Content uploaded by Emmanuel Baccelli
Author content
All content in this area was uploaded by Emmanuel Baccelli
Content may be subject to copyright.
Routing Across Wired and Wireless Mesh Networks:
Experimental Compound Internetworking with OSPF
Juan Antonio Cordero Fuertes, Matthias Philipp, Emmanuel Baccelli
Abstract—As wireless mesh networks aredeployed, anew
concept emerges: compound internetworks, i.e.,internetworks
that contain both wired networks and wireless mesh networks.
Routing is one of the key challenges that arises in compound
internetworks: indeed, while specific routing protocols are
typically used for wired networks on one hand, and for wireless
mesh networks on the other hand,it has been observed that
operating asingle routing protocol to manage acompound
internetwork as awhole brings several advantages. In this
realm, the IETF has thus standardized protocol extensions
to Open Shortest Path First(OSPF,the routing protocol
used by morethan 50 %of the wired routers in today’s
Internet), enabling OSPF to operate simultaneously on wired
networks, and on wireless mesh or moderately mobile ad hoc
networks (MANETs). This paper evaluates the performance of
OSPF coupled with such aprotocol extension for MANETs on
an experimental compound internetwork testbed. This paper
reports on experiments carried out with OSPF operating simul-
taneously over Ethernet and 802.11b.Despite the limitations of
the testbed, these experiments provide both aproof-of-concept
and complementary results compared to prior work in the
domain, which was mostly based on simulations, and focused
on wireless ad hoc network scenarios only.
I.INTROD UCTIO N
The advent of packet radio networks and lead to new
concepts such as Mobile Ad hoc Networks (MANETs)
or wireless mesh networks. Since the late 1990’s, several
research efforts havefocused on enabling communication
in such wireless multi-hop networks in which topology is
spontaneous and dynamic. This resulted in different routing
protocols being designed for wireless mesh networks and
MANET operation, the most prominent to date being OLSR
[13] and AODV [15]. OLSR is for instance deployed in many
urban wireless community mesh networks around the world
[21].
In practice, the interconnection of wired networks with
wireless mesh networks in the Internet gives birth to Au-
tonomous Systems (AS) containing both mesh and fixed
networks –such an AS is hereafter denominated compound
AS.An approach for routing in acompound AS consists
in splitting the AS into several routing domains, and then
to use aseparate routing protocol for each domain: on one
hand MANET-specific protocols in wireless mesh networks,
and on the other hand, routing protocols designed for wired
networks. An alternativeapproach consists in using asingle
routing protocol for the whole compound AS. The former
approach requires the presence in the AS of routers with
specific hardware and software capabilities, denominated
gateways,in order to provide connectivity between different
routing domains. The latter approach, explored in this paper,
reduces the cost of network maintenance and operation
as no such gateways are needed [2], and naturally avoids
bottlenecks created by such gateways. This is, however,at
the expense of increasing the complexity of the employed
routing protocol, which thus needs to handle the diverse
characteristics of wireless mesh and wired networks with
the same core mechanisms.
In today’sInternet, IGP routing [16] is essentially provided
by two protocols based on the link state algorithm :OSPF
[17] [10], and ISIS [18]. This paper explores the use of
OSPF to provide routing in compound ASes, and evaluates
its performance on acompound internetwork testbed.
A. Related Work
Leveraging the commonality between OLSR and OSPF
(both based on the link state algorithm), operation of OSPF
on MANETs has recently been considered. This approach
may indeed be cost-effectiveconcerning the integrating of
wireless mesh networks in existing infrastructure, as more
than 50% of routers in the Internet already run OSPF.
Early studies and proposals such as [14] [19] paved the
way to IETF standardizing several OSPF protocol extensions
for operation over ad hoc networks [9] [6] [8]. Various
studies have evaluated and compared the performance of
these extensions in mobile ad hoc scenarios [7], [12], [20],
essentially via simulations. Further improvements of these
extensions havebeen proposed in [3], [4], still comparatively
evaluated via simulations in MANET only scenarios. In
contrast, this paper evaluates OSPF on areal testbed that
consists of awireless mesh network interconnected with a
fixed wired network.
