Good Literature Review About Rl=4ht2σ2π E-Δv22σ2dt, 1
Vehicular Ad Hoc Network, abbreviated as VANET, is a system where vehicles like cars have the possibility of digitally communicating with one another, and communicating with a stationary object, for example a building or a traffic light. These communications are of short-range types. Concretely, the three types of communications provided in VANET are: inter-vehicle communication, vehicle-to-roadside communication and inter-roadside communication. In all these communications, there are no usage of a base station or any additional dedicated infrastructure like an access point to facilitate the digital communications. This is a unique property of VANET.
Through VANET, vehicles may be able to communicate among one another about a collision or an obstacle on the road, without most of the vehicles having seen the said collision or obstacle. In this sense, VANET is a self-organizing traffic information system. In VANET, each vehicle is considered to be a node and each of them is capable of passing information from itself to another node, i.e. to another vehicle. The transmission of information from the source node to the destination node is a path, and its procedure to find a path to it can be guided through a routing protocol. The stability of a path is an important issue to study in VANET since each of the nodes in VANET is most likely a moving vehicle, and not stationary. The movement of vehicles can be quite fast so much so that a path can be broken when a vehicle is not within the transmission range of another.
2. Link Reliability
The most challenging aspect of VANET communications is when one of the communicator, also known as a node, is a moving vehicle. With moving nodes, the communication link between nodes are relatively unstable due to a fast changing topology in the vicinity of the moving node. Kirtiga et al. in 2014 proposed a measure of a probabilistic link reliability r(l) between two vehicular nodes. This measure is computed as an integral with respect to an input quantity known as a transmission link between node i and node j over a certain time length which can be denoted as T. The integral kernel is a Gaussian-like probability density function. Mathematically, the link reliability r(l) can be expressed as follow:
where Δv=vr-μvr, vr is the magnitude of the relative velocity between two mobile nodes, μvr is the mean of vr, σ is the standard deviation of the distribution of the relative velocity, and H is the typical communication range in the VANET system, which was taken to be 300 meters by Kirtiga et al (2014). In this approach as proposed by Kirtiga et al., the variation in the velocity of the moving nodes, as well as the direction of the moving nodes are taken into account, as reflected in Equation (1). Prior to this, techniques calculating the linkage between two mobile nodes failed to take the aforementioned factors into account.
The modelling of the movement of vehicles are done with the assistance of graphs. Here, each node in a graph represents a vehicle. The linking between two nodes are time-ordered. There can be more than one path that links a source node and a destination node. Kirtiga et al. assigns each link which links two intermediate nodes along a path connecting the source and the destination node with a reliability value r(l). The total reliability value R of an entire route linking a source and a destination node is the product of all the reliability values along the route. Mathematically, this can be written as follow:
R=Π rlij, (2)
where i and j are two nodes found along a route linking the source and the destination node.
However, the reliability link is relatively mute on the issue of security attacks, malicious nodes and hackers. The quantity assumes that the reliability value can take into account how malicious nodes for example, can alter the connection between the source node and the destination node, and hence alter the path’s total reliability value.
3. Security Issues
In any wireless system such as VANET, security issues need to be addressed. If information to a vehicle node is tampered with, the vehicle node in question may be given mischievous instructions that could lead to road accidents. The security goals of VANET can be broadly categorized into five main components or criteria, which are, the authentication, the confidentiality, the integrity, the availability and the non-repudiation of a data.
Matthew and Kumar in 2013 described the authentication criteria as the assurance that the nodes in a communication link are the correct nodes matched in identity to the supposed nodes in a database. Therefore, for this to happen, each vehicle should have been registered and be uniquely identified with an authority prior to being driven on the road.
The confidentiality criteria of a data is due to the fact that all data should remain private and confidential, and not be disclosed to unauthorized figures. Disclosure of this information may lead to leakages of vital information about the nodes in a communication link, thereby allowing tampering and hacking of information to occur. Accidents can happen if the hackers intend to overwrite the information with new instructions that may cause the nodes to behave in a hazardous manner.
The integrity criteria of a data describes the fact that each data that has been received is assured to be exactly the same as the data that was sent out by a node, which has been authorized to do so by an authority. No modifications or compression of a data is allowed that renders the data sent being not the same as the data received.
