Review of Routing Protocols in VANETs

(1)

Review of Routing Protocols in VANETs

by

Rana Moazzam Tufail

B.E. Dawood College of Engineering and Technology, Karachi, Pakistan, 2008

A Project Submitted in Partial Fulfillment of the Requirements for Degree of MASTER OF ENGINEERING

in the Department of Electrical and Computer Engineering

c

Rana Moazzam Tufail,2016 University of Victoria

(2)

i

Supervisory Committee

Dr. Fayez Gebali, Supervisor

Department of Electrical and Computer Engineering

Dr. Samer Moein, Departmental Member

(3)

ii

Abstract

Vehicular Ad Hoc Network (VANET) is becoming an important technology which col-laborating ad hoc network, wireless LAN (WLAN) and cellular technology to attain intelligent communication mechanism. Due to rapidly changing topology, obstacles in communication network and limited mobility in VANET, there is a need of intelligent and efficient routing protocol which promise improved efficiency in terms of minimizing delay, increase throughput and reliability. A review of most recent protocols is presented by using few parameters of the network, location verification, clustering, routing technique, delay, control overhead and forwarding strategy. The review discusses the advantages and disadvantages of routing protocols for vehicular ad hoc networks. It inspects the need behind the design of these routing protocols. The review also includes Physical layer and MAC protocol structure for current vehicular ad hoc networks. Finally the review concludes discussing issues with routing protocol and Physical layer and MAC protocol with regard to VANETs.

(4)

List of Figures

1.1 VANET Communication Model [1] . . . 1

1.2 Cellular, Ad hoc and Hybrid Networks [2] . . . 2

2.1 Taxonomy of VANET Routing Protocols . . . 4

2.2 Route Discovery Mechanism [2] . . . 6

2.3 Route Maintenance Mechanism [3] . . . 7

2.4 Height of each node for updated message delivery [4] . . . 8

3.1 IEEE 802.11p Channel Frequency Band [5] . . . 15

3.2 IEEE 802.11p Protocol Stack and Sub-layer of PHY [6] . . . 16

4.1 Data link layer of WAVE Protocol Stack [7] . . . 20

4.2 Node Priority Process [8] . . . 21

(7)

List of Tables

3.1 Comparison of PHYs implementations in IEEE 802.11a and IEEE 802.11p[21] 17

4.1 EDCA parameter settings for applications in IEEE 802.11p [8] . . . 21

(8)

vii

Acknowledgements

I would like to express my honest gratitude and deepest appreciation to my supervisor, Dr. Fayez Gebali for his guidance, time, knowledge and support in the pursuit of my studies and in the completion of this project. I am deeply thankful and grateful to my lovely parents, my brother Mohsin, without his support and motivation, it would not have been possible. I would also like to thank my wife, my sisters and all those who have supported me throughout this entire process, by keeping me harmonious, motivated and helping me put the pieces together. I will be grateful forever for their love.

(9)

viii

Dedication

This work is dedicated to my mother and my father, they are my strength, courage and belief. I love you both.

(10)

Chapter 1 Introduction

VANET has become very challenging technology in recent years. It’s unique due to highly dynamic in nature and intermittent connectivity. Figure 1.1 explains the communication types in VANET. Two types of communication takes place, Vehicle to Vehicle (V2V) and Vehicle to Infrastructure (V2I). Topology of network has changed from virtual to real as nodes are replaced by vehicles acting as a router and client at the same time to share information in between using wireless links. Operating mechanism of VANET is based on information collected from traffic and road environment such as traffic congestion, accidents, warning messages etc. There are two types of communication in Vehicle to Vehicle, single hop communication and multiple hop communication. There are various research projects which are related to VANET such as CarNet, NoW, DRIVE, Fleet Net and CarTALK. There are number of applications such as Net access, Security distance warning, Driverless Vehicles, Cooperative driving, Auto Parking, Driver help and Vehi-cle collision warning. Need of efficient routing protocol is necessary in order to adapt dynamic mobility of VANETs.

(11)

2

Figure 1.2 shows the architecture of Vehicular Ad hoc Networks(VANETs), categorized into three major parts (a) Pure Cellular, (b) Pure Ad hoc, (c) Hybrid. The network of Pure Cellular Architecture works in such a way that the access points and cellular tower are connected to the internet to facilitate vehicular applications. In this architecture vehicles can easily communicate by connecting internet with a wireless ac-cess point or cellular gateway. Because of some geographic boundaries nodes can only communicate with each other. However, the sensors help in informing about the traffic conditions and also help in solving police crimes. In Pure Ad Hoc architecture, the nodes execute vehicle to vehicle communication with each other. Roadside Cellular gateways and access points help the vehicles which have wireless networking devices in communi-cating with one another. Numerous applications in parts of urban monitoring, driving assistance, safety and entertainment have used communicating units to access vigorous information outside their network and share it through ad hoc infrastructure less com-munication. As far as the hybrid architecture is concerned, it offers richer contents and better flexibility in content sharing. In hybrid architecture the nodes act as servers and they share the information like peers. These nodes are mobile thus, they make data transmission less consistent.

Figure 1.2: Cellular, Ad hoc and Hybrid Networks [2]

Highly dynamic topology is one of the most important feature of VANETs. The topol-ogy always keeps on varying because the vehicles move at a high speed. For instance if there is 250m radio range between two vehicles then their link would last for almost 10 seconds. Vehicles follow a certain mobility pattern that is a function of the under-lying roads, traffic lights, speed limit, traffic condition and drivers’ driving behaviors. Because of the particular mobility pattern, evaluation of VANET routing protocols only makes sense from traces obtained from the pattern. In frequently disconnected network

(12)

3

the link between the vehicles can rapidly disappear as the nodes transmit information. This problem is aggregated by heterogeneous node density e.g Urban environment and Rural environment. In addition non busy hours result in low node density, which results in disconnectivity. A robust routing protocol needs to recognize the rapidly changing topology and provides alternate paths to ensure smooth communication. In VANETs, propagation is not free due to many obstacles on and off the road like buildings, trees, pedestrians and vehicles. A VANET propagation model should be smart enough to take obstacle in consideration which can cause wireless communication interference. Nodes in VANETs are not subject to power and storage limitation as in sensor networks, another class of ad hoc networks where nodes are mostly static. Nodes are assumed to have ample energy and computing power. Therefore, optimizing duty cycle is not as relevant as it is in sensor networks. Nodes are assumed to be equipped with sensors to provide information useful for routing purposes. Many VANET routing protocols have assumed the availability of GPS unit from on-board Navigation system. Location information from GPS unit and speed from speedometer provides good examples for plethora of in-formation that can possibly be obtained by sensors to be utilized to enhance routing decisions.

(13)

Chapter 2 Routing Protocols

A routing protocol governs the way that two communication entities exchange informa-tion. It includes the procedure of establishing a route, decision in forwarding and action in maintaining the route or recovering from route failure. This section describes different routing protocols proposed in the literature where a single data packet is transported to the destination node without any duplication due to the overhead. As shown in Figure 2.1, protocols are classified into four categories, Topology Based, Position Based, Broad cast and Geo cast based.

(14)

5

2.1 Topology Based Routing Protocols

Routing protocols use already available links to transmit the data in networks. Dynamic routing decisions in the network are made by efficient routing protocols. Topology based routing Protocols are classified into Proactive and Reactive.

Proactive routing carries the distinct feature, regardless of the request from the node, the routing information such as the next forwarding hop is maintained in the background. Flooding of control packets in the network are constant to maintain the path or link among any pair of nodes. Due to that table is built in a node with each entry in table points to next hop node toward a specific destination. Advantage of table driven routing is that there is no searching or route discovery as destination path is already maintained in background.Though it provides minimal latency for real time applications, most of its bandwidth is consumed by unused paths, that creates overhead particularly in high mobility. Protocols are normally based on shortest path algorithm.

Reactive routing protocols are opposite in nature to Proactive routing protocols, table is not maintained when topology changes. In Reactive routing, route only initiates when nodes want to communicate with each other. It helps minimizing the overhead on network as this is the only communication taking place in the network. In order to send data, query packets are flooded into the network in search of a route to destination. Path is stored until that other node is irresponsive.

2.1.1 FSR-Fisheye State Routing Protocol

FSR [2] is a link state routing and maintains full topology map at each node, periodi-cally exchange HELLO packets and periodiperiodi-cally exchange of topology tables within local neighbors instead of flooding the network. Updates are frequently sent to nearby desti-nation then to remote destidesti-nation. In order to reduce size of update routing message, topology table use different frequencies for different entries depending on hop distance to the current node.

2.1.2 AODV-Ad hoc On Demand Routing Protocol

Figure 2.2 explains the route establishment process in Ad hoc On Demand Distance Vector (AODV) [4]. Route Request (RREQ) is generated in search of destination, each node which receives RREQ, store sending node address in routing table. When request finally reaches to destination, a Request Reply (RREP) is sent back to the same path. To

(15)

6

keep the update routing information and to prevent loops AODV uses Sequence number maintained at each node and carried by all routing packets.