II.THEOSPF PROTOCOLA ND ITSMANET EXT ENSIO N
OSPF [10], [17] is alink-state routing protocol for IP
networks. This implies that each router maintains alocal
instance of the Link-State Database (LSDB), representing
the full network topology –with the objectiveof the protocol
being that each router should havethe same information in
its local instance of LSDB and, thus, the exact same viewof
the network topology.Paths to every possible destination are
derived from the Shortest Path Tree (SPT) that every router
computes, by way of Dijkstra’salgorithm. Although OSPF
supports network partitioning in several routing areas,the
work described in this paper explores the use of asingle
area for routing in the whole internetwork.
Routers acquire local topology information and advertise
their own presence by periodically exchanging Hello mes-
sages with all their 1-hop neighbors (i.e.,neighbor sensing).
With such signaling, each router becomes aware of its
immediate network topology,i.e.its 2-hop neighborhood.
This also allows verification of bidirectional connectivity
with 1-hop neighbors (then called bidirectional neighbors).
Each router also explicitly synchronizes its local instance
of LSDB with asubset of its bidirectional neighbors. Links
between arouter and its synchronized neighbors are called
adjacencies,and are required to form anetwork-wide con-
nected backbone, connecting all routers in the network, in
order to ensure paths can be computed correctly.
Finally,routers also acquire remote topology information
by way of receiving Link State Advertisements (LSA). Each
such LSA lists mainly the current adjacencies of the router
which generated the LSA. LSAs are disseminated through
the entire network in reliable fashion (explicit acknowledge-
ments and retransmissions) via the backbone formed by
adjacencies; this operation is called LSA Flooding.Thus,
anyrouter which has formed adjacencies must advertise this
periodically by way of originating an LSA and performing
LSA flooding.
Remote topology information is then used for the con-
struction of the Shortest Path Tree: each router computes the
shortest paths over the network graph described in the set of
received LSAs it, by way of Dijkstra’salgorithm.
A. Packet Types
Routers in OSPF use fivetypes of messages and packets
to exchange topology information over the networks, some
of them havebeen already mentioned in this section: Hello
packets are used for neighbor sensing, Database Description
(DBDesc) packets are exchanged for LSDB synchronization
and Link State Advertisements (LSAs) are used for topology
reliable flooding and update. After the exchange of DBDesc
packets, arouter in process of LSDB synchronization may
request to its synchronizing neighbor the retransmission of
particular LSAs –these requests are sent by way of Link
State Request (LSReq) packets. Several LSAs may be sent
in asingle Link State Update packet (LSU). Several LSA
acknowledgements may also be grouped in asingle Link
State Acknowledgment (LSAck) packet.
B. Interface Types for Wired Links
Rules for flooding and adjacencyhandling vary for the dif-
ferent interface types supported by OSPF.In broadcast and
non-broadcast multiple access (NBMA) interfaces, the flood-
ing procedure is mainly managed by Designated Routers
(DRs). ADesignated Router is elected among routers whose
interfaces are connected to the same link.Such aDR forms
adjacencies with all the routers connected to the same link,
and it becomes responsible for flooding of LSAs, origi-
nated by routers on that link. In point-to-point and point-
to-multipoint interfaces, all links are synchronized and all
interfaces participate in LSA flooding.
C. MANET Interface Type
The MANET interface type is defined in the extension
of OSPF for operation over MANETs. Three different ex-
tensions havebeen standardized by the IETF [6], [8], [9],
each of which specifies mechanisms to optimize topology
description, flooding and LSDB synchronization in wireless
ad hoc environments.
The experiments carried out used RFC 5449 [9]. Wireless
interfaces following this specification select aset of Multi-
Point Relays (MPRs) among its bidirectional neighbors [13].
The set of MPRs selected by the wireless interface of arouter
must ensure that every packet received from the router reach
all 2-hop neighbors of the selecting interface in 2hops (MPR
coveragecriterion). Alink between arouter and one of its
MPRs is denominated MPR link.
LSA flooding is then performed through MPRs, meaning
that an LSA transmitted (originated or forwarded) by an in-
terface is retransmitted by the MPRs of such interface. Links
between interfaces and their MPRs are synchronized and
thus become adjacent.Moreover,each interface describes
in its LSAs the set of MPRs and MPR selectors. As the
set of adjacencies based on MPR selection may not provide
aconnected subgraph, links from one additional router in
the network (denominated synchrouter) are also declared
adjacent to ensure adjacencyset connection [5]. The Shortest
Path Tree is then constructed over the set of adjacencies.