The availability of a data should be guaranteed by a working system. Checks should be ensured that no battery power that generates the signal, sending and receiving one is drained. In addition, the routing protocol should be ensured not to be jeopardized by any hackers.
The non-repudiation criteria of a data is described by Matthew and Kumar (2013) as one that ensures that the vehicle node that has sent a piece of information cannot not be in denial that it was the sender. Likewise, the vehicle node that has received a piece of information is not allowed to be in denial that it has received the information.
Security attacks in VANET can be abundant, and are broadly categorized into several types by Matthew and Kumar (2013). The categories, include Sybil attack, denial of service (DoS), sinkhole attack, illusion attack, position attack, selective forwarding and misbehaving and faulty nodes.
In a Sybil attack, the mischievous node creates a multiple identity of itself. Concretely, there is only one node mapping to one vehicle in the physical world, but in the digital world, there are multiples of this one vehicle node. Consequently, other nodes which may receive information from one of this duplicate nodes, might be misled into thinking that there is a traffic congestion ahead, and thereby will choose alternative routes to reach their respective destinations. In this situation, the mischievous node is a selfish one, since no other nodes will be near it, thus having the entire road for itself. In order to prevent a Sybil attack from succeeding, a node can observe other nodes for a certain time interval, and the data from neighboring nodes are compared with the information gathered by itself. Any anomalies from the comparisons will be flagged as a potential attack.
In a DoS attack, the aim is to block nodes from accessing services and resources in the VANET network. This can be done through jamming the network, akin to spamming in emails. Consequently, nodes under DoS attack will lose connection with the network, which without information received from other nodes, may cause accidents next. The sinkhole attack acts differently from a DoS attack, but has similar intention. In this attack, the attacker node forces all information to pass through itself. Consequently, the attacker node who is now capable of controlling the information traffic, and thereby the traffic on the road, may be able to selectively reroute other vehicle nodes, or create malicious data packets to be sent to others, thereby causing accidents.
In an illusion attack, the malicious node deliberately create false signals or alter the signal originating from its own vehicle, before broadcasting it to other nodes. Consequently, other neighboring nodes fail to see the physical traffic but an imaginary traffic created by the malicious node.
An attack can also be in the form of a deliberate alteration of the distance information in the network, such that a transmission of data can be forced to pass through the node under malicious attack. Lu et al. (2013) described that the attacker will attempt to change the information on the distance value between the destination node and itself as less than other neighboring nodes. Through such modification, one of the hop on the routing will always be through the malicious node. Not only is the security of the data information in transmission compromised, the network routing overhead is also increased. It is important the overhead is not increased, as it can cause traffic accidents due to an information not reaching the destination node in a fast enough manner.
Similarly, Lu et al. (2013) remarked that the azimuth information can also be tempered. The azimuth information is the information on the angle between an imaginary line connecting two nodes, and an imaginary line drawn on the road. To transmit azimuth information, the law of right hand is used in protocols, such as VANET GPCR and GeoDTN+Nav, to transfer the information from the source node to the destination node. When a malicious attack occur, the law of right hand is no longer obeyed. Instead, the information is forwarded to another series of nodes, which some of them are not in the supposed route as determined by the law of right hand, and one of the nodes in the series is that of the malicious node.
An important consideration for VANET prior to a large-scale implementation of it in the real world is its speed of information exchange while at the same time, respecting the security and the safety of the data transmission. Asif et al. in 2010 introduced a mechanism to preserve these two factors in VANET, without compromising too much of one for the other, at a conference for communication software and networks. Prior to this, proposed techniques by others have resorted to sacrificing one of the factor for the other, mostly at increasing the speed of information exchange, but unfortunately with a side effect of reducing the security standard of the transmission. At the moment, the standard security mechanism used in the VANET community is known as the PKI/ECDSA security mechanism. This has been criticized by Asif et al. as needing too high a computational cost, thus reducing its practical usage in emergency situations, where information exchange, especially safety messages between vehicle nodes should be fast.
The crux of the idea from Asif et al. in 2010 is that the vehicles and nodes in their immediate vicinity form a trusted group relationship, thereby enhancing group communication with security standards. This framework used by them is a hardware-based one, and the trusted group relationship is build based on an important module, coined as Trusted Platform Module or TPM for short. The group entity can be formed based on which cell in the geographical location that a vehicle node belongs to. These cells are imaginary boxes that divide a road into segments. In each group, there will be a group leader and a number of group members. The group leader can be selected via its node position.
In the grouping method as proposed by Asif et al., the hardware used is a TPM chip which will be embedded in each vehicle. The chip contains not only the core TPM module, but an additional four sub-modules, known as an asymmetric module or ECC, a symmetric module, a random number generator or RNG, and a hash module. The TPM module functions to ensure that there is security during the transmission of messages, and that each component has not been tempered. The ECC module functions by generating a digital signature. This signature has a public key and a private key. The private key is burned into the TPM chip during manufacturing stage. The public key is widely distributed to the users in a VANET system. Seed numbers for the keys are generated by the random number generator. The hash value is provided by the hash module using the Secure Hash Algorithm or SHA1 technique. The hash value is used when selecting a group leader within a vehicular trusted group.
The symmetric module functions similarly to the EEC module, but is used primarily in event-driven information transfers. Event-driven information are usually safety and alarm-related messages to alert other vehicles in the vicinity of a hazard that one of the node, usually the group leader node has encountered. The other type of information transfer involves periodic safety messages which transfers information like the speed of the node, and its location and direction. This will be transferred using the EEC module, which has a slower transfer rate than the symmetric module, due to its slower generation of digital signature generation and verification.
Any node can leave the trusted group, for example when it leaves the cell that specifies the group. Likewise, any external node can be linked to the trusted group, and be a member of it at any time when it is within the vicinity of the trusted group. To check the feasibility of the trusted group technique applied in the real physical world, it is imperative that Asif et al. conduct simulations and present their results thereof in the future.
Another type of cellular group-style wireless networking is the Vehicular Mesh Network, or VMESH for short, which was proposed by Zang et al. in 2007. Here, vehicles which are within the transmission range of one another will form a group-like structure in VANET, also known as a mesh. If one of the node in a mesh has an information, it is assumed that all other nodes in the said mesh will also carry the same information via broadcasting the data from the initial node which receive that information to the other nodes. This is a useful and resource-saving networking method. As an example, if a mesh contains node A, B and C, and A happens to encounter an obstacle on the road. This information about an obstacle will be broadcasted to B and C, and a resolution to the obstacle will be determined, for example, with B and C slowing down or B and C completely avoid the obstacle altogether by diverting from their original course. Hence, it is not necessary that B and C experiences the encounter like what A did.
In VMESH, a temporary event such as the occurrence of an obstacle on the road, which is assumed that it will be removed from the road after a certain period of time, is known as a transient event of interest or EOI for short. In addition, the information stored in a node within a mesh that has an encounter with an EOI is in the form of a region of interest information or ROI for short. ROI defines the physical region that encloses the EOI, for example the exact location in coordinates and the size of the obstacle.
An information that is stored in one mesh can be passed to another mesh if both meshes are within a distance d from one another. This distance is not a distance set in stone, but is rather tunable. However, this distance d is calculated based on the location of where an obstacle has occurred for example, if the said information is about the obstacle itself. Suppose there is no other mesh within that region defined by the distance d, information loss occurs when the initial mesh that first encounter the EOI has travelled out of the said region. Hence, if there is a mesh within the region, the information about the EOI is still available when the initial mesh that first encounter the obstacle has travelled out of the region containing the EOI.
Liu et al. in 2010 has performed a simulation of a VANET scenario which utilized the VMESH protocol. The simulation lasted for three hours. Several assumptions were made which formed the inputs to the model. Firstly, the arrival time of a vehicle to a certain location was described by a Poisson distribution. Secondly, a free flow traffic was assumed. Thirdly, EOI could randomly appear on the road at any given time. Fourthly, the information about an EOI can be transmitted instantaneously to other nodes within a mesh. As Liu et al. used the VGSim software to perform their simulation, the vehicle nodes travel following the Nagel-Schreckenberg model which was a model introduced by Nagel and Schreckenberg in 1992.
The routing protocol of VANET can be broadly categorized into two, which are, the topology based routing protocol and the geography based routing protocol, as described by Paul et al (2011). The topology based routing protocol can be further divided into the proactive routing and the reactive routing protocols. In a proactive routing protocol, each node keeps a table of information about other nodes that are connected to it. An example of a proactive routing protocol is the Fisheye State Routing also known as FSR. The drawback to this is that the network load is tremendously large. Furthermore, the time needed to process the routing tables is large.
The reactive routing protocol avoids the problems face by the proactive routing protocol. In this protocol, nodes only start a route discovery if it requires information or need to communicate with another node. Consequently, the network load is tremendously reduced. An example of a proactive routing protocol is known as the Ad Hoc On Demand Distance Vector routing protocol. In this protocol, routing can be done in a unicast or a multicast mode. The drawback to this protocol is that there is, on average, a need for more time before an initial connection between two nodes can be established.
The geography based protocol does not keep any routing tables. Rather, it relies on the Global Positioning System also known as the GPS device to decide on routes. Nonetheless, the GPS is not available at all times, particularly when the vehicle node is in a tunnel, as the signal from the GPS is obstructed by the density of matter above the tunnel. In order to solve this problem, a system known as the Delay Tolerant Network (DTN) can be used. In this system, when a node is disconnected from other nodes, the data packet which should be sent to the other nodes is recalculated based on some metrics, and forwarding to other nodes is done thereafter. Another way is to drop completely the node from the contact list kept by each neighboring nodes. Here, each node has a beacon that transmits a short data packet to other nodes at each specified time interval. If no data is received from a node after a certain time period, the node is considered not within the vicinity of the neighboring nodes. To combine the advantages of both the geography based routing protocol and the topology based routing protocol, Lochert et al. (2003) proposed a new protocol known as the Geographic Source Routing protocol, also known as the GSR.
4.1 Data Broadcasting and Transmission
Broadcasting of data in a consistent manner to the environment in VANET can be quite costly and inefficient, particularly when the data is large in size. Furthermore, since vehicles are mobile nodes and thereby the topology of the network can be very dynamic, this type of communication between nodes is not effective at all. Janech et al. in 2012 proposed a solution to this based on the concept of a distributed database system. Their proposal also addressed the possibility of disconnection of nodes momentarily in a VANET which can be high in a dynamic network topology. It was proposed that communications be done in one of two methods. The first method is known as a pull method, and the second one is known as a push method.
In the pull method, a query node will first enquire for a piece of information from a data node. The data node then checks to see if it possess the necessary information, or nodes in its immediate vicinity may have. The data node will connect the query node to the respective node having the necessary information. If partial information can only be retrieved from any of the nodes the data node has searched from, the data node will return the query node with the partial information. It then awaits for further instruction from the query node. Upon acquiring further instruction from the query node, the data node will connect the query node with another possible node. This new node is decided based on the possibility that it may be able to assist in providing the necessary remaining information that the query node requires.
In the push method, a response is sent out via a broadcasting message from the data node at a certain specified time interval. This broadcast can be received by whichever node that may be in its vicinity. This response is an information based on a query that has been preprogramed. Any node upon receiving the broadcast, will check if the information contained in it is a data that is needed. If affirmative, the data will be processed accordingly.
For these methods, three algorithms were introduced by Janech et al. They are the Simple Pulling Algorithm (SPA), Independent Replication Algorithm (IRA) and the Dependent Replication Algorithm (DRA). The SPA algorithm is the simplest of the three. This algorithm starts with a query by a node to a list of data nodes, and then receives a response from them.
The IRA algorithm does similar to the SPA algorithm, but also additional steps. The IRA also collects data that a node may not require at that particular time, but deemed possibly useful in the future. The node also broadcasts, at a specified time interval, information regarding its local database to its environment. Based on the information contained in this local database, all other nodes in the vicinity will update their own local databases respectively.
The DRA algorithm is the most complicated of the three. It combines the techniques in both the IRA and the SPA algorithm. A key feature of the DRA algorithm is that each data that is available in a local database of a node is assigned a timestamp and an identifier. The timestamp will be used to determine if the data should be retrieved in any future queries. The identifier is used to identify the node that has sent the broadcasting signal to it.
A simulation on the feasibility of the methods had been done by Janech et al. The simulation was done using a tool known as adhocsim.fri. Each vehicle node was assign a travel speed. The vehicles were simulated to move in a uniform linear motion manner. When encountering a road section, the vehicles were simulated to slow down before continuing at the initially assigned speed at a new road section. In addition, the simulation had included the possibilities of buildings forming obstacles that may block the propagation of the broadcasting signal. For building corners, Janech et al. have assign a degree of tolerance that the broadcasting signal may propagate successfully. This is an effect of diffraction of the radio signal wave, which allows the signal to bend slightly when it encounters any obstacles.
Furthermore, Janech et al. have set a high number of cars on the road per simulation to reflect the possibility of a road during busy hours. The number was set to as high as 400 per simulation, and may change in any intermediate stages based on an exponential distribution. The broadcasting range of the signal was set to a maximum distance of 800 m.
All three algorithms, that is, the SPA, IRA and DRA algorithms were tested. Not surprisingly, the SPA algorithm, which is the simplest of the three algorithms, required the least amount of bytes for a transmission of data. The DRA algorithm required more amount of bytes than that of SPA. The IRA algorithm required the most amount of bytes among the three algorithms. In IRA, the nodes collectively produced the largest amount of replicas of data. Furthermore, in the IRA, the nodes distributed and shared the most amount of data. Due to this, it contributes a drawback in that the network load is high in the IRA compared to the other three algorithms. Given the conducive results that Janech et al. obtained from IRA, the authors will be focusing on the IRA algorithm, and will attempt to reduce the network load to improve the applicability of the said algorithm.
When a node intends to transmit a data to another node, it is possible that a transmission collision can occur when a neighboring node is transmitting out its data to the same node. Various protocols have been proposed to solve this issue. One of them is known as a directional antenna-based MAC protocol. Here, the space around the transmitter is divided equally in the angular space into N transmission angles. Not only can this reduce transmission collision, it can also help increase the usage of a channel and the capability to receive more transmission.
Nasipuri et al. in 2000 and Huang et al. in 2002 proposed a bi-directional antenna-based MAC protocol. The bi-directional is of interest, since it is designed around the fact that a single road can only accommodate at least one flow of traffic, and at most two flows of traffic, with each one flowing in parallel but in the opposite direction relative to the other. As any vehicle node can be surrounded by at most eight other vehicle nodes at any given time, this fact can be exploited in implementing VANET. Eight nodes since there will be one node in front of the said node on the same lane, one at the back on the same lane, three on the left lane, and another three on the right lane, to make eight nodes in total. This fact is used engineer an antenna which can be divided into eight directions for transmitting or receiving signals.
Ko et al. in 2000 proposed yet another directional-based protocol, also known as a Directional MAC protocol or D-MAC for short. In this protocol, a node acting as a transmitter will broadcast a request to send (RTS) signal to initialize a handshake to its neighboring nodes prior to sending any other data. When the receiving node is ready to receive the data, the receiver will send out a clear to send (CTS) signal after receiving an RTS signal. In a simple scenario involving three nodes, P, Q, and R, operating under a D-MAC protocol, when P intends to transmit some data to Q, first, P needs to initialize a handshake by sending out a directional RTS signal to Q. However, if R is within the vicinity, node R will also receive the RTS signal from P. Thereafter, R will block its antenna during the entire time period that P and Q is communicating with one another. Hence, by doing this, node R prevented itself from sending an RTS to node P during the communication period between node P and Q. It is important to note that R can still send RTS signals to other nodes nearby in a direction which is not blocked during the communication period between P and Q.
Directional antenna-based protocols need to ensure that communication is established within a time limit, where beyond this limit, communication may not be possible. This impossibility is due to having one of the node, for example the receiver moving at a speed that is too fast for it to return an RTS signal to the transmitting node. This problem is known as a deafness problem as described by Kim et al. in 2014. In fact, the deafness problem is a waste of resources, as any failed attempt to establish a communication with another node, will render retrials until it reaches a maximum number of retrials permitted. This leads to a reduction in the network performance.
While not able to solve the deafness problem, Nasipuri et al. in 2000 proposed a solution to the maximization of the resources in a node. Before a node attempts to communicate with another node using a certain channel, it has already tagged another channel as the data transmission channel which is idle. When the clear to send signal is received by the transmitting node, the same transmitting node will be able to start data communication process using the idle data transmission channel immediately without the necessity of a waiting time.