Figure 2.2: Route Discovery Mechanism [2]

2.1.3 DSR-Dynamic Source Routing Protocol

DSR [2] objective is to provide highly reactive process by implementing routing tech-nique with very low overhead and quick reaction to frequent topology changes. DSR does not require periodic HELLO messages as it is beaconless. DSR dynamically sends the packets in network, upon receiving the request destination node send reply and carries in header the route traversed packet, due to that path is established between source and destination node. Source node can receive and store multiple replies from destination which can be utilized in case of link termination. Here DSR has an advantage over AODV, instead of flooding the network in case of failure it has an alternate route to re-establish communication. Figure 2.3 showing available links to destination and infor-mation of route error is delivered through same path. Source have multiple routes to reach destination in case of link failure.

(16)

7

Figure 2.3: Route Maintenance Mechanism [3]

2.1.4 DSDV-Destination Sequenced Distance Vector Routing

Protocol

DSDV [9] deals with the routing loop problem. It provides loop free route by using short-est path algorithm, it carries dshort-estination sequence number in packet header. Protocol is carried by two types of packets, full dump and Incremental. First type, full dump packets contain routing information of all nodes which are broadcasted to neighbors and incremental packet deliver updates. Bandwidth is affected in full dump packets and the incremental packets affect overhead in networks. Both types make DSDV unsuitable for highly Dynamic VANETs.

2.1.5 TORA-Temporally Ordered Routing Algorithm

(TORA) [4] is a source initiated on demand routing protocol and it finds multiple routes from a source node to a destination node. The three basic functions of TORA are Route Creation, Route Maintenance and Route Delete. Route creation is done by QRY and UPD packets. QRY keeps the destination address for which the algorithm is running. UPD keeps the height of node I (Hi) for packet broadcasting. The height of destination is set to 0 and all other nodes’ height set to NULL. The source broadcasts a QRY along destination node’s ID. Node when receives a reply packet, will update its height only when height in reply packet has minimum of all heights from reply packets it has received till yet. After that, Reply packet will be rebroadcasted by the node. Invalid routes are erased by flooding clear packet (CLR) in the network. The advantages of TORA are that

(17)

8

the execution of the algorithm gives a route to all the nodes in the network and minimize communication overhead on topological changes. Maintenance of routes is complexed as TORA allow route to every node in network, particularly in highly active VANETs. Figure 2.4 is showing route Re-establishment on link failure 5-7, new reference level is 5.

Figure 2.4: Height of each node for updated message delivery [4]

2.1.6 ZRP-Zone Routing Protocol

ZRP [10] is a combination of proactive and reactive routing protocol. Network is di-vided into different zones, each zone contains number of nodes. Proactive routing is used if the packet is destined within the zone area and Reactive routing is used outside of zone. Longer routes are affected by overhead in proactive routing, ZRP minimizes control overhead for longer routes and eliminating the delays within zone. Disadvantage of ZRP protocol is that it’s not suitable for high density and rapidly changing topology of VANETs, because it works with proactive approach in large size zones and reactive in small zones.

2.1.7 DYMO-Dynamic On-demand Routing Protocol

DYMO [11] protocol is a reactive multi hop routing protocol. Like AODV protocol, sequence number is used to provide loop free paths. In DYMO a route request pro-cess aims to maintain information about all intermediate nodes. In addition, each node participating in an ongoing route discovery process have to gather information about a requested node as well as intermediate nodes. Specifically at higher density level, which

(18)

9

happens more often in VANETs, routing and transport protocols can cause the network overhead. Congestion is un avoidable when establishing a new path in the network and retransmission of packets will only create more congestion.

2.1.8 Pros and Cons of Topology Based Protocols

In order to route packets from source to destination, Topology based protocols utilize link information available in the network.

Pros

Discovery is not required.

Low latency for real time applications e.g Audio/Video streaming. No periodic messages.

Support unicast, multicast and broadcast message.

Cons

Frequent network changes may cause congestion in network. Huge amount of available bandwidth consumed by unused paths.

More control overhead as no control messages being triggered even on link failure.

2.2 Position Based Routing Protocols

In geographic (position-based) routing, node makes a decision on position of packet destination and next hop neighbor’s position. Neighbor’s position is determined by periodically sent Beacon messages. Nodes are neighbors if they fall under same radio range. Each node knows its’ location in Geographic routing and OBU (On board Unit) having GPS, so location of destination is already known to sender. As geographical routing protocol do not follow traditional protocol mechanisms i.e sharing of link state information with neighbor node or maintenance of routing table, it means less overhead and more scalability and more suitable for dynamic environment like VANETs.

2.2.1 AEGRP-An Enhanced Geographical Routing Protocol

The AEGRP [12] selects a best route based on road segments with variables like traffic saturation, road length, distance and velocities of vehicle. Each packet calculates the

(19)

10

path based on given road variables. Neighbors update the sender about vehicle velocities and distance on receiving the broadcast request message which contains a query about number of intersections, lanes and road length. On finding multiple routes to destination, source will chose the route with higher density due to better transmission coverage. If road densities are same, then priority will be given to shorter distance. On failure of finding neighbor node, it carry and forward packet till the discovery of appropriate neighbor. Protocol has a prediction behavior for neighbor discovery and it does that by calculating self velocity,distance and position. Network is suppose to work better with high velocities of vehicle.

2.2.2 WNPRP-Wagon Next Point Routing Protocol

The WNPRP protocol [13] assumes that the range of a Wagon in the network is around 500 m. Also each Wagon in the network should be able to gain sufficient knowledge about the nearby nodes. This is done by sending ‘Hello Message’ periodically with the nearby Wagons. This help in gaining the information such as the position of Wagon, speed and direction at which the Wagon is moving. The source from where the Wagon starts and the destination point is marked with the help of GPS. Wagon gather the information about network from its location. If data need to be transmit from source to destination, source will filter and discard the information about vehicles going on other route. Vehicles on the same route would be considered to avoid dispersion of packets.

2.2.3 GeoSVR-A Stateless Map Based Routing Protocol

GeoSVR [14] is combination of node location and digital map. Vehicle density is taken into account before route setup. If next hop is destination, source will transmit the packets directly. Second step is to select next hop, protocol selectively finds a neighbor within a range to avoid packet loss caused by wireless channel. Disadvantage of this protocol is strict mechanism of choosing nodes which can cause delay and loss of packets.

2.2.4 CAR-Connectivity-Aware Routing Protocol

CAR [15] is topology based on destination location, route selection process, data forward-ing and path maintenance, usforward-ing the concepts of Anchor and Guards. Every node in CAR transmits route finding request, on receiving request each node updates hop count, minimum and average neighbor list and send it back to sender node. If link breaks in

(20)

11

between the transmission CAR uses a technique GUARD to recover from link failure. CAR shows better performance in packet delivery and minimizing routing overhead in the network.

2.2.5 Pros and Cons of Position Based Protocols

Pros Scalability

Maintenance and discovery of routes is not required. Efficient in rapidly changing mobility pattern. Low overhead.

Cons

Obtaining exact location.

Does not guarantee connectivity indoors or underground locations e.g tunnels. Obstacles on highways

2.3 Broadcast Based Routing Protocols

Protocol is used for flooding broadcast messages in the network. In case of emergency information needs to be broadcasted so other vehicles should know about it. Protocol broadcast the message to all neighbor nodes which can intensify the transmission. Pro-tocol has low packet loss ratio and more reliable in transmitting important information in the network.

2.3.1 EAEP-Edge Aware Epidemic Routing Protocol

EAEP [1] improves reliability, uses bandwidth efficiently and propagates information in an efficient manner. Protocol eliminates overhead of periodic Hello messages being exchanged among vehicles and simplifies the network maintenance. Each node push its own position to transmits packets in order to eliminate beacon messages. EAEP after receiving this information use number of transmission from front and back nodes to calculate the probability for either nodes retransmit the packets or not. EAEP addresses

(21)

12

flooding issue but the disadvantage is, it does not deal with link failures in network and increase packet delivery ratio.

2.3.2 DV-CAST-Distributed Veehicular Broadcast Routing

Pro-tocol

DV-CAST [16] keeps an update info about neighbors in order to initiate communication. Protocol works on multi hop scheme. DV-CAST gets information about the network from periodic beacon messages, it deals with different parameters of the network e.g vehicle density state, traffic lights, neighbor nodes etc. When source cannot find enough connected nodes it won’t broadcast, the packet will be stored till more nodes gets into broadcast range. Packet will be discarded if no node is found. Protocol enable message duplication awareness in nodes using flag parameter. DV-CAST is suitable for both high and low traffic density because it reduces broadcasting overhead. Disadvantage is data transmission delay and high control overhead.

2.3.3 SRB-Secure Ring Broadcast Routing Protocol

SRB make node rings based on their receiving power, rings are grouped as Inner, Outer and Secure rings [2]. Protocol controls retransmissions in the network and limit them to the rings to minimize overhead in order to achieve more reliable and stable network.

2.3.4 DADCQ-Distribution-Adaptive Distance With Channel

Quality Routing Protocol

DADCQ [17] aims for large networks with large node distribution. Nodes are selected on their geographical location before broadcasting a packet. Receiving node will make a decision to rebroadcast message on destination location, packet will not be rebroadcast if destination is close which minimizes network delay and increase network efficiency. Disadvantage is it creates a message overhead.

2.3.5 UMB-Urban Multi-hop Broadcasting Routing Protocol

UMB protocol [18] is developed to eliminate the issue of hidden node and packet collision simultaneously initiate communication in multi-hop broadcast. Nodes in UMB protocol

(22)

13

do not account previous network knowledge to forward packets and their acknowledg-ment. Protocol tries to reach to the farthest node while broadcasting. UMB is efficient in a network with high data loss and high traffic density.

2.3.6 Pros and Cons of Broadcast Based Protocols

Pros

Due to smaller number of nodes it has a high efficiency on highways. Reliability, since packet delivered to destination via multiple nodes. Effective in minimizing overhead due to broadcast storm mechanism. Cons

Due to reachability beyond transmission range it consumes significant network bandwidth.

Nodes receive duplicate messages due to flooding(Broadcast nature of protocol) Cause longer data transmission delays in network.

2.4 Multicast Based Routing Protocols

Multicast protocols use multi hop communication to send messages from sender to par-ticular cluster nodes. They are sub-classified into two classes Cluster based routing and geocast based routing Geocast Based Routing Protocol: Geocast routing protocols [9] belongs to a multicast routing protocol which based on sending packets from a source to a particular group of destinations. In geocast routing protocol one node can broadcast to other nodes fall under same geographical range, marked as zone of relevance (ZOR). Nodes under same geographical area are called members, if member goes out of that boundary then packet will be dropped. Zone of forwarding (ZOF) is point of interaction between zone and non-zone members. ZOF developed to provide a reliable packet’s de-livery in highly dynamic topology. Disadvantage of these protocols is transmission delay due to intermittent connectivity in network.

2.4.1 ROVER-Robust Vehicular Routing Protocol

ROVER [9] is a geographical multicast protocol. ROVER has somewhat similar AODV protocol mechanism, but more efficient and consistent as it only flood network with

(23)

14

control packets and unicasts the data packets. Another difference is that ROVER reply back to the node it received the packet rather then sending it to source node, which helps in creating more stable path to forward packets. ROVER assumes that each node have geographic location, identity and a map. Node initiate route discovery within its ZOR by sending route request message which contains source address, ZOR address, location and route sequence number. Another node only accepts the packet if it falls under his ZOR or ZOF, otherwise it drops the packet. Node on accepting the packet will reply including its ID and retransmits the packet. ROVER is considered as reliable routing protocol in VANETs, except couple of drawbacks, higher control overhead and delay in data delivery caused by retransmission of packets.

2.4.2 DG-CASTER-Direction-Based Geocast Routing

Proto-col for Query Dissemination

DG-CASTOR [19] gives an idea of link availability. Protocol does a node prediction in the network, identifies neighbors which have tendency to communicate with source at a particular time period. Protocol is designed for commercial use in VANETs. The main aim of DG-CASTOR is to build an essential commonality that is based on future location prediction of moving nodes in the network. This prediction behavior is known as Rendezvous group.

2.4.3 Pros and Cons of Multicast Based Protocols

Pros

Reduced Power consumption Reduced transmission overhead. Reduced control overhead.

Assure packet delivery in highly dynamic topology. Cons

Consume bandwidth.

(24)

Chapter 3 Physical Layer of VANET

The Vehicle to anything (V2X) communication system is an essential part of the Intelli-gent Transport Systems (ITS). Commercial and safety applications e.g traffic efficiency, driverless cars are matter of concern and need enhancements. The IEEE community is working on a new standard technology IEEE 802.11p modified for ITS communica-tion [17]. The IEEE 802.11p is expansion of IEEE 802.11 and also known as Wireless Access in the Vehicular Environment (WAVE). It uses the mechanism initially provided by IEEE 802.11 to operate in the Dedicated Short Range Communication (DSRC). Fea-ture of DSRC is higher transfer rates and low communication delays for small areas. It provides exchange of data between vehicles (V2V) and vehicle to roadside infrastructure (V2I) up to 1000m with transmission rate 3 Mbps to 27 Mbps and node speed up to 161 mph. IEEE 802.11p operates on about 9 channels and the frequency band used by each channel is described in Figure 3.1. CH172-5.860 GHz and CH184-5.920 GHz are safety dedicated channels. The first addresses to security solutions and other protect against congestion on other channels. Transmission broadcast and link creation is done by Channel CH178-5.890GHz which is a control channel. There is 5 MHz in the begin-ning of the band used as guard band (GB).

