FANETs have evolved using the concept of MANETs and VANETs. To apply routing protocols to FANETs is one of the common challenges. Unlike FANETs, generic ADHOC networks are not capable of sending information quickly and correctly is some situations like at the time of natural disaster like flooding, earthquakes and even in military battlefield. Optimized Link-State Routing (OLSR), and Predictive-OLSR (P-OLSR) are the techniques for dynamic topology adaptation in accordance with the mobility pattern. OLSR is a routing protocol where every node maintains one or more tables representing the entire topology of the network. P-OLSR which is designed for FANETs is an extension of Optimized Link-State Routing. OLSR have a benefit of having routes quickly available whenever required. Simulation results verify that the proposed scheme could significantly increase the performance of flying ad hoc networks. 1 INTRODUCTION A Wireless ad hoc network is a set of different network nodes forming the temporary networks without the aid of any infrastructure or any centralized administrator. They do not have gateway, each node can act as the gateway. The wireless arena has been experiencing exponential growth in 21st Century. The IEEE Standard 802.11 is used for ad hoc networks which specifies Medium Access control and Physical layer. In ad hoc network every node has its own transmission range, which in turn combine to form a bigger transmission area. Nodes use hopping technique to transmit data. To make this process more effective , an appropriate routing algorithm should be implemented. Routing protocols are designed based on the requirement such as when the data is to be exchanged, what data should be exchanged, how routes are calculated etc. In proactive algorithms routes are calculated in the before hand , so nodes use these routes for information ex1 change. OLSR routing algorithm is an example of proactive. Whereas in reactive algorithms route is calculated only when there are nodes that need to be communicated. AODV is an example of reactive algorithm. Multiple paths between any two given nodes can have better throughput and recovering connection failure is also easy , but overhead of route discovery in multi path routing more when compared with a single path routing algorithm. In table driven routing protocol information about each node to other node is updated. Where as in on demand routing ,nodes calculate if they have to exchange data from each other. They don’t have to update information. For a routing algorithm to be best, it should overcome the routing problems in ad hoc network. Some of the routing problems are given below: • Asymmetric Link: If node 1 can hear node 2, that does not imply vice versa, hence routing algorithms should be asymmetric. • Redundant connection: It consumes lot of time to update routing table , if there are many redundant connections. • Interference: Interference is very common as the data transfer is done trough waves. Natural effects like weather, scattering can cause interference. • Dynamic topology : Since nodes can enter and leave network freely, it causes node to make frequent changes in their routing table. FANETs are the special form of MANETs and VANETs. They have much mobility degree and because of this topology changes more frequently than that of MANETs and VANETs. Distance between the nodes is also longer than that of the VANETs and MANETs. Hence communication range in FANETs also should be longer than that of others. Since the topology changes in an unpredictable manner, robust algorithm with dynamic routing are better to maintain communication among the nodes in the FANETs. Dynamic routing in ad hoc network is an reactive algorithm where the nodes recover the routes each time a node transfers data to other node. Since FANETs have high mobility nodes, route breakage is very common. Hence route is maintained in order to minimize the breakage. Network Simulator widely known as NS2, is an event driven simulation tool that has proved useful in studying dynamic nature of FANETs. NS2 enables users by providing with ways to specify network protocols and simulate their behavior. The paper has three sections. In the first section , OLSR protocol is analyzed , in the second part P-OLSR is analyzed. Third section is about the simulation results. 2 DESCRIPTION 2.1 OLSR OLSR is a proactive routing protocol that obtain the strength of link state algorithm. By proactive, it refers that the entire topology of the network is maintained by every node. Some of the OLSR performance include: 1. It declares only links with its multipoint relay selectors. 2. Every node select from its neighbor, who can read 2 or who can broadcast messages, those nodes are termed as multipoint relays (MPRs). 3. The link state information is partially distributed in the network, where the link is obtained only between an MPR and its MPR selectors. Only MPR can retransmit in this protocol. For instance, consider a node A having nodes B, C, and D as its neighbors. So, if A considers C as MPR, then whatever packet A sends is received by all the other three nodes. However, only C can rebroadcast that packet. Hence, this protocol significantly reduces the number of retransmissions in a broadcast procedure and can play a significant role for dense ad hoc networks. OLSR is not a central entity based but works on a completely distributed manner. Since the nodes send periodic control messages, it is not compulsory to have a reliable transmission of messages. Types of packets: 1. Hello packets: Hello packets are mainly used to know the state of the link. Every node transmits their neighbor’s information and the information about their link to its neighbors. 