Directional-based MAC protocol has other weaknesses. In this protocol, it is not necessary that two nodes attempting to handshake and thereby establish a communication, will have the same directional transmission scope. Furthermore, the size of the signal coverage for each node may differ from one another. In this situation, if a transmitter, which incidentally has a much greater transmission range than the receiving node, will be able to send an RTS, but not receive any CTS or any signal at all from the receiving node. This is known as a hidden terminal problem as described by Kim et al. in 2014.
In a 2011 paper by Sjoberg et al., the hidden terminal problem in the carrier sense multiple access (CSMA) MAC protocol and self-organizing time division multiple access (STDMA) MAC protocol was examined. In CSMA, every node which intends to perform a transmission has to first perform a carrier sensing operation during a specified sensing period to determine if the channel for transmission is clear or busy. If it is busy, the node may not perform the transmission. Instead, it must perform a backoff procedure, that is, to wait for a randomized period of time before resuming the sensing operation. When at any time, the channel is free, the node may transmit without halt.
The STDMA protocol is similar to CSMA but differs in the sense that each node can access a channel in a timely manner without the need to wait for a medium to be clear before performing a transmission. Here, each node with the intention to transmit will listen to the channel. Then, the node will determine at which time slots the node will do the transmission bearing in mind that the time slots chosen are free beforehand, meaning no other nodes prior to it has chosen those time slots for their own transmission. Consequently, there is no randomized period of waiting time in the STDMA protocol.
The impact of the hidden terminal problem can be quantified through measuring the packet reception probability. The hidden terminal problem studied by Sjoberg et al. was formalized in a thought experiment in the following manner:
Let K and N be two distinct nodes. K is transmitting a data packet k within time t. During time t, N is also transmitting a packet denoted as n. Let M be another node intended to receive packet k and n. Furthermore, N is not an intended receiver of packet k. The radius Rk centered around node K defines the region where nodes within it are intended to be receiver of packet k. Hence, node M is inside the circle as defined by the radius Rk but node N is not. The radius Rn centered around node N defines the region where nodes within it are intended to be receiver of packet n. Hence, node M is also inside the circle as defined by the radius Rn but node K is not. Therefore, since K and N are transmitting packets at the same time, and node N is an intended receiver of both, there exist a data collision as experienced by node N. This data collision is defined as the hidden terminal problem in the study done by Sjoberg et al. (2011).
A case study using simulations was done by Sjoberg et al. (2011). In the simulation, a two-way highway scenario was set up, where five lanes were located for each direction. The vehicle arrival time is Poisson distributed. The speed of the vehicles was assumed to follow a Gaussian distribution with a 1 m/s standard deviation, and an average that depended on which lane the vehicle was travelling on. To make the simulation closer to the real world, fading was considered, and was assumed to model according to the Nakagami fading channel model.
Several reasons were stated by the authors as to why STDMA performed better compared to the CSMA protocol. Firstly, in the STDMA protocol, freely available time slots for transmission is broadcasted hence other nodes will already have partially know beforehand the slots to select for transmission. Secondly, in CSMA, the transmissions are not synchronized unlike in the STDMA rendering the existence of a higher interference in CSMA compared to STDMA. In STDMA, the synchronization is due to the partitioning of the time for transmission into slots indirectly a way to schedule the transmission for all nodes intending to transmit.
On the hidden terminal problem issue, Sjoberg et al. (2011) found that, if hidden terminal problem was experienced in either CSMA or STDMA, the packet reception probability decreased in comparison with when there was no hidden terminal problem. However, what was observed is that the decrease in the packet reception probability was about in the order of one percent. This indicates that the hidden terminal problem is not a serious matter of concern for a VANET system that utilizes either the CSMA or the STDMA protocol.
5. Simulation Software
MOVE uses the Java language, and is an attempt to model real-life traffic system into a user-friendly graphic user interface (GUI) display. The output of MOVE is a file containing the vectors of the directional movement of the vehicles. In MOVE, the user can create manually the road map, or allow MOVE to randomly generate one, or as a third option, import an existing real world road map from other programs, such as Google Earth.