(25)

16

3.1 Physical Layer Architecture

The physical layer (PHY) is an interface between the MAC protocol and the media re-sponsible for sending and receiving frames. The PHY of the IEEE 802.11p is similar to the one of IEEE 802.11a [6]. It consists of two sub layers as shown in Figure 3.2. Physical Layer Convergence Protocol (PLCP) and Physical Medium Access (PMD). PLCP com-municates with the MAC and also converts the Packet Data Unit (PDU) coming from MAC to form an OFDM frame. The Physical Medium Access (PMD) defines the details of transmission and reception of individual bits on physical medium.e.g radio channels and fiber links. PMD is responsible to handle data encoding and perform modulation.

Figure 3.2: IEEE 802.11p Protocol Stack and Sub-layer of PHY [6]

3.2 Modulation

DSRC uses Orthogonal frequency division multiplexing (OFDM) modulation to multi-plex data, it divides the radio signal to multiple smaller sub-signals [6]. Main reason for using OFDM is, it utilizes the spectrum efficiently by allowing overlap. Transmission of all the sub carriers are simultaneous. OFDMA- Orthogonal frequency division multiple access offer shared access to multiply users by assigning subsets of subcarriers to indi-vidual users, allows lower data rates from multiple users. However, negative impact can be caused by high mobility such as message reception failure or packet errors. Message reception failure is due to nodes/receivers move out of sender transmission range dur-ing safety related message transmission. The other disadvantage is increase in packet

(26)

17

error rates and subsequently lower channel capacity due to high mobility cause intense Doppler’s spread on OFDM. Failure or delay in IEEE 802.11p networks could lead to hazardous situations, so those networks have to be very efficient and robust.

Regardless of many benefits, Physical layer encounter various challenges which are unique in comparison to other wireless networks such as : Collision Avoidance among vehi-cles in high mobility environment. Latency for VANET safety applications has to be 50ms and must not over 100ms. Doppler’s spread caused by high packet error rate and OFDM being sensitive to frequency offset may effect with lower channel capacity due to high mobility. To address these issues of IEEE 802.11p, few changes has been done in Physical layer parameter. Sub-carrier is using half clocked mode. For example, single IEEE 802.11a OFDM channel consists of 52 sub-carriers, 48 among them used for transmit data and 4 for pilot carrier, but single channel of IEEE 802.11p uses equal num-ber of sub-carriers changing the bandwidth/channel from 20 MHz to 10 MHz resulting in decrease of Doppler’s spread and interference. Data rate in 802.11p is 3 to 27 Mbps which is 6 to 54 Mbps in 802.11a. Comparison between parameters of IEEE 802.11a and IEEE 802.11p is shown in table 3.1. All the time parameters are double due to change in bit rate ( 20MHZ to 10MHz), subcarrier spacing is half and rest of the parameters are same as in IEEE 802.11a.

(27)

18

3.3 Challenges of PHY Layer

Still there are several challenges faced by 802.11p like: Effect of noise in bit and symbol energy, multipath effects, channel variation, channel estimation, network coverage range and bit rate enhancement techniques. Though these issues are partially addressed. An efficient PHY will eliminate these problems. PHY should be robust and scalable with low latency and minimum BER. Different transmission scenarios like urban, highways may cause PHY to perform within variable channels because rain, dust and other environmen-tal elements affect the transmission method. Apart from that, various technical factors can affect the PHY performance (Transmission quality), like unused carries, modulation, encoding, data rate and frame size.

