degree ratio IDRi(uv)for link ibetween uand vas:
IDRi(uv)=kvPv(k)
Pmax
.(8)
The ratio reflects the utilization of the channel assigned to
link i.Pmax is the maximum tolerable interference power at
receiver and can be calculated by (7). kvPv(k)is sum of
undesired powers at node vfrom other transmissions. Note
that if there is no interference, the IDR is 0, implying that
entire bandwidth of this channel is available for link i.Onthe
contrary, an IDR of 1 will indicate that the channel has been
fully occupied by other links, and no additional bandwidth
can be allocated for link iuntil the ratio becomes less than 1.
Based on this definition, we evaluate the ABITF at link ias
ABI T Fi=(1 IDRi)Bi
ETXi
(9)
where Bidenotes the physical bandwidth of link i, and ETXi
[21] denotes the expected transmission attempts to achieve a
successful transmission over link i.Thevalue(1 IDRi)Bi
indicates the available bandwidth for a transmission under the
inter-flow interference. The physical meaning of (9) can be
interpreted as: given the transmission failure probability pi,
a successful transmission needs ETXiattempts in average;
the bandwidth is effectively used for only one of the ETXi
transmission attempts.
2) Intra-flow interference: There exists intra-flow interfer-
ence if two links belonging to the same path work on the
same channel and are located within each other’s interference
range, i.e. within r(2) hops. We define a new concept of
sub-path spanning r+2 hops, based on the observation that a
link will potentially interfere with another link at most r+2
hops away. A sub-path with r=2is illustrated in Fig.2. In
general, a H-hop path contains Hr1sub-paths.
Under the impact of intra-flow interference, a sub-path
is equivalent to a virtual link. The reason is that a new
packet can enter a sub-path only after the previous one leaves.
The achievable bandwidth over a sub-path can be iteratively
obtained from the achievable bandwidth over two interfering
links. For example, consider two neighboring co-channel links
iand jalong a path. Links iand jhave bandwidth Bi
and Bj, respectively. Since the two links can not be active
simultaneously, the equivalent achievable bandwidth under the
intra-flow interference (ABIRF) over links iand j, denoted as
ABI RF (ij), satisfies
L
ABI RF (ij)=L
Bi
+L
Bj
(10)
A B C D E
1213
Fig. 2. Illustration of a multi-radio multi-channel path.
where Ldenotes the packet length. It can then be obtained
that
ABI RF (ij)= BiBj
Bi+Bj
.(11)
Extending the ABI RF (ij)result to the whole sub-path can be
iteratively implemented. In each iteration, consider those links
having been processed as one virtual link, with the bandwidth
set as the ABIRF value already obtained; and then apply the
computation of (11) over the virtual link and the next-hop
link. Note that the impact of inter-flow interference on path
capacity can be conveniently integrated with the intra-flow
interference impact by using the ABITF value given by (9)
as link capacity in ABIRF computation, instead of using the
physical bandwidth.
3) Multi-radio achievable bandwidth: The multi-radio
multi-channel connection makes the capacity analysis of a
sub-path more complicated. When two links work on dif-
ferent channels through different radio interfaces, they could
send/receive packets simultaneously without interference. It is
possible that the two end-hops of a sub-path are co-channel
links, while other hops in the middle may work on different
channels. The iterative procedure discussed above to compute
the ABIRF for a co-channel sub-path can also be extended
to the multi-channel sub-path, with consideration that the
achievable bandwidth over two links is min(Bi,B
j), when
link iand jwork over different channels. Specifically, the
iterative steps to compute the ABIRF for a given multi-radio
sub-path is as follows:
Step 1 : For the first link of the sub-path, set ABI RF equal
to the ABITF associated with the channel over which
the link works.
Step 2 : Go to the next link, say link i, and check whether
the channel of the next link is different from those
used by the previous links or not. If different, go to
step 3; otherwise go to step 4.
Step 3 : In this case, set
ABI RF = min(ABI RF, AB IT Fi)(12)
and then go to step 5.
Step 4 : In this case, set
ABI RF =ABI RF ×AB IT Fi
ABI RF +ABI T Fi
(13)
and then go to step 5.
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE INFOCOM 2009 proceedings.
Step 5 : Check whether all the (r+2) hops of the sub-path
have been considered or not. If not, go to step 2;
otherwise, terminate the iteration.
For a H-hop path including multiple sub-paths, let
ABI RFjdenote the achievable bandwidth over the jth sub-
path. The multi-radio achievable bandwidth can be computed
as:
MRAB = min (ABI RFj)(14)
for j=1,2, ...H r1.IfHr10,wesetj=1,
which means the path is so short that there is only one sub-
path along the whole path. The computation in (14) exploits
the bottleneck concept, but is applied at the sub-path level
instead of the link level.
C. WEED Metric
1) Weighted end-to-end delay: Over a multi-radio multi-
channel path, the MRAB metric is integrated with the EED
metric to form a weighted end-to-end delay metric as:
WEED =α×
H
i=1
EEDi+(1α)×NP·L
MRAB (15)
where 0α1is tunable weight factor. In the WEED
metric, the first part represents the accumulation of the delivery
delay due to hop-by-hop transmissions in a store-and-forward
manner; the second part represents the extra delay due to
the interference nature of a multi-hop wireless network. In
particular, the NPinduced in (15) denotes the total number
of packets queued in the buffers along the path, because the
interference effect described by the MRAB applies to all the
packets that are being served by the path.
We would like to emphasize that the WEED metric contains
not only the end-to-end delay information regarding a single
packet transmission, but also the transmission delay for a block
of packets due to the bottleneck bandwidth MRAB. Therefore,
selecting a shortest path based on the WEED metric tends to
minimize both the short-term and the long-term delay.
2) Monotonicity analysis: It has been indicated in [11]
that monotonicity is one of the necessary properties of a
routing metric to result in a consistent and loop-free routing
implementation. For example, the well-known WCETT metric
[13] is monotonic. We here prove that WEED also has the
property of monotonicity.
Consider a given path, it is obvious that the H
i=1 EEDi
part will becomes larger when one more hop is attached to
the path. For the MRAB part, one more hop may lead to two
possible cases. In the first case, the number of sub-paths does
not change; the new hop just makes the (Hr1)th sub-
path one hop longer. According to the iterative computation
given in Section IV-B , the ABIRF of a sub-path is less
than or equal to the previous value when one more hop is
included. Thus, the NP·L
MRAB part achieves a larger/equal value
when the path goes longer. In the second case, a new sub-
path is incurred by the new hop, while the existing sub-paths
will not change. The “minimization” operation in obtaining
the MRAB value guarantees that MRAB will not increase
when one more sub-path is generated. Hence, NP·L
MRAB part
achieves a larger or equal value too in the the second case,
when one hop is involved. Since both parts constituting the
WEED metric becomes larger or maintains equal when the
path goes longer, WEED is monotonic. The proof applies to
both left-monotonicity and right-monotonicity.
D. Channel Diversity Coefficient
A challenging issue being widely studied in the area of
multi-channel wireless networks is how to quantify the chan-
nel diversity for a given path. Intuitively, an ideal quantity
describing the channel diversity incorporates the impacts from
various aspects, including the number of hops, the number of
channels, and the interference relationship among the links.
Our development efforts in the above have demonstrated that
the MRAB metric indeed takes all of the factors into account.
Therefore, we define a channel diversity coefficient (CDC)
based on the MRAB as
CDC =MRAB
Bs
(16)
where Bsdenotes the achievable bandwidth of the same path
if all links of this path work over the same channel, defined
as the single-channel path capacity. For the convenience of
comparison, we adopt the smallest ABI T Fivalue as the
single-channel path capacity Bs. Thus, CDC is always larger
than or equal to 1, and a higher CDC indicates a better channel
diversity.
V. ROUTING PROTOCOL DESIGN
A. Basic DSR Implementation
We implement the proposed EED/WEED based routing
by modifying the dynamic source routing (DSR) protocol
[12]. We select DSR due to the following reasons: (i) DSR
is one of the most popular protocols in multi-hop wireless
networks, and the implementation codes are publicly available.
(ii) We have noticed that some well-known routing metrics,
such as ETX, ETT and WCETT, are implemented based on
DSR; a common implementation will significantly facilitate
the performance comparison among different routing metrics.
(iii) The EED/WEED metric does not have the property of
isotonicity [11], so the source routing approach (adopted by
DSR) is preferred to guarantee the optimality, consistency, and
loop-freeness in routing.
With basic DSR, a node attempts to find a route to a given
destination by initiating a route request (RREQ) message.
Every RREQ has a unique broadcast ID to prevent routing
loops and redundant flooding. An intermediate node will
further broadcast a RREQ only when the broadcast ID appears
for the first time; also the node will insert its address in the
source route field of the RREQ message. Once the destination
receives the RREQ, it will reverse the hop sequence of the
received path and insert the reversed path into the source route
field of a route reply (RREP) message, which is then unicasted
back to the source node. The source node will determine a
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE INFOCOM 2009 proceedings.
shortest path to a destination based on the path information it
receives from the RREP messages. With DSR, all the RREP
messages and data packets carry a complete path between a
certain source/destination pair in their source route fields. All
the nodes can overhear such path information and store it in
their route caches for later use. In cases that a node finds that
a packet could not be successfully delivered over a link after
a maximum number of retransmission, the node will return
aroute error (RRER) to the source node of the path; every
intermediate node receiving the RRER message will mark this
link invalid.
B. EED/WEED Based Routing
1) EED link metric: The EED metric itself can be used
as a routing metric, especially for single-channel multi-hop
networks. We consider that a directional link is defined by an
upstream end and a downstream end, and the communication
between two neighbor nodes is through two directional links.
To obtain the EED link metric, a downstream node needs to
monitor the transmission failure probability over the link and
knows the number of packets in the upstream node’s buffer. In
our routing protocol, each node periodically broadcasts probe
packets to its downstream neighbors at a predetermined rate
λ, with the number of packets Min its buffer carried in
each probe. Each downstream node maintains a neighbor list.
When the downstream node receives a probe packet, it will
update the value Mfor the corresponding upstream node in
its neighbor list. Moreover, a downstream node will count the
number of probes received from each upstream node during a
period T;useVito denote the number of probes received from
the upstream node associated with link i. The transmission
failure probability over link ican be estimated as pi=Vi
λT .
The pivalue will also be stored into the neighbor list. After a
downstream node finishes processing a received probe packet,
the probe will be discarded.
A new field, called link metric is established in the RREQ
message to store the metric value of each link. Once a node
receives a RREQ, it first checks its own neighbor list to get the
values of Miand pi, and then computes the EED according to
(5). The metric EEDiwill be inserted into the RREQ message
if the node needs to continuously forward it. In addition, every
node collects the EED metrics carried in the RREQ or RREP
messages and store them in a link cache. The cached link
metrics can be used to establish a network topology. If a node
has a packet for a destination with the path information not
established yet, the node can apply the Dijkstra’s algorithm to
compute an EED based shortest path based on the link cache
information.
2) WEED over a multi-radio path: To implement the
WEED based routing, each radio interface in the network is
uniquely identified with a separate IP address for each inter-
face. By considering each interface as the entity involved in the
routing, the operations of probing, maintaining neighbor list
and link cache, estimating the transmission failure probability
are similar to the node based case. There is an additional field
in all control messages to denote which interface is processing
510 15 20 25 30
0.682
1.024
1.364
1.705
2.046
2.387
2.728
3.069
Flow rate (pkts/second)
Total throughput (Mbps)
ETX
ETT
EED
(a) Total network throughput
510 15 20 25 30
0
0.5
1
1.5
2
2.5
Flow rate (pkts/second)
End−to−end delay (seconds)
ETX
ETT
EED
(b) Average end-to-end delay
Fig. 3. The routing performance in grid topology versus flow rate.
the message. To calculate the MRAB of each sub-path, two
extra fields need to be added to each record in the neighbor
list: the channel ID indicating the channel associated with
a link, and the IDR value computed according to (8) based
on power monitoring at the downstream node. When a node
forwards a RREQ message, the channel ID, the IDR values,
and EED all are attached as link metrics. Once the destination
receives the RREQ, it will return a RREP to the source node,
which duplicates all the routing information retrieved from
the RREQ. Based on the RREP messages, the source can
finally compute a shortest path regarding the WEED metric,
according to (15).