In MOVE, a road map needs two main inputs which are the nodes and edges. Here, nodes are different from the definition of a node in VANET. Nodes in MOVE refer to an element of the road, such as a junction or a traffic light. Meanwhile, edges refer to the road itself. Attributes associated to an edge include the speed limit and the length of the road.
The movements of vehicles in MOVE is specified in the Vehicle Movement Editor, either manually or generated automatically. The user can specify the values to properties, such as the speed, and the time of the vehicle at a certain location. A unique feature of MOVE is that the user can also simulate the movements of public transport, including buses by entering as an input, the arrival time and stopping duration of buses at each stop. This makes MOVE a good tool to use for policymakers to implement new routes for public transport, or to analyze the conditions of the existing routes.
Vaishali and Pradhan have demonstrated the functionality of MOVE by generating a 150 nodes scenario with a 4 square kilometers road map. Information exchange in VANET is done at a rate of 4 packets per second, where each packet contains 64 bytes of information. In addition, they found that the packet delivery ratio correlates positively with the number of vehicles. However, they cautioned that this increase is only up to a certain limit before radio interference factor starts to take over and thereby deteriorate the delivery capabilities.
The VANET system has been studied for its effectiveness, drawbacks and ways to overcome those drawbacks by various researchers. At the moment, there are still issues regarding the security of a transmission, the reliability of a link and the effectiveness thereof in transmitting a data information. Currently, VANET systems can only be studied through simulation using softwares such as MOVE, since experimenting a VANET model in the physical world is costly and inefficient. Having said that, VANET remains attractive due to its minimal requirements which can be implemented without the fuss of having to prepare dedicated infrastructures on the road side to facilitate wireless communication.
Huang, Z. et al. “A Busy Tone-Based Directional MAC Protocol for Ad Hoc Networks.”
MILCOM. IEEE, 2002.
Janech, J. et al. “Comparation of Strategies for Data Replication in VANET Environment.”
Int. Conf. on Advanced Info. Networking and Appl. Workshop. IEEE, 2012. 575-80.
Khairnar, V.D. and S.N. Pradhan. “Mobility Models for Vehicular Ad-Hoc Network
Simulation.” International Journal of Computer Applications 11(4):8-12.
Kim, B. et al. “A Survey on Qualitative Analysis of Directional VANET MAC Protocols.”
Int. Journal of Contents 10(2):9-17.
Kirtiga, R. et al. “Reliable Graph Based Routing in VANET Environment.” Int. Conf. on
Recent Trends in Info. Tech. IEEE, 2014. 1-6.
Ko, Y.B. et al. “Medium Access Control Protocols Using Directional Antennas in Ad Hoc
Networks.” MILCOM. IEEE, 2002.
Liu, B. et al. “Assessing the VANET’s Local Information Storage Capability under
Different Traffic Mobility.” INFOCOM. IEEE, 2010. 1-5.
Lochert, C. et al. “A Routing Strategy for Vehicular Ad Hoc Networks in City
Environments.” Intelligent Vehicles Symposium. IEEE, 2003. 156-61.
Lu, C. et al. “Analysis of VANET Security Based on Routing Protocol Information.” ICICIP.
IEEE, 2013. 134-8.
Matthew, M.E. and A.R.P. Kumar. “Threat Analysis and Defence Mechanisms in VANET.” Int. Journal of Adv. Research in Comp. Sci. and Software Engineering 3:1 (2013):47- 52.
Nagel, K. and M. Schreckenberg. “A Cellular Automaton Model For Freeway Traffic.”
Nasipuri, A. et al. “A MAC Protocol for Mobile Ad Hoc Networks Using Directional
Antennas.” WCNC. IEEE, 2000. 1214-19.
Paul, B. et al. “VANET Routing Protocols: Pros and Cons.” International Journal of
Computer Applications 20:3 (2011):28-34.
Wagan, A.A. et al. “VANET Security Framework for Trusted Grouping using TPM
Hardware.” Int. Conf. on Communications Software and Networks. IEEE, 2010. 309-12.
Zang, Y. et al. “A Novel MAC Protocol for Throughput Sensitive Applications in Vehicular
Environments.” Vehicular Tech. Conf. IEEE, 2007.