(28)

Chapter 4 MAC Protocol of VANET

VANET creates a challenging scenario with quickly changing environment where move-ment of nodes are very high. Moreover it’s hard to keep track of nodes due to high mobility. In absence of Access point/Base station in decentralized network, it is difficult to manage the scalability problems. Frequency reuse or cell structure of cellular network is not possible to large extent. Moreover, efforts are being made in US and Europe to use single standard of frequency channel for transmission. Thus, the same radio spectrum will be shared by most vehicular communication links in a limited area, which will cause interference. The MAC method selects when a node has access to use the common com-munication channel. Which MAC method need to be used in a specific comcom-munication network is subjected to network topology and application. In a centralized network AP or BS has information about all the nodes within their range and traffic has to route through them. So, distribution of resources like frequencies and time slots can be done by MAC protocol. But, in ad hoc networks where there is no central mechanism and extreme movement of nodes in a given area, it is hard to implement an efficient MAC protocol. In VANET situation, MAC mechanism should be self- organizing, scalable and reliable. Moreover, absence of centralized network in ad hoc scenario, MAC mechanism has to be scalable so it can accumulate increasing number of nodes and resources. In VANET traffic safety applications, there are no controlled number of nodes and not known in advance. The MAC protocol for that reason has to satisfy the fairness, relia-bility and delay requirements even in high node density network.

WAVE protocol stack is shown in Figure 4.1. Layer is divided into two parts, MAC and logical link control (LLC) layer. LLC is responsible for point and multipoint communi-cation between wireless and wired channels.

(29)

20

Figure 4.1: Data link layer of WAVE Protocol Stack [7]

4.1 Medium Access Control

The IEEE 802.11p implicates the contention-based channel access EDCA (Enhanced Distributed Channel Access) as the MAC algorithm, it is a basic type of Distributed Coordination Function (DCF) by 802.11. EDCA applies Carrier Sense Multiple Access (CSMA) with Collision Avoidance (CSMA/CA) [8]. In CSMA/CA a node initially listen to the channel before starting communication and if the channel is free for Arbitration Inter-frame space(AIFS), the node will start to transmit by selecting a random back-off time. If the medium is or becomes busy in between that period the node will perform a backoff procedure, i.e. node will defer the access for randomized time period. To ensure significant safety communications and to find a reliable method of message transmission in rapidly changing topology, the 802.11p MAC protocol use various Access Classes (ACs) i.e queues for data traffic as shown in Figure 4.2. Data traffic is classified into four ACs : Background traffic (BK), Best Effort traffic (BE), Video traffic (VI) and Voice traffic (VO). Different ACs use different AIFSN and CW values, as shown in Table 4.1, it describes a set of four AC’s for station operation with mechanism to calculate CWmin and CWmax for each AC’s.

(30)

21

Figure 4.2: Node Priority Process [8]

Table 4.1: EDCA parameter settings for applications in IEEE 802.11p [8]

Ranking of transmission in EDCA is realized by a new Interframe Space (IFS) instead of AIFS, which is an extension of the back-off procedure in DCF [5]. The Short Interframe Spacing (SIFS), PCF Interframe Space (PIFS), and DCF Interframe Space (DIFS), are new AIFS values for various Access Class (AC) that are brought in EDCA.

(31)

22

4.2 Challenges of MAC Protocol

MAC protocol needs more enhancement in order to meet the highly dynamic topology of VANETs. Future work can be done on following areas:

Better Throughput : Safety related messages are necessary to be transmitted between vehicles sporadically and to have a better traffic control, Vehicle to RSU communication needs improvement. Therefore, high throughput is very important in VANETs.

Scheduling optimization: The proposal of multiple channel configuration has also been given rather than using single channel like Control Channel (CCH) and Service Channel (SCH). The method will guarantee the minimum delay while trans-mitting safety messages and making sure their delivery in dense traffic scenario. Traffic control: By using Back-off algorithms in MAC protocol the increase in traffic flow cause more contention periods. This can trigger packet collision be-fore reaching to destination, or there is a possibility that node want to broadcast would unable to initiate communication and in result cause maximum packet loss, therefore to avoid collisions and packet loss, traffic control needs to be taken care of.

(32)

Chapter 5 Conclusion

There are number of routing protocols designed for VANET. Many of them addressing to specific situation and specific issue. For example, CAR routing protocol address the issue where node receives an inaccurate info about neighbors and their location. On the other hand DSR works efficiently in low traffic scenarios and reacts well to rapidly changing topology, not effective when mobility is very high. Regardless of the fact that protocols are dealing with particular problems in routing environment, there is no agreed ground to authenticate their performance. Simulation tools or environments are not yet able to test routing protocols performance with regard to VANETs. In summary, new protocols are being designed, they are progressing and becoming established. Further, PHY layer and MAC protocol for VANET are discussed. Review has been done about parameters of both layers that are reformed in new technology IEEE 802.11p, finally highlighted the issues and possible research areas.