VI. PERFORMANCE EVA L U AT I O N
A. EED-Based Routing in a Single-Channel Network
The EED metric by itself can be used as an efficient routing
metric, since it effectively captures not only the queuing delay
at the network layer but also the retransmission delay at the
MAC layer. We use NS2 simulation results for a single-channel
wireless mesh network to demonstrate the performance, with
comparison with the two well-know metrics ETX and ETT.
The MAC protocol used for simulation is 802.11b.
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE INFOCOM 2009 proceedings.
510 15 20 25 30
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
Flow rate (pkts/second)
Total throughput (Mbps)
ETX
ETT
EED
(a) Total network throughput
510 15 20 25 30
0
0.5
1
1.5
2
2.5
3
3.5
4
Flow rate (pkts/second)
End−to−end delay (seconds)
ETX
ETT
EED
(b) Average end-to-end delay
Fig. 4. The routing performance in random topology versus flow rate.
We specifically consider two network topologies. The first
is a grid topology over a 1400m×1400marea. The area is di-
vided into 200m×200msquare cells, with each cell containing
one network node in its center. Four flows are deployed at the
1st, 3rd, 5th, and 7th rows of the grid, respectively, with the
source/destination nodes of each flow located at both ends of
the row correspondingly. The other topology randomly places
40 nodes in a 1000m×1000marea with necessary adjustment
to maintain the connectivity. We also run 4 flows over the
random topology. For both topologies, the transmission range
is 250m, and the interference range is 550m. All flows are
constant-bit-rate (CBR) flow, with the packet size of 512 bytes.
Fig. 3 and Fig. 4 present the total network throughput
and end-to-end delay, under ETX, ETT, and EED metrics,
respectively, versus the flow rate. The queue size at each
node is 20 packets, and the link metric update interval in our
EED implementation is 50 seconds. In the two figures, it is
explicitly demonstrated that EED metric can result in much
better end-to-end delay performance than ETX and ETT, under
both the grid and random topologies. Regarding the network
throughout, ETT and EED has the similar performance, while
outperforming the ETX. The reason that ETT and EED have
similar throughput performance is that both of them exploit the
20 40 60 80 100 120 140 160
2.2
2.25
2.3
2.35
2.4
2.45
2.5
2.55
2.6
2.65
2.7
Update interval (seconds)
Total throughput (Mbps)
(a) Total network throughput
20 40 60 80 100 120 140 160
1.25
1.3
1.35
1.4
1.45
1.5
1.55
1.6
1.65
Update interval (seconds)
End−to−end delay (seconds)
(b) Average end-to-end delay
Fig. 5. The impact of EED update interval on routing performance.
transmission failure probability for computing the link metric,
while the transmission failure probability is directly related to
the MAC throughput [7]. In most of the cases, ETT has slightly
higher throughput, which is due to the larger computation
overhead with EED and implementation overhead due to path
change incurred by the random queue length behavior. In
addition, we observe in the grid topology that the throughput
curves under all the three metrics become flat when the flow
rate exceeds a certain level, which indicates that the network
is approaching its maximum capacity.
We next examine the impact of EED update interval on
the routing performance. The grid topology is simulated, with
the queue size at each node being 20 packets and flow rate
25 packets/second. Fig. 5(a) and 5(b) present the network
throughput and the end-to-end delay versus the EED updating
interval, respectively. From the figure, it can be seen that
both inappropriate small and large intervals result in low
throughput and large delay. On one hand, an inappropriate
small update interval induces over-frequent link metric update
and results in a large bandwidth overhead. On the other hand,
an inappropriate large update interval will not timely respond
to a congested link and result in unnecessary packet loss due
to a full buffer. Fig. 5 demonstrates that an optimal update
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE INFOCOM 2009 proceedings.
510 15 20 25 30 35 40 45 50
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
Queue size (pkts)
Total throughput (Mbps)
ETX
ETT
EED
(a) Total network throughput
510 15 20 25 30 35 40 45 50
0
0.5
1
1.5
2
2.5
3
3.5
Queue size (pkts)
End−to−end delay (seconds)
ETX
ETT
EED
(b) Average end-to-end delay
Fig. 6. The impact of queue size on routing performance.
interval exists that can lead to largest throughput and smallest
end-to-end delay.