III.TESTBEDDESCR IPTION
This section describes the characteristics of the employed
networking testbed. Section III-A presents the distribution of
computers in the testbed and the network topology that they
form. Section III-B details the implications of such topology
in OSPF routing.
A. Interfaces Configuration and Network Topology
The testbed is composed of 6fixed computers
(routers/hosts) attached to two interconnected networks: a
wired network and awireless network. Table Iindicates the
network interfaces of each computer.For more details about
computers’ hardware, see the Appendix.
Computer Abbr.Wired ifs. Wireless ifs.
server Seth1, eth2 –
hybrid1 h1eth1 wlan0
hybrid2 h2eth1 wlan0
wless1 w1–wlan0
wless2 w2–wlan0
wless3 w3–wlan0
TABLE I
NET WORKINTERFACESOFT ESTBEDCOMPUTE RS.
1) Physical Topology: The internetwork connecting these
computers was deployed in the Computer Science Lab (Lab-
oratoired’Informatique,LIX) of ´
Ecole Polytechnique, in
Paris (France). Three scenarios –I, II and III– were config-
ured over the resulting internetwork. These scenarios permit
to test the communication between computers wless3 and
server,for different situations. The physical distribution
of computers at LIX is displayed in Figure 1.
Positions of computers do not change, except for the case
of wless3,which has adifferent position for scenario Iand
for scenarios II and III, as shown in Figure 1.
Sh
1
h
2
w
2
w
1
w
3
w
3
(II, III)
(I)
10 m
Fig. 1. Computers position over the plan of LIX.
2) Logical Internetwork Topology: Each scenario corre-
sponds to aspecific internetwork topology.Figure 2indicates
the internetwork topology graphs for scenarios I, II and
III. In the wired network, computers communicate through
the IEEE 802.3 (Ethernet) standard protocol, server is
connected with hybrid1 by way of interface eth1 and
with hybrid2 by way of interface eth2, as shown in
Figure 2. In the wireless network, interfaces communicate
through the IEEE 802.11b WLAN standard protocol, and all
wireless routers (hybrid1,hybrid2,wless1,wless2
and wless3)use their wireless interface wlan0. The topol-
ogy that results from wireless reachability among computers
hybrid1,hybrid2,wless1,wless2 and wless3 is
modified by means of MACfiltering in order to disable links
h1←→ h2,w1,3←→ w2,w1,3←→ h2and w2←→ h1.
In scenario III, link h1←→ w1is suppressed by disabling
interface wlan0 at computer hybrid1.
S
h
1
h
2
w
2
w
3
eth1
eth1 eth2
eth1
wlan0
wlan0
wlan0
wlan0
S
h
1
h
2
w
2
w
3
w
1
eth1
eth1 eth2
eth1
wlan0 wlan0
wlan0 wlan0
wlan0
S
h
1
h
2
w
2
w
3
w
1
eth1
eth1 eth2
eth1
wlan0
wlan0 wlan0
wlan0
I II III
Fig. 2. Considered topologies for scenarios I, II and III.
B. OSPF Routing Configuration
All interfaces use the extended OSPFv3 routing protocol,
wired and wireless interfaces using different interface types.
Wired interfaces are configured as point-to-point interfaces,
as specified in RFCs 2328 [17] and 5340 [10]. Wireless
interfaces are configured as MANET interfaces,as specified
in the MPR-OSPF MANET extension (RFC 5449 [9]).
1) OSPF Adjacencies and MPRs: According to the spec-
ification of OSPF and MPR-OSPF extension, all links in any
of the considered topologies for scenarios I, II and III are
adjacent. Within the wired network, every point-to-point link
is an adjacency.In the wireless network, wireless links are
adjacent if theyare MPR links. The list of MPRs of every
wireless interface, for each scenario, is displayed in Table II.
