A Reinforcement Learning Network based Novel Adaptive Routing Algorithm for Wireless Ad-Hoc Network

Scientific Essay 2015 5 Pages

Engineering - Communication Technology


A Reinforcement Learning Network Based Novel Adaptive Routing Algorithm for Wireless Ad-Hoc Network

Jagrut Solanki 1

1 Student of M.E

1 Department of Computer Science and Engineering

1 Narnarayan Shastri Institute of Technