Journal of Physics: Conference Series

PAPER • OPEN ACCESS Related content

- Fuzzy Logic-based Intelligent Scheme for
Simulation-Based Packet Delivery Performance Enhancing QoS of Vertical Handover
Decision in Vehicular Ad-hoc Networks
Evaluation with Different Parameters in Flying Ad- F. Azzali, O. Ghazali and M. H. Omar

- Performance Analysis of AODV Routing

Hoc Network (FANET) using AODV and OLSR Protocol for Wireless Sensor Network
based Smart Metering
Hasan Farooq and Low Tang Jung
To cite this article: A V Leonov and G A Litvinov 2018 J. Phys.: Conf. Ser. 1015 032178
- Delay-dependent asymptotic stability of
mobile ad-hoc networks: A descriptor
system approach
Yang Juan, Yang Dan, Huang Bin et al.

View the article online for updates and enhancements.

This content was downloaded from IP address [Link] on 30/05/2018 at 06:16

International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

Simulation-Based Packet Delivery Performance Evaluation

with Different Parameters in Flying Ad-Hoc Network
(FANET) using AODV and OLSR

A V Leonov, G A Litvinov

Omsk State Technical University, 11, Mira Ave., Omsk, 644050, Russia

E-mail: Kot@[Link]

Abstract. The article examines node connectivity as one of the most significant quality
characteristics of any information and telecommunication network. The article is aimed to
determine FANET parameters providing PDR (Packet Delivery Ratio) more than 90% for the
specified area. All results were received on the basis of the experimental study. Imitation
modeling based on an ns-2 network simulator was applied as a research method.
Characteristics of the network based on the simulation model utilizing the AODV and OLSR
routing protocols have been evaluated. The authors studied also the connectivity dependence
on such network parameters as the nodes number and the transmission range.

1. Introduction
Unmanned aerial vehicles (UAVs) have been quickly improving and are widely used, while wireless
data transmission technologies are also being developed rapidly. All this underlies the appearance of
the new communication networks types. One of these communication networks is Flying Ad-Hoc
Network – FANET [1].
It attracts specialists’ attention because it is a convenient platform for solving various tasks, mostly
related to the territories survey and facilities monitoring (see Fig. 1).

Figure 1. Example use of FANET in traffic monitoring.

Indeed, FANET represents a decentralized self-organizing network. Its nodes not only provide the
link with their neighbors, but they also forward the traffic passing through them.

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

2. FANET
Recently the world scientific society has been investigating the concept of using a UAVs group to
form UAV ad hoc network. This network will be able to perform one or more tasks.
FANET started to be researched and deployed intensely in the civilian sector almost at the same time
as a mini-UAV separate class was popularized [2]. This class is a multi-rotor system with an electric
power unit.
A high level of node mobility is characteristic of FANET; it is well over that of the other self-
organizing networks such as MANET (Mobile Ad Hoc Network) and VANET (Vehicular Ad Hoc
Networks) (see Fig. 2). Consequently, the network topology changes frequently and rapidly.

Figure 2. MANET, VANET and FANET.

As for the distances between nodes, in FANET, they are usually greater than in VANET or
MANET [3]. Thus, to organize UAVs connectivity, a longer range of communication should be
provided, which in its turn will have an effect on the radio channel nature and the equipment choice.

2.1. Network topology

In terms of topology, FANET is a distributed, peer-to-peer and multiply connected network, which
nodes connect "on the fly" following the "with each other" principle. It allows high availability to be
achieved, while a large number of links between UAVs gives a wide choice of traffic routes inside the
network.
This topology allows the task completion area to be expanded by data broadcasting, and the maximal
broadcasting time to be increased (up to round-the-clock) by successive or gradual UAV replacement
[4].
Besides, when one or several UAVs are disabled, intentionally or unintentionally, FANET enables an
improved network survivability by rearranging the network topology automatically.

2.2. Routing
The routing protocols utilized in FANET are required to feature the automatic search for the best route
(route group) to provide one or several qualitative parameters for the process of data transmission and
reception. Thus, one of the major tasks in FANET is to search for the route or the route group among
the network nodes. Moreover, such route parameters as connectivity (the route quantity), route length
(the number of hops or iterations in the route) and transit nodes share (the network nodes involved in
routing), influence the traffic performance in the network [5] .
During the investigation, the most widespread and widely used routing protocols in FANET were used
on the network level, that is reactive protocol AODV (Ad hoc On-Demand Distance Vector) RFC
3561 and proactive protocol OLSR (Optimized Link-State Routing) RFC 3626 [6].

2.3. AODV
In the reactive protocols, the node looks for the path to the destination point only when it has a packet
to deliver. To make a link, the node either uses an existing route or creates a new one based on the
data about available channels.
AODV protocol adapts quickly to dynamically changing network links [7]. Its overhead costs for data
processing and storage and low, and initialization time is short. The sequence number mechanism is
used in AODV to delete routing loops. Low, moderately and relatively high mobile nodes can be

2
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

manipulated and a mixed type of traffic is supported by this protocol.

Dynamic multiply connected routing with possibility of "self-starting" is realized in an AODV
algorithm. In AODV, routes to new destinations are received quickly and the routes to inactive nodes
are not required to be supported.
The distinctive feature of this protocol is the utilization of the recipient sequence number for every cell
in the routing table. The sequence number is generated by the recipient itself and, together with the
other data, it is sent by the route to the destination node. Sequence numbers usage provides freedom
from loops and it does not make programming more complex. When choice should be made between
two routes, the node must select the one with the largest sequence number.

2.4. OLSR
Proactive protocols are supposed to provide routing tables on each node, and these tables include
routes for reaching any network subscriber. To maintain the data integrity and reliability related to the
network structure, the nodes in such protocols exchange control messages.
OLSR is a variation of a classical link state routing protocol according to the requirements of ad hoc
network [8]. The concept of MultiPoint Relay (MPR) underlies OLSR. This paradigm lowers the
transmitted data amount significantly as compared to a traditional broadcast process with each node
forwarding the arriving messages to all its neighbors.
Proactive protocols are supposed to provide routing tables on each node, and these tables include
routes for reaching any network subscriber. To maintain the data integrity and reliability related to the
network structure, the nodes in such protocols exchange control messages.
OLSR is a variation of a classical link state routing protocol according to the requirements of ad hoc
network. The concept of MultiPoint Relay (MPR) underlies OLSR. This paradigm lowers the
transmitted data amount significantly as compared to a traditional broadcast process with each node
forwarding the arriving messages to all its neighbors.
Moreover, periodically each node sends control packets, and thanks to that the protocol is immune to
partial loss of control messages, this being a frequent case with broadcast packets in wireless ad hoc
networks.

3. Problem statement
The concept of FANET utilization revealed a lot of problems to researchers and scientists. Their
solution has been widely investigated by an increasing number of scientific groups recently.
One of the major problems is the loss of connectivity. The connectivity represents the ability to
deliver data from the source node to the destination. Mini UAVs can travel with high speed (up to 60
m/s) [9], and during this movements working nodes can be switched off and new ones can be
connected. Hence the network topology is prone to fast and frequent variations. So there are the
transmitted data delivery routes [10].
The FANET communication nodes area is limited by the antenna parameters, the transmitter and noise
powers and other transmitters jamming [11].

3.1. Connectivity
Let us assume that connectivity depends on the distance between nodes, while the node transmission
range represents a 2D disk with the radius R and uses omnidirectional antennas having a circular
pattern. Two nodes are considered as connected if they can transmit and receive the data from each
other directly or via other nodes. Otherwise, the data transfer service is not available between the
nodes [12] (see Fig. 3).

3
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

Figure 3. Modeling the topology of UAV ad hoc network

The nodes number depends on the specific network purpose, and it is likely to be selected based on the
connectivity. However, the network nodes may change their location at the time of operation and
break down. Thus, the nodes distribution can be supposed to be random, just like the connectivity [13].
The nodes number depends on the specific network purpose, and it is likely to be selected based on the
connectivity. However, the network nodes may change their location at the time of operation and
break down. Thus, the nodes distribution can be supposed to be random, just like the connectivity.

3.2. Connectivity probability determination

The generalized Erdos-Renyi model allows us to describe the connectivity probability of random
graph . A graph containing nodes sets the network model, and an edge
between vertices and in the graph is specified with probability , where is the random quantity
of edges in the graph [14].
Then, the probability space is defined in the following way:
(1)
where:

.
If one captures some graph , where:

(2)
where is mathematical expectation of the edges number in a graph, then for the random graphs
described according to this model, there is a known theorem stating that , where c is a
constant, then for the graph is almost never connected, and for , it is almost always
connected.
In fact, expression determines the edge probability threshold. If it is exceeded, the network is
connected with a probability higher than 0.5 [15].
For sufficiently large n and c, one may estimate the probability of the graph connectivity as:

. (3)

4
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

4. Experiment
Since there are no available real UAVs, and to generate a real FANET scenario requires as much
resources as there are people, the network simulator was utilized to perform UAVs movement and to
enable communication between them.
The network environment that is maximally close to the real world one must be proposed in the
simulation task. Taking this into account, one should try to find real-life scenarios of FANETs
utilization. The current section describes the air-to-air scenario that was utilized in the simulations.
Next the experimental setup including the parameter setting for ns-2 simulation will be presented [16].

4.1. Performance Metric

To estimate the different configurations fitness or quality (tentative solutions), a communication cost
function was defined with the use of a wide-spread metric for this area, that is the packet delivery ratio
(PDR) [17].
Packet delivery ratio (PDR). Some data packets are generated by an application and delivered by a
routing protocol [18]. A data packet is taken as delivered when it is received by the destination node in
full and without damages.

(4)
where is all packets received by the receiver successfully, is all packets
produced by the senders.

4.2. Scenario
Following this idea, a FANET sample was generated by mapping an area of 3,600×1,200 m2 [19]. A
rectangular space was chosen to urge the utilization of longer routes between the nodes than possibly
occurring in a square space, with the node density being equal.
Air-to-air wireless communications [20]: The communication between the ground base stations and
UAVs sets a number of constraints on the transmission ranges. To avoid them, UAVs are able to
connect with each other over pure ad hoc architecture. Moreover, one can use such wireless
communication when supporting multi-hop communications and different applications whenever the
data packet should be transmitted to another node outside of the range. Such topology is applicable in
the analysis of the node density effect on the performance.
The radio signal transmission-reception is assumed to be stable, and all UAVs transceivers have
identical settings. Under ideal conditions the network nodes are within the line of sight (LOS), that is
without obstacles, and without multiple possible paths of signal transmission and refractions [21].
To simulate the Atheros AR9344, the network interface physical radio characteristics for each mobile
node were selected, including the receiver sensitivity, the antenna gain and transmission power [22].

4.3. Mobility Model

Being a general-purpose network simulator, ns-2 does not provide a way of direct realistic FANET
simulations definition, with the nodes following the UAVs behavior. The Random Waypoint (RWP)
model was used for this problem solution to create realistic mobility models (Fig. 4). This model is
applied in the majority of simulation scenarios to create multiple movements with straight trajectories.
In these scenarios, each node chooses a random destination, then it moves with a random velocity and
pause time at the destination. When the pause time ends, a random destination is chosen by the node
again with a random velocity and an analogous pause time on the basis of fixed probabilities [23].

5
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

Figure 4. Random waypoint (RWP) mobility model.

A number of UAVs (from 30 to 150) are applied in the simulation with specified UAVs
direction/speed changes. During the simulation we considered different levels of UAV density, UAV
velocity (between 0 and 60 m/s), and network activity (from 15 to 75 connections).

4.4. Traffic Generation Model

In this FANET scenario, a specific data flow representing various possible communications precisely
was specified. The data flow model runs from 15 to 75 cycles of the CBR (Constant Bit Rate) network
application via UDP source agents built in UAVs, thus interconnection between each other is realized
by our mobility model. CBR data packet size is 512 bytes, while the packet rate is 10 packets per
second [24]. For future recreation purposes Table I includes the other simulation parameters. A fixed
data rate was selected because our objective was not to investigate the maximum throughput, but to
study the AODV and OLSR abilities to locate and keep routes successfully.

Table 1. Simulation Parameters In NS-2

Parameter Value
Simulation time 300 s
Simulation area 3,600×1,200 m2
Number of UAVs 30, 50, 100, 150
UAV speed 0-60 m/s
Propagation model Friis
Radio frequency 2.437 GHz
Channel bandwidth 6 Mbps
Antenna model Omni
Mac protocol 802.11g
Transmission range of UAVs 250, 500, 750, 1000
Routing protocol AODV (RFC 3561), OLSR (RFC 3626)
Transport protocol UDP
CBR data flow 15, 25, 50, 75
TCP sources were not applied as TCP offers the network a conforming load, which means it alters
times of sending packets based on its evaluation of the network ability to transmit packets. As a result,
the protocols differ both in the times of data packet origination performed by its sender and in the node
position when the packet is transmitted, thus making a direct comparison between them impossible
[20].

4.5. Experimental Settings

The simulation phase is performed by running an ns−2 simulator (version ns−2.34) based on the UM-
OLSR (ver. 1.0 patch) implementation of OLSR. Five separate runs of each optimization method were
made on Lenovo ThinkStation P320 SFF with Intel Xeon vPro E3-1245 v6, 16 GB of RAM, and O.S.

6
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

Ubuntu Desktop 16.04.3 LTS 64-bit.

A different UAV mobility and interactions are used by each single iteration, based on independent
random sources inside each simulation, hence contributing to the generalization of the results.

5. Modelling results

5.1. Packet Delivery Ratio

In the simulation results analysis we obtained the dependencies of the network connectivity providing
PDR>90% on the network parameters.
Figures 5 and 6 demonstrate simulation modelling results of 30-node network with the UAV
transmission range of 250-1000 m with 250 m increments.

Figure 5. AODV protocol PDR dependence on the transmission range and velocity for 30 nodes.

7
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

Figure 6. OLSR protocol PDR dependence on the transmission range and velocity for 30 nodes.
Figures 7 and 8 show AODV and OLSR protocols PDR dependence on the nodes number, a
transmission range and velocity for 50 UAVs with a node transmission range of 250-1000 m in
increments.

Figure 7. AODV protocol PDR dependence on the transmission range and velocity for 50 nodes.

Figure 8. OLSR protocol PDR dependence on the transmission range and velocity for 50 nodes.
Figures 9 and 10 represent the imitation modeling results for the 100-node network for AODV and
OLSR protocols with a UAV coverage radius in the range of 250-1000 m with 250 m increment.

8
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

Figure 9. AODV protocol PDR dependence on the transmission range and velocity for 100 nodes.

Figure 10. OLSR protocol PDR dependence on the transmission range and velocity for 100 nodes.

Figures 11 and 12 show PDR dependence on the nodes number, a transmission range and velocity for
150 nodes in AODV and OLSR protocols. The UAV transmission range changes from 250 to 1000 m
at 250 m interval.

9
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

Figure 11. AODV protocol PDR dependence on the transmission range and velocity for 150 nodes.

Figure 12. OLSR protocol PDR dependence on the transmission range and velocity for 150 nodes.

The results shown in Fig. 7, 8, 9, and 10 demonstrate that with 50 and 100 UAVs, connectivity
providing the required PDR level was achieved.
It follows from the data presented in Fig. 7 and 8 that sufficient connectivity probability can be
guaranteed by a smaller number of UAVs.
The desired PDR values were achieved with the used simulation settings and network
configuration. As the simulation results demonstrate, the transmission range should be m
and UAVs number should be from 50 to 100, to guarantee the required connectivity between nodes.
The diagrams in Fig. 9 and 10 demonstrate that a further increase in the nodes number does not
improve the quality indicators. Therefore, one may assume there is an optimal UAVs number, as well

10
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

as the transmission range for the chosen network topology, data transmission technology and traffic
parameters.

5.2. Dropped Packets

There are a number of reasons for packet loss in FANET. Dropped packets represent a significant
issue, which can be caused by many factors, e.g. network topology, node mobility, link interferences
and network overload.
Dropped Packets: packet loss takes place in a network if one or several packets travelling through a
network cannot reach their destination. One may estimate the total number of packets dropped during
the transmission as follows:

(4)
where is Sent Constant Bit Ratio; is Received Constant Bit Ratio.
In a wireless ns-2 model, packets are dropped due to several reasons. Data in Table II explain these
reasons [25].

Table 2. Explanation of Packet Drops on NS-2

Explanation of Packet Drop
MAC LAYER
COL Collusion
DUP Duplicate
ERR MAC Error
STA MAC State invalid
BSY MAC Busy
RET Retry Count Exceed
ROUTER LAYER
LOOP Loop on Route
TOUT Time Out
IFQ Interface Queue Full
ARP ARP Queue Full
OUT Outside Subnet
CBK Call Back
TTL Time to Live
NRTE No Route Exceed
GENERAL
END Simulation End

Figures 13, 14, 15, and 16 illustrate the congestion that occurs due to the buffer overload at the
Interface Queue (IFQ) level for 50 and 100 UAVs.

11
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

Figure 13. Dropped packets for 50 nodes of AODV protocol.

Figure 14. Dropped packets for 50 nodes of OLSR protocol.

12
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

Figure 15. Dropped packets for 100 nodes of AODV protocol.

Figure 16. Dropped packets for 100 nodes of OLSR protocol.

The node becomes a bottleneck when the packets arriving at the node have a higher rate than it can
transmit. If this is the case, a large number of packets are dropped at the IFQ buffer, which leads to a
throughput reduction. An overload control mechanism is needed to solve the issue. The flow rate must
be adjusted to reduce the excess packets number and to control the network overload for reducing the
packet losses and improving network performance.
Figures 13, 14, 15, and 16 demonstrate the absence of the packets dropped due to the ERR, STA,
BSY, TOUT and OUT, and drops due to the COL are few in number.
Even though in most cases packets are dropped because of Callback (CBK) and No Route Exceed
(NRTE), dropped packets appear on the MAC layer if after the 7th attempt of RTS (Request To Send)

13
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

it receives no CTS (Clear To Send) message. In such case, the MAC layer activates the router layer
because RTS RET signifies that the destination node is unavailable, which in its turn means that a
break is on the previously found route to destination. If this is the case, the router layer can drop all
packets in its queue. In the ns-2 simulation trace file, it is called as CBK. In addition, the route layer
does not send packets anymore, and the route establishment procedure is initiated.
Even if the node begins the route establishment procedure, there are still some neighboring nodes
able to send packets directed to a broken link to this node. When such packet arrives at the node, it
may be dropped because there is no route. This is called NRTE.
The queries from Link layer arrive at the Address Resolution Protocol (ARP) module that translates
the hardware addresses into the network one. If the hardware address for destination exists in ARP, it
is recorded into the MAC header of the packet. Elseways, an ARP query is transmitted, and the packet
is temporarily moved to cache. Every unknown destination hardware address has a buffer for a single
packet. If ARP receives additional packets to the same destination, the packet that has been buffered
before is dropped. When the hardware address for the packet's next hop is known, the packet is put
into the interface queue.
Each packet has a TTL value (according to RFC 791 [26] / RFC 1812 [27]). If the waiting time is
longer than TTL, the packet will expire. As the nodes number in the entire system increases,
contention is higher for a model based on contention, resulting in more dropped packets. In the
proposed model, when the nodes number rises, the packet waiting time rises too since the slot number
is fixed. More dropped packets result, because the TLL value of packet expires (see Fig. 14, 16).

6. Conclusion
FANET is the initial step towards cost-effective self-organizing networks. FANET Imitation modeling
in the ns-2 simulator, undertaken for the present research, demonstrated OLSR protocol operability.
Connectivity has several definitions, and their common element is that two nodes are called connected
if there is some criterion above the threshold. The authors defined the connectivity in FANETs in
terms of quality of service (QoS) metric, namely, packet delivery ratio.
The simulation results showed that FANET connectivity with PDR>90% is achieved when the
nodes number is 50 and 100, while their transmission range is 500 m.
The factor affecting UAV ad hoc network connectivity was experimentally proven to be the nodes
number distributed over the area and their transmission range.
However, there is also a need to rework the model to make it closer to the real network. This can
be achieved by setting "background" relevant traffic from other nodes, taking into account
administrative traffic, not related to routing, utilizing real video files as generated video traffic, and so
on.
It seems useful to introduce routing protocols, more modern and adapted for FANET, in a
simulation model.

References
[1] Bekmezci İ, Sahingoz O K and Temel Ş 2013 Flying Ad-Hoc Networks (FANETs): A survey Ad
Hoc Netw. 11 1254–70
[2] Valavanis K P and Vachtsevanos G J 2015 Handbook of Unmanned Aerial Vehicles (Dordrecht:
Springer Netherlands)
[3] Gupta L, Jain R and Vaszkun G 2015 Survey of Important Issues in UAV Communication
Networks IEEE Commun. Surv. Tutor. 1–32
[4] Maxa J-A, Mahmoud M-S B and Larrieu N 2017 Survey on UAANET Routing Protocols and
Network Security Challenges Ad Hoc Sens. Wirel. Netw.
[5] Tareque M H, Hossain M S and Atiquzzaman M 2015 On the Routing in Flying Ad hoc
Networks Computer Science and Information Systems (FedCSIS) Computer Science and
Information Systems (FedCSIS) (Lodz) pp 1–9
[6] Vasiliev D S, Meitis D S and Abilov A 2014 Simulation-based comparison of AODV, OLSR

14
International Conference Information Technologies in Business and Industry 2018 IOP Publishing
IOP Conf. Series: Journal of Physics: Conf. Series 1015 (2018)
1234567890 ‘’“” 032178 doi:10.1088/1742-6596/1015/3/032178

and HWMP protocols for flying Ad Hoc networks Internet of Things, Smart Spaces, and Next
Generation Networks and Systems (Springer) pp 245–252
[7] Perkins C, Belding-Royer E and Das S 2003 Ad hoc on-demand distance vector (AODV)
routing
[8] Clausen T and Jacquet P Optimized Link State Routing Protocol (OLSR) RFC 3626
[Link]
[9] Anon 2017 The world’s fastest drone hits 179mph Mail Online
[10] Frew E W and Brown T X 2009 Networking Issues for Small Unmanned Aircraft Systems J.
Intell. Robot. Syst. 54 21–37
[11] Olsson P-M, Kvarnström J, Doherty P, Burdakov O and Holmberg K 2010 Generating UAV
communication networks for monitoring and surveillance Control Automation Robotics &
Vision (ICARCV), 2010 11th International Conference on (IEEE) pp 1070–1077
[12] Perkins C E, Malinen J T, Wakikawa R, Nilsson A and Tuominen A J 2002 Internet
connectivity for mobile ad hoc networks Wirel. Commun. Mob. Comput. 2 465–482
[13] Shilin P, Kirichek R, Paramonov A and Koucheryavy A 2016 Connectivity of VANET Segments
Using UAVs Internet of Things, Smart Spaces, and Next Generation Networks and Systems vol
9870, ed O Galinina, S Balandin and Y Koucheryavy (Cham: Springer International Publishing)
pp 492–500
[14] De Pellegrini F, Miorandi D, Carreras I and Chlamtac I 2007 A graph-based model for
disconnected ad hoc networks INFOCOM 2007. 26th IEEE International Conference on
Computer Communications. IEEE (IEEE) pp 373–381
[15] Eletreby R and Yağan O 2017 Connectivity of inhomogeneous random key graphs intersecting
inhomogeneous Erd\Hos-Rényi graphs Information Theory (ISIT), 2017 IEEE International
Symposium on (IEEE) pp 2920–2924
[16] Hayat S, Yanmaz E and Muzaffar R 2016 Survey on Unmanned Aerial Vehicle Networks for
Civil Applications: A Communications Viewpoint IEEE Commun. Surv. Tutor. 18 2624–61
[17] Galinina O, Andreev S, Balandin S and Koucheryavy Y 2017 Performance Evaluation of
COPE-like Network Coding in Flying Ad Hoc Networks: Simulation-Based Study vol 10531
(Cham: Springer International Publishing)
[18] Oubbati O S, Lakas A, Zhou F, Güneş M, Lagraa N and Yagoubi M B 2017 Intelligent UAV-
assisted routing protocol for urban VANETs Comput. Commun. 107 93–111
[19] Anon The Drone Chart - 2017 Drone/Quadcopter Comparison
[20] Broch J, Maltz D A and Johnson D B 1999 A performance comparison of multi-hop wireless ad
hoc network routing protocols Parallel Architectures, Algorithms, and Networks, 1999.(I-
SPAN’99) Proceedings. Fourth InternationalSymposium on (IEEE) pp 370–375
[21] Tiwari A, Ganguli A, Sampath A, Anderson D S, Shen B-H, Krishnamurthi N, Yadegar J, Gerla
M and Krzysiak D 2008 Mobility aware routing for the airborne network backbone Military
Communications Conference, 2008. MILCOM 2008. IEEE (IEEE) pp 1–7
[22] Anon Atheros - WikiDevi
[23] Bujari A, Palazzi C E and Ronzani D 2017 FANET Application Scenarios and Mobility Models
(ACM Press) pp 43–6
[24] Zheng X, Qi Q, Wang Q and Li Y 2017 An adaptive density-based routing protocol for flying
(Ad Hoc networks)
[25] Nayak A K, Rai S C and Mall R 2016 Computer network simulations using NS2 (Boca Raton:
Taylor & Francis, CRC Press)
[26] Postel J Internet Protocol
[27] Baker F Requirements for IP Version 4 Routers

15