We further evaluate the effect of queue size on the rout-
ing performance, with results illustrated in Fig. 6. The grid
topology is considered with the EED update interval of 50
seconds and the flow rate of 20 packets/second. Regarding
the throughput as shown in Fig. 6(a), we can see that ETT
performance better for small buffer sizes, while EED better
for large ones. The reason is that when the buffer size is small,
in most of the cases all the buffers are full, where the EED
could not exploit more benefit compared to the ETT. The extra
computation overhead and route updating overhead, however,
will lead to a smaller network throughput. When the buffer is
large, EED can select a path with more buffer space, which
will lead to less tail-dropping of the packets and thus a higher
throughput. Regarding the delay as shown in Fig. 6(a), EED
achieves the smallest end-to-end delay in all the cases.
B. Metric Comparison in Multi-channel Environment
In this part, we use an example to illustrate the effectiveness
of WEED in quantifying the capacity of a multi-radio multi-
channel path. Since there is no much reference on NS2
simulation of DSR-based routing over multi-radio wireless
networks, we are still developing such a NS2 simulation
S 5 8 D
S 3 2 D
S 2 5 4 D
S 2 4 6 D
123
12
1
1231
12
13
0.2 0.2 0.25
0.1 0.3 0.35
0.2 0.15 0.25 0.4
0.1 0.2 0.3 0.2
I
II
III
IV
Fig. 7. WEED and WCETT Examples.
TAB L E I
PARAMETERS USED IN NUMERICAL ANALYSIS
Parameters Values
Packet length 600 bytes
Link bandwidth over channel 1 8 Mbps
Link bandwidth over channel 2 12 Mbps
Link bandwidth over channel 3 6 Mbps
Minimum contention window 0.02 ms
Maximum retransmission times 5
package. Thus, we resort to numerical analysis here to obtain
the results.
The path selection example is shown in Fig.7, where the
interference range r=1hop. There are four candidate
paths with three possible channels. The numbers within each
circle indicate the number of packets waiting in the buffer
of that node. The channel assignment is marked above the
link, and the transmission failure probability below. Other
necessary parameters for WEED computation are illustrated
in Table I. For the convenience of analysis, we ignore the
power monitoring part for inter-flow interference analysis,
while the intra-flow interference analysis here is sufficient
for demonstrating the capability of WEED in quantifying the
channel diversity.
We compare the path selection based on WEED and
WCETT, with results shown in Table II and III. WCETT
prefers path I over II, and IV over III, while WEED prefers
the other way. By investigating the configuration of Path I and
II, it is obvious that WEED makes a different decision from
WCETT’s because it takes the number of packets in the buffers
into consideration. The advantage of WEED over WCETT can
be demonstrated by comparing path III with path IV. Since the
two interfering links over channel 1 are located further away
in path III than in path IV, it is intuitively clear that path
III suffers a less amount of intra-flow interference. Moreover,
both path III and path IV consist of the same set of links (two
channel-1 links, one channel-2, one channel-3), so path III will
achieve a higher path capacity than path IV due to the less
interference over it. Thus, WCETT makes a wrong decision,
while WEED makes a correct one. The reason for WCETT’s
wrong decision is that it deems that co-channel links along a
path always interfere with each other, disregard the distance
among them. With r=1, it is obvious that the two channel-1
links in path III will not interfere with each other, in fact.
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE INFOCOM 2009 proceedings.
TAB L E II
NUMERICAL RESULTS FOR WCETT [13]
Path H
i=1 ETTimax(Xj)WCETT (β=0.5)
I2.3881 1.0667 1.7274
II 2.1612 1.5898 1.8755
III 3.2873 1.75 2.5187
IV 3.0238 1.5238 2.2738
TABLE III
NUMERICAL RESULTS FOR WEED
Path H
i=1 EEDiMRAB WEED (α=0.5) CDC
I16.6991 6Mbit/s 13.5496 3.0
II 5.8638 4Mbit/s 5.9319 2.0
III 13.8753 6Mbit/s 11.33765 4.0
IV 13.6927 4Mbit/s 14.04635 2.6667
Table III illustrates that the MRAB and CDC value can
quantitatively demonstrate the capacity difference between
two multi-radio multi-channel paths. Moreover, the MRAB
and CDC values also reveals the interesting insight that the
relationship between the channel diversity and the length of
the path in terms of hop count is not monotonic. For example,
path III achieves better channel diversity than path I, being
one-hop longer; but path I has better channel diversity than
path IV, being one-hop shorter.
VII. CONCLUSION
In this paper, we aims at designing link/path metrics that
can lead to path selection with the minimum end-to-end delay,
while a high network throughput can also be achieved. The
paper has key contributions in two aspects: 1) Based on the
concept of network/MAC cross-layer design, both the queuing
delay in network layer and transmission delay in the network
layer are included in the EED link metric computation; 2)
A generic iterative approach is developed to compute the
achievable bandwidth over a multi-radio multi-channel path,
which captures the complex interaction among hop count,
channel assignment, and inter/intra flow interferences to form
the WEED path metric. A side benefit of our EED/WEED link
metric computation is a quantitative channel diversity coeffi-
cient. We demonstrate the performance of EED/WEED based
routing via extensive numerical analysis and NS2 simulation
results.
REFERENCES
[1] P. Gupta and P. R. Kumar, “The cpacity of wireless networks,IEEE
Trans. Inform. Theory, vol. 46, no. 2, pp. 388–404, Mar. 2000.
[2] M. Gastpar and M. Vetterli, “On the capacity of wireless networks: the
relaycase,” in Proc. IEEE INFOCOM, 2002, pp. 1577–1586.
[3] A. E. Gamal, J. Mammen, B. Prabhakar, and D. Shah, “Throughput-
delay trade-off in wireless networks,” in Proc. IEEE INFOCOM, 2004,
pp. 464–475.
[4] I. F. Akyildiz, X. Wang, and W. Wang, “Wireless mesh networks: a
survey,Computer Networks., 2005, pp. 523–530.
[5] G. Jakllari, S. Eidenbenz, N. Hengartner, S. V. Krishnamurthy, and
M.Faloutsos, ”Link Positions Matter: A Noncommutative Routing
Metric for Wireless Mesh Network, in Proc. IEEE INFOCOM, 2008,
pp.744–752.
[6] C. Koksal and H. Balakrishnan, “Quality-Aware Routing Metrics
for Time-Varying Wireless Mesh Networks,” IEEE J. Select. Areas
Commun., vol. 24, no.11, pp. 1984–1994, November.2006.
[7] Y. Cheng, X. Ling, W. Song, L.X. Cai, W. Zhuang, and X. Shen A
Cross-Layer Approach for WLAN Voice Capacity Planning,” IEEE J.
Select. Areas Commun., vol. 25, no. 4, pp. 678–688, May 2007.
[8] A. Raniwala, T.-c. Chiueh, “Architecture and Algorithms for an IEEE
802.11-Based Multi-Channel Wireless Mesh Network, in Proc. IEEE
INFOCOM, 2005, pp. 2223–2234.
[9] A.Cerpa, J. L. Wong, M. Potkonjak and D. Estrin, “Temporal Properties
of Low Power Wireless Links: Modeling and Implications on Multi-Hop
Routing,” in Proc. ACM MobiHoc, 2005, pp. 414–425.
[10] K. N. Ramachandran, E. M. Belding, K. C. Almeroth and M. M.
Buddhikot, “Interference-Aware Channel Assignment in Multi-Radio
Wireless Mesh Networks, in Proc. IEEE INFOCOM, 2006, pp.1-12.
[11] Y. Yang, J. Wang, “Design Guidelines for Routing Metrics in Multihop
Wireless Networks, in Proc. IEEE INFOCOM, 2008, pp.1615 - 1623.
[12] D. B. Johnson, D. A. Maltz and Y. Hu, “The Dynamic Source Routing
Protocol for Mobile Ad Hoc Networks (DSR),” in IETF, INTERNET-
DRAFT, 2003, April.
[13] R. Draves, J. Padhye, and B. Zill, “Routing in Multi-Radio, Multi-Hop
Wireless Mesh Networks, in ACM MOBICOM, 2004, pp. 114–128.
[14] R. Draves, J. Padhye, and B. Zill, ”Comparison of Routing Metrics for
Static Multi-Hop Wireless Networks, in ACM SIGCOMM, 2004, pp.
133–144.
[15] Y. Yang, J. Wang, and R. Kravets, “Designing Routing Metrics for
Mesh Networks,” in Proc. IEEE Workshop on Wireless Mesh Networks
(WiMesh), 2005.
[16] J. So, N. H. Vaidya, “Load-Balancing Routing in Multichannel Hybrid
Wireless Networks With Single Network Interface,” in IEEE Trans. Veh.
Technol., vol. 56, no. 1, pp. 342–348, Jan, 2007.
[17] T. Liu, W. Liao, “Capacity-Aware Routing in Multi-Channel Multi-Rate
Wireless Mesh Networks, in Proc. IEEE ICC, 2006, pp. 1971–1976.
[18] K. Jain, J. Padhye, V. N. Padmanabhan, and L. Qiu, “Impact of
Interference on Multi-hop Wireless Network Performance, in AC M
MOBICOM,, 2003, pp. 66–80.
[19] X. Li, H. Chen, Y. Shu and X.Chu, “Energy Efficient Routing With
Unreliable Links in Wireless Networks, in Proc. IEEE International
Conference on Mobile Adhoc and Sensor Systems (MASS), 2006, pp.
160–169.
[20] H. Zhai, Y. Fang ”Impact of Routing Metrics on Path Capacity in Multi-
rate and Multi-hop Wireless Ad Hoc Networks, in Proc. IEEE ICNP,
2007, pp. 86–95.
[21] D. S. J. De Couto, D. Aguayo, J. Bicket, and R. Morris, A High-
Throughput Path Metric for Multi-Hop Wireless Routing, in ACM
MOBICOM, 2003, pp. 134–142.
[22] K. Kim, K. G. Shin, “On accurate measurement of link quality in multi-
hop wireless mesh networks,” in ACM MOBICOM, 2006, pp. 38–49.
[23] H. Zhai, J. Wang and Y. Fang, “Distributed packet scheduling for
multihop flows in ad hoc networks, in Proc. IEEE WCNC, 2004, pp.
1081–1086.
[24] Y. Xiao, K. Thulasiraman and G. Xue, “QoS routing in communication
networks: approximation algorithms based on the primal simplex method
of linear programming,” in IEEE Trans. Comput., vol. 55, no. 7, pp.
815–829, July. 2006.
[25] H. Li, Y. Cheng, C. Zhou, “Multi-hop effective bandwidth based routing
in multi-radio wireless mesh networks,” in Proc. IEEE Globecom, 2008.
[26] J. Tang, G. Xue, and W. Zhang, “Interference-Aware Topology Control
and QoS Routing in Multi-Channel Wireless Mesh Networks, in ACM
MobiHoc, 2005, pp. 68–77.
[27] The Network Simulator - NS2, http://www.isi.edu/nsnam/ns/.
[28] “Wireless Lan Medium Access Control (MAC) and Physical Layer
(PHY) specifications,” ANSI/IEEE Std 802.11: 1999 (E) Part 11,
ISO/IEC 8802-11, 1999.
[29] A. Raniwala, T.-c. Chiueh, ”Centralized Channel Assignment and
Routing Algorithms for Multi-Channel Wireless Mesh Networks, in
ACM SIGMOBILE Mobile Computing and Communications Review,
2004, pp. 50–65.
[30] A. Abdrabou and W. Zhuang, “Service time approximation in IEEE
802.11 single-hop ad hoc networks,” IEEE Trans. Wireless Commun.,
vol. 7, no. 1, pp. 305-313, Jan. 2008.
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE INFOCOM 2009 proceedings.
    • "Therefore, achieving fast data delivery and response for the massive data generated in WSNs poses new research challenges. Data delivery delay has been extensively studied in forwarding quality measurement789, sensor network routing, and scheduling10111213 . Most of the works focus on investigating the metrics to characterize the forwarding quality or minimizing average path delay. "
    [Show abstract] [Hide abstract] ABSTRACT: Many time-sensitive applications impose high requirement on real-time response. There exist many algorithms and routing protocols for efficient data packet delivery. However, previous works set the same retransmission threshold for all the relay nodes along a delivery path. The method decreases the probability of a packet being transmitted through the delivery path within given deadline. In this paper, we focus on finding the optimal retransmission thresholds for the relay nodes, such that the summation of the probability of a packet being transmitted to the next relay node or destination node within the specified deadline is maximized. A distributed greedy algorithm that can be run on sensor node is proposed, which enables the sensor node to adaptively set the optimal retransmission threshold. To avoid dropping the packet forwarded to the destination within given deadline with high probability, we develop a packet dropped protocol based on probabilistic delay bound. Experimental results show that the proposed protocols have better performance.