It can be observed that all links are MPR links, and
therefore all are adjacent. In this topology,the presence of a
Interface I II III
hybrid1:wlan0 w1w1–
hybrid2:wlan0 w2w2w2
wless1:wlan0 –w2w2
wless2:wlan0 w3w1w1
wless3:wlan0 w2w1w1
TABLE II
MPRS SE LECT EDBYEACHWIRE LES S INTE RFACE,FOREACHSCENARIO.
synchrouter (see section II-C) is thus redundant.
2) OSPF Flooding: Flooding in the wired network is
performed through adjacent links, i.e.S←→ h1and S←→
h2.In the wireless network, flooding is performed:
•through the MPR links (from awireless router towards
its MPR), and
•through all links connecting an interface to ahybrid
router (hybrid1 and hybrid2).
IV.EXPERIMENTSAND RESULTS
For each scenario (I, II and III), communication between
wless3 and server is tested by way of two experiments:
•Transmission of ICMPv6 [11] requests (pings)from
wless3 to server.The measure of time between
the transmission of an ICMP request and its reply
corresponds to the Round Trip Time (RTT) of the ping
through the evaluated path.
•Transmission of aconstant bit rate data UDP flowfrom
wless3 to server.Comparison between packets sent
and packets received permits to test the quality of the
traversed paths and the wireless links that compose them
in each scenario. For adetailed description of these
UDP flows, see Table III in the Appendix.
Displayed results in this section showthe most prominent
averaged measurement values over tens of samples (see the
Appendix for details); acomplete description of the obtained
results can be found in [1]. The three considered scenarios
(I, II and III) are complemented by another scenario in
which information is transmitted and measured through the
wired link h1←→ S.Results on this scenario are added
for completeness. Section IV-A presents the results obtained
in both experiments, for each scenario, in terms of quality
of wireless links. Section IV-B examines the amount and
structure of OSPF control traffic.
A. Wireless Mesh Communication
Figures 3.a and 3.b display the results of the performed ex-
periments, in particular the delay for ICMP requests (pings)
and the packet delivery ratio of CBR UDP data flows.
Figures 3.a and 3.b indicate the degradation of the quality
of communication between routers wless3 and server
as the number of wireless links between them increases.
As expected, the wired link h1←→ Shas an almost-
ideal behavior: 100% PDR and no significant delay.The
negativeimpact of wireless links in the path from source
to destination is close-to-linear with the number of traversed
(a) Packet Delivery Ratio (PDR)
(b) Round TripTime (RTT)
Fig. 3. PDR of UDP flows and RTT of ICMP requests, depending on the
number of wless hops.
wireless links, as shown in Figure 3.a: more than 30% of
transmitted packets are lost in the first wireless link, and such
percentage increases about a15% per additional wireless link
included in the path. Figure 3.b shows that such degradation
is also evident in terms of round trip time (RTT). Replies
to ICMP requests are immediately delivered through awired
link, but the average and the variation of delays growwith
the number of wireless links involved –is in the order of
tens of miliseconds for 2and 3wireless links.
While the impact on communication due to the use of
wireless links depends on the specific topology and the
network technology that is used, two conclusions can be
drawn from these experimental results. As each additional
wireless hop in the route of data packets in the network
implies asignificant degradation of the quality of com-
munication, routing in wireless networks should preserve
the principle of shortest (wireless) paths, meaning that the
number of wireless links traversed by data packets should
be minimized. This confirms the conclusions of simulation-
based studies such as [20] which highlights the importance
of not sacrificing path optimality for less control traffic.
Moreover,in the context of compound internetworks with
both wired and wireless links, it is obvious that ’optimal’
does not necessarily mean ’the least number of hops’, as
implicitly used in previous work such as [7], [12], [3], [4].
Indeed, in compound internetworks, it is better for apath to
use wired links than wireless links, whenever possible, even
if it means more hops in the end. This observation confirms
that metrics used should indeed be able to track link quality,
and that OSPF sepcifications such as [9] [6] [8] should be
completed with astandard way to do that in MANETs.
B. OSPF Control Traffic Pattern
Fig. 4. Control traffic overhead at server:eth1.
Figures 4, 5and 6display the evolution of OSPF control
traffic transmitted by wireless interfaces wless3:wlan0
and hybrid1:wlan0,on one side, and the wired interface
server:eth1,on the other.The fivepacket formats used
in OSPF (Hello, LSUpdate, LSRequest, LSAck and DBDesc,
see section II-A) can be distinguished in these figures.
Measurements were taken with the topology of scenario I,
each point indicating the number of bytes sent within an
interval of 5seconds. The traffic load of the internetwork
was composed of aCBR UDP data traffic flowfrom wless3
towards server (see Table III), and OSPF control traffic.
The figures showthe structure of such control traffic, in
terms of number of bytes, during the first 335 seconds of
network operation, i.e.,after routers’ startup. All interfaces
are configured with the same OSPF parameters (see Table
IV), in order to facilitate the comparison between control
traffic patterns of each of them.
1) Hello Packets: The amount of Hello packets sent by
each interface is kept constant along the monitored time. As
HelloInterval=2sec,interfaces transmit 2.5Hello packets
per interval of 5seconds. The length of an average Hello
packet is significantly longer in wireless interfaces (Figures
5and 6) than in wired interfaces (Figure 4). For the same
number of neighbors, Hellos from server:eth1 have40
bytes while those from hybrid1:wlan0 have75.34 bytes
(average values). This is due to the fact that Hello packet
format in RFC 5449 [9] includes additional information
about link costs, adjacencies and MPR selection, which is
added to the format specified in OSPF [17] and OSPFv3
[10].
2) LSDB Synchronization: The existence of ongoing
LSDB synchronization processes during the monitored time
interval can be noticed in the OSPF control traffic structure
Fig. 5. Control traffic overhead at wless3:wlan0.
by way of the presence of Database Description (DBDesc)
packets. The fact that these packets are only present, for
wired interfaces, in the first part of the monitored interval
(from t=0secto t=10sec,as shown in Figure 4indicates
that links become synchronized only when the routers are
switched on. In contrast, DBDesc are transmitted in the
whole monitored interval for wireless interfaces. Thisis
consistent with the fact that wired links are mostly stable
and therefore there is no need to repeat synchronization
process after the first LSDB exchange. Wireless links, in
contrast, are more prone to packet losses and link failures,
and need thus to be synchronized several times during the
network lifetime, even in the absence of router mobility.
The same phenomenon can be observed with LSRequest
packets, which can only be sent during the last phase of
the LSDB synchronization process, when the synchronizing
neighbors havecompleted the exchange of DBDesc packets.
These observations are also consistent with simulation-based
studies such as [20] which havestudied the impact of link
quality on OSPF control traffic.
Fig. 6. Control traffic overhead at hybrid1:wlan0.
3) Link State Updates, Requests and Acknowledgements:
LSUpdate packets contain one or more Link State Advertise-
ments (LSAs). Such LSAs can be either originated by the
sending interface, either originated by another interface and
flooded (forwarded) by the sending interface. Transmission
of LSUpdate packets follows acommon pattern in all the
interfaces in the internetwork. Thus, pattern consists of
periodic peaks followed by time intervals (valleys)in which
the number and size of LSUpdate transmissions is lower and
roughly constant.
Despite the common pattern in the LSUpdate traffic,
several differences can be observed between wired and wire-
less interfaces. This section concentrates on three particular
aspects: peak width, height of valleys between consecutive
peaks and transient state (after routers are switched on).
a) Transient state: Immediately after switching on,
wireless interfaces transmit ahigh number of packets –
mostly,LSUpdate packets sent in response to LSRequest
packets received during the first LSDB synchronization
processes in all wireless links (Figures 6and 5, between
t=0secand t=50sec). This amount of transmissions
involves traffic rates above220Bps (1.1kBper interval of
5sec), then decreases and stabilizes in aslightly lower level
(maximum peak of 130Bps). The opposite behavior is found
in wired interfaces (Figure 4), in which the initial transient
period of lowLSUpdate traffic rate (about 26Bps =130B
5sec for
server:eth1)is followed by asteady period in which the
minimum LSUpdate rate is slightly higher (about 29Bps =
145B
5sec for server:eth1). These different behaviors can be
explained by the different roles that flooding has over wired
and wireless links. Due to their stability,packets sent over
wired links are mostly forwarded packets –that is, theycome
from other interfaces than those involved in the links. In the
first instants in which there is no flooding over the network
because adjacencies havenot been formed in the network
and flooding links havenot yet been identified, the overall
traffic traversing such wired links is temporarily low.The
opposite is observed in wireless links.