(33)

Bibliography

[1] B. Paul, M. Ibrahim, M. Bikas, and A. Naser, “Vanet routing protocols: Pros and cons,” arXiv preprint arXiv:1204.1201, 2012.

[2] K. C. Lee, U. Lee, and M. Gerla, “Survey of routing protocols in vehicular ad hoc networks,” Advances in vehicular ad-hoc networks: Developments and challenges, pp. 149–170, 2010.

[3] G. K. Walia, “A survey on reactive routing protocols of the mobile ad hoc networks,” International Journal of Computer Applications, vol. 64, no. 22, 2013.

[4] N. A. Rakhi and V. Dhakad, “Comparison between aodv, tora and dsr for provid-ing qos guarantees in mobile adhoc network,” International Journal of Scientific Engineering Research, vol. 4, no. 12, 2013.

[5] R. Popescu-Zeletin, I. Radusch, and M. A. Rigani, Vehicular-2-X communication: state-of-the-art and research in mobile vehicular ad hoc networks. Springer Science & Business Media, 2010.

[6] A. Abdelgader and W. Lenan, “The physical layer of the ieee 802.11 p wave commu-nication standard: the specifications and challenges,” in Proceedings of the World Congress on Engineering and Computer Science, vol. 2, 2014.

[7] L. Miao, K. Djouani, Y. Hamam, and B. Wyk, “A survey of ieee 802.11 p mac protocol,” Multidisciplinary Journals in Science and Technology, Journal of Selected Areas in Telecommunications (JSAT), vol. 1, 2011.

[8] C. Han, M. Dianati, R. Tafazolli, R. Kernchen, and X. Shen, “Analytical study of the ieee 802.11 p mac sublayer in vehicular networks,” IEEE Transactions on Intelligent Transportation Systems, vol. 13, no. 2, pp. 873–886, 2012.

(34)

25

[9] S. Allal and S. Boudjit, “Geocast routing protocols for vanets: survey and geometry-driven scheme proposal,” Journal of Internet Services and Information Security (JISIS), vol. 3, no. 1/2, pp. 20–36, 2013.

[10] T. T. Le, D. D. Tan, and D.-S. Kim, “Zone routing determination for izrp based on bee-inspired algorithm,” in Information Fusion (FUSION), 2012 15th International Conference on. IEEE, 2012, pp. 750–756.

[11] C. Sommer and F. Dressler, “The dymo routing protocol in vanet scenarios,” in 2007 IEEE 66th Vehicular Technology Conference. IEEE, 2007, pp. 16–20.

[12] K. N. Qureshi, A. H. Abdullah, R. W. Anwar, M. Anwar, and K. M. Awan, “Aegrp: An enhanced geographical routing protocol for vanet,” Jurnal Teknologi, vol. 78, no. 4-3, 2016.

[13] A. Chinnasamy, S. Prakash, and P. Selvakumari, “Wagon next point routing proto-col (wnprp) in vanet,” Wireless Personal Communications, pp. 1–13, 2016.

[14] Y. Xiang, Z. Liu, R. Liu, W. Sun, and W. Wang, “Geosvr: A map-based stateless vanet routing,” Ad Hoc Networks, vol. 11, no. 7, pp. 2125–2135, 2013.

[15] P. Darisini and N. Kumari, “A survey of routing protocols for vanet in urban sce-narios,” 2014.

[16] Y. MohanSharma and S. Mukherjee, “A contemporary proportional exploration of numerous routing protocols in vanet,” International Journal of Computer Applica-tions, vol. 50, no. 21, pp. 14–21, 2012.

[17] M. Slavik and I. Mahgoub, “Spatial distribution and channel quality adaptive proto-col for multihop wireless broadcast routing in vanet,” IEEE Transactions on Mobile Computing, vol. 12, no. 4, pp. 722–734, 2013.

[18] E. Spaho, L. Barolli, G. Mino, F. Xhafa, and V. Kolici, “Vanet simulators: A survey on mobility and routing protocols,” in Broadband and Wireless Computing, Communication and Applications (BWCCA), 2011 International Conference on. IEEE, 2011, pp. 1–10.

[19] B. T. Sharef, R. A. Alsaqour, M. Ismail, and S. M. Bilal, “A comparison of vari-ous vehicular ad hoc routing protocols based on communication environments,” in

(35)

26

Proceedings of the 7th International Conference on Ubiquitous Information Man-agement and Communication. ACM, 2013, p. 48.

Review of Routing Protocols in VANETs