2. Topology Control Packets: Topology control packets are used to give information about the MPRs of that particular node. 3. MID packet: If a node has multiple interface, then these MID packets are used to send the information of those interfaces A HELLO message consists information of its neighbors and their link status. The addresses of all the neighbors to which there exists a valid bi-directional links. And addresses of the neighbor that received the HELLO messages. All onehop nodes receive the message. Every node in the network periodically sends the HELLO messages to its neighbors. As a node receives a HELLO message, it updates its MPR selector table. The question is how one can select the multi point relay. It is selected independently by each node on the network. The set of MPRs is selected in such a way that all the two hop neighbors are selected, and this information can be retrieved from the HELLO messages received by this node. Only the information of bi directional link nodes is taken. Whenever the node of the network is updated or a node fails, the multipoint relay is recalculated. The information of MPR should also be mentioned to other nodes. This can be sent with topology control packets. The TC messages 3 sent by the MPR includes its own sequence number to the list. The topology control messages that are sent throughout the network helps in building the topology of the network. The interval of TC messages may depend on the MPR selector set, that if a change occurs then sent early. A node records information about all the multipoint relays, using the information obtained from TC messages. The information of all the multipoint relays in a topology table. The table consists of 1. MPR selector set sequence number, 2. Destination address, 3. Last hop node address. This topology table is executed in the following way: The entry in the topology table is discarded if the MPR selected sequence number is greater than the sequence number than sequence number in the received message. Second, if there is an entry of last-hop address is smaller than the sequence number in the received, the entry is removed. By using the Third, for each MPR selector received in the TC message, a new topology entry is recorded in the topology table. Each node consists of a routing table, to send the packets for other destinations in the network. Routing has information of all the nodes, hence if any of the nodes are updated or declined then this routing table needs to be recalculated. The following procedure 1. All the entries from the routing table are deleted. 2. Starting with one hop neighbors as destination nodes, new entries are added into the table. 3. The topology table records that are not used while calculating the routing table are removed. 2.2 P-OLSR Since FANETs have no central infrastructure, they are prone to isolated attacks and node failures. UAV move rapidly with high speeds hence the topology varies very rapidly. Nodes must react to this rapid change and be able to update their routing tables automatically. It can be achieved by using a fast and reactive algorithm. OLSR fails to track fast topology changes of a FANET. In P-OLSR routing algorithm relative speeds are used to calculate the weighted ETX factor that is expected transmission count. Consider node A and node B , ETX is used as measurement for quality of the wireless network link. It is defined as ET X A,B = 1 ?f ?r where ? f is the probability that the data packet is successfully sent. ? r is probability that the acknowledgement packet is successfully sent. ETX of a particular route R is sum of ETX metrics of links composing the routes A,B. ET X R = P (A,B)?R ET X A,B ? f ,? r are the receiving ratios which are measured using a link probe packet that is hello packet. The frequency in which hello messages are broad-casted is called Hello interval. A node takes a while before noticing that wireless link quality have reduced, and before realizing about the broken or a poor link it sends the data packets on a broken link causing interruption 4 to the service. Relative speeds are used to learn how the link quality is going to change. By assuming that each node has knowledge about its neighboring nodes, ET X A,B = e v A,B l ? r f r r Where v˜ A,B l is relative speed between nodes A,B and BETA is a non negative value. If nodes moves towards each other relative speeds will be negative making ETX value less than 1 and vice versa if the nodes are moving away from each other.Hence the link between the nodes that are moving closing towards each other is preferred to that of the nodes moving away from each other. GPS information is collected by hello messages, each time ETX is calculated it will have its upgrade GPS information. Following is the formula to calculate instantaneous relative velocity between A,B v˜ A,B l = d A,B l ?d A,B l?1 tl ?tl?1 where tl and ti?1 are time of arrival for last and second last hello messages respectively. Implementation of P-OLSR is done through sharing hello messages through which each node will have to share GPS coordinates. Then each node uses its neighboring co-ordinates to calculate its relative speeds and share them through hello and TC messages. TC messages are topology control messages that are used to construct routing table. Figure1 is the structure of hello message in OLSR. It has the 1st block which is 8 bytes carries information about node. Where as in the figure2 which is format of modified P-OLSR hello message, have 16 bytes. It has Figure 2.1: structure of hello message in OLSR Figure 2.2: structure of hello message in OLSR 5 2 empty bytes that are used to communicate the relative speeds between the nodes. Similarly in TC message format the OLSR and POLSR differ by one byte which is used to communicate relative speeds between nodes.