b) Peak width: Peaks are narrower in wired inter-
faces (∼10secfor server:eth1)than in wirelessin-
terfaces (∼25secfor hybrid1:wlan0,∼30secfor
wless3:wlan0). For interfaces attached to wireless links,
there is ahigh probability that atopology change causes
anewtopology update before the LSRef resh interval –
therefore, intervals between consecutivetransmission of in-
terfaces’ topology descriptions are shorter than LSRef resh
and the width of the peak increases. In stable wired links,
in contrast, intervals between consecutivetransmissions are
closer to the LSRef resh parameter and, therefore, LSUp-
date transmission events are less spread in time.
c) Depth of valleys: Besides the peaks caused by trans-
mission of its own topology description, either periodic or
as areaction to atopology change, two other events may
lead an interface to transmitLSAs: (i) forwarding of LSAs
originated by other interfaces in the internetwork, and (ii)
retransmission of LSAs not acknowledged by their intended
destinations. Both additional events explain the presence of
valleys with significant traffic rate, i.e.,anon-zero minimum
level of LSUpdate transmissions in the monitored interfaces.
In wired (reliable) links such as server:eth1,such trans-
missions are caused by flooding, and involveabout 25Bps
(127Bper interval of 5sec). Wireless interfaces such as
wless3:wlan0 have a minimum LSUpdate transmission
rate of about 16Bps (80Bper interval of 5sec)caused by
LSA retransmissions and flooding.
V.CO NCLUSIO N
This paper addresses routing with OSPF across compound
Autonomous Systems, i.e.,internetworks combining wired
networks and wireless mesh networks. Compound internet-
works are bound to become acommon phenomenon as
wireless mesh networks are being deployed and coexist
with wired networks. This paper reports on experiments
carried out on acompound internetwork testbed confirm
prior simulation-based studies analyzing OSPF control traffic
over wireless links. Weobserved that even without mobility
and asmall number of wireless neighbors, link synchro-
nization involves acontinuous and substantial amount of
traffic. This observation confirms that reducing the number
of synchronized links is apriority when using OSPF over ad
hoc networks. Furthermore, data path quality was observed
to significantly decrease with each additional wireless hop
involved in the path. This observation suggests that wired
links should be preferred to wireless links whenever possible,
and highlights the need to complement current OSPF specifi-
cations with astandard way to track link quality in compound
internetworks. The importance of leveraging every possible
wired connection indicates that all links in the compound AS
should be taking into consideration for route computation.
As this implies alarge number of gateways between routing
domains when several routing protocolsare used in the AS,
and this is costly,asingle routing protocol is prefered in
order to avoid gateways.
APP END IX
•Testbed Hardware
–Wired interfaces: DECchip 21140.
–Wireless interfaces: Broadcom BCM4306 WLAN.
•Testbed Software
–Operating System: Ubuntu v.10.04 with kernel 2.6.32.
–Routing Protocol Implementation: ospf6d daemon of
Quagga/Zebra routing suite v.0.99.15.
∗Wired interfaces: Point-to-point.
∗Wireless interfaces: MANET (RFC 5449 [9]).
•Setup for PDR and RTT Measurements
–Routers were switched on within [0,2]sec.
–PDR results averaged over 84 iterations.
–ospf6d daemon NOTrestarted in each iteration.
–UDP flows, started 60sec after ospf6d daemon switch-
on:
Nominal sender bit rate 100 pkts/s
Packet payload 1024 bytes
CBR real traffic rate ∼300 kbps
Flowduration 5min/flow
TABLE III
CHA RACTERISTICSOFTRA NSMIT TEDU DPFL OWS.
–RTT results averaged over 60 iterations.
–ICMPv6 requests did not overlap with UDP flows.
•Setup for Control Traffic Measurements
–Routers were switched within [0,2]sec.
–Results averaged over 84 iterations.
–ospf6d daemon restarted in each iteration.
–UDP flows: see Table III.
•OSPF Parameters: see Table IV.
Name Value
HelloInterval 2sec
DeadInterval 10 sec
RxmtInterval 5sec
AckInterval 2sec
MinLSInterval 5sec
MinLSArrival 1sec
LSRefreshInterval 60 sec
TABLE IV
OSPF PARAMETERS.
REFERENCES
[1] Cordero, J. A.; Philipp, M.; Baccelli, E. (2011). Compound
Wired/Wireless Internetworking with OSPF,Research Report #7642,
INRIA, Jun. 2011.
[2] Cordero, J. A.; Baccelli, E.; Clausen, T.; Jacquet, P.(2011).
Wired/Wireless Compound Networking. In: Wang, X. (Ed.). Mobile
Ad-Hoc Networks: Applications,InTech Publishers, ISBN 978-953-
307-416-0.
[3] Cordero, J. A.; Clausen, T.; Baccelli, E. (2011). MPR+SP –Towards
aUnified MPR-based MANET Extension for OSPF.Proc. HICSS’44,
Hawaii (US).
[4] Baccelli, E.; Cordero, J. A.; Jacquet, P.(2010). Optimization of
Critical Data Synchronization via Link Overlay RNG in Mobile Ad
Hoc Networks. Proc. MASS’44,San Francisco (US).
[5] Cordero, J. A. (2010). MPR-based Pruning Techniques for Shortest
Path Tree Computation. Proc. 18th SoftCOM,Split (Croatia).
[6] Roy,A.; Chandra, M. (2010). RFC 5820, Extensions to OSPF to
Support Mobile Ad Hoc Networking,IETF,Mar.2010.
[7] Baccelli, E.; Cordero, J. A.; Jacquet, P.(2009). MPR Techniques with
OSPF on Ad Hoc Networks. Proc. 4th ICSNC,Porto (Portugal).
[8] Ogier,R.; Spagnolo, P.(2009). RFC 5614, Mobile Ad Hoc Network
(MANET) Extension of OSPF Using Connected Dominating Set
(CDS) Flooding,IETF,Aug. 2009.
[9] Baccelli, E.; Jacquet, P.; Nguyen, D.; Clausen, T.(2009). RFC 5449,
OSPF Multipoint Relay (MPR) Extension for Ad Hoc Networks,
IETF,Feb.2009.
[10] Coltun, R.; Ferguson, D.; Moy,J. (2008). RFC 5340, OSPF for IPv6,
IETF,Jul. 2008.
[11] Conta, A.; Deering, S.; Gupta, M. (2006). RFC 4443, Internet Control
MessageProtocol (ICMPv6) for the Internet Protocol Version 6
(IPv6) Specification,IETF,Mar.2006.
[12] Henderson, T.; Spagnolo, P.; Pei, G. (2005). Evaluation of OSPF
MANET Extensions,Tech.Report D950-10897-1, Boeing, Jul. 2005.
[13] Clausen, T.; Jacquet, P.(2003). RFC 3626, Optimized Link State
Routing Protocol (OLSR),IETF,Oct. 2003.
[14] Henderson, T.et al. (2003). AWireless Interface Type for OSPF,
Proc. MILCOM’03,pp. 137-145, IEEE ComSoc, Boston (US).
[15] Perkins, C.; Belding-Royer,E.; Das, S. (2003). RFC 3561, Ad hoc
On-Demand Distance Vector (AODV) Routing,IETF,Jul. 2003.
[16] Halabi, S.; McPherson, D. (2000). Internet Routing Architectures,
2nd Edition, Cisco Press, ISBN 1-57870-233-X.
[17] Moy,J. (1998). RFC 2328, OSPF Version 2,IETF,Apr.1998.
[18] Oran, D. (1990). RFC 1142, OSI IS-IS Intra-domain Routing Proto-
col,IETF,Feb.1990.
[19] F.Baker et al. (2003). Problem Statement for OSPF Extensions for
Mobile Ad Hoc Routing,IETF Internet Draft, Sept. 2003.
[20] Baccelli, E.; Cordero, J. A.; Jacquet, P.(2010). OSPF over Multi-
Hop Ad Hoc Wireless Communications,In: IJCNC, Vol.2, No.5, Sept.
2010.
[21] Urban Wireless Community Mesh Networks: Freifunk Berlin
www.freifunk.net, GUIFI Barcelona www.guifi.net, OPEN AIR
Boston www.openairboston.net