    Full-text · Article · Sep 2015
    • "However, the authors did not consider the channel allocation problem, which might degrade the performance in cognitive radio networks. In [6], the authors studied how to select a path with the minimum cost in terms of expected end-to-end delay (EED) in a multiradio wireless mesh network. However, the path selection relies on the information of the queueing length at the nodes, which adds much complexity. "
    [Show abstract] [Hide abstract] ABSTRACT: Delay issue is a challenge in multi-hop cognitive radio networking because of the dynamic change of channels available to the cognitive radio nodes. To address this challenge, we propose a channel allocation and rate allocation scheme with low complexity in this paper. Our scheme is flexible so that it can react to the dynamic change of channel availability, and it can minimize the delay by considering the primary user activities. First, we model the primary users' activity, channel availability and the interference among the cognitive radio nodes in a cognitive radio network environment, and show how to implement a channel allocation algorithm and a rate allocation scheme in multi-hop cognitive radio networks. Second, we formulate an optimization problem to minimize the end-to-end delay of the network. We consider the channel availability constraint and the primary user activities jointly, and analyze the end-to-end delay in multi-hop cognitive radio networks. In order to reduce the computing complexity, we propose a graph coloring based channel allocation and rate selection algorithm using gradient descent method. Furthermore, we show the performance of our schemes and compare them with existing schemes for different scenarios of channel availability and number of channels through simulations. Our work brings insights on how to make rate selection and channel allocation in multi-hop cognitive radio networks.
    Full-text · Conference Paper · Aug 2015 · International Journal of Distributed Sensor Networks
    • "IRU metric is the ETT weighted with the number of the interference links, while CATT metric extends IRU by considering the effect of packet size and raw data rate on the links because of the use of multiple channels. WEED (Weighted end-to-end Delay) [7] involved with the delay raised in source to destination packet delivery. Some existing QoS routing protocols operate with the knowledge of the available bandwidth of each link [2], [4],[6], [16], [17], [18], [19]. "
    [Show abstract] [Hide abstract] ABSTRACT: Wireless mesh networks are capable of facilitating wireless local area network coverage, Internet broadband access and network connectivity for stationary or mobile hosts for both network operators and customers at low costs. The strengths and weakness of metrics used in routing protocols are reflected directly in WMN's characteristics. Metrics associated with link bandwidth have received more attention in recent years since they help to provide efficient Quality of Service (QoS) to users. However, there are horrible problems in achieving high network throughput. In this paper, complete solutions have been proposed for enhancing the routing process. We put forward a new routing metric called Composite Available Bandwidth (CAB) that captures non-dominated maximum available bandwidth path for routing. An enhanced Hop-by-Hop Routing protocol has been suggested which satisfies the consistency property. Consistency Property guarantees that each node makes a proper packet forwarding decision, which helps the data packets to traverse over the intended path. Our comprehensive simulation experiments also show that our proposed metric outperforms existing path metrics in identifying high-throughput paths. 1. Introduction A wireless mesh network (WMN) consists of a largenumber of wireless nodes. The nodes form a wireless overlay to cover the service area while a few nodes are wired to the Internet. WMN has to provide great support to diversified multimedia applications for its users since it is part of Internet. It is absolutely necessary to provide efficient Quality-of-Service (QoS) support in this kind of networks [1]. Seeking the path with the maximum available bandwidth is one of the fundamental issues for supporting QoS in the wireless mesh networks. The available path bandwidth is stated as the maximum additional rate a flow can push before saturating its path [2]. Therefore, if the traffic rate of a new flow on a path is no greater than the available bandwidth of this path, accepting the new traffic will not disturb the bandwidth guaranteed of the existing flows. This paper has attention on the problem of identifying the maximum available bandwidth path from a source to destination, called the Maximum Bandwidth Problem (MBP). MBP is a subproblem of the Bandwidth-Constrained Routing Problem (BCRP), the problem of identifying a path with at least a given amount of available bandwidth [3]. In the
    Full-text · Conference Paper · May 2015 · International Journal of Distributed Sensor Networks
Show more

Recommended publications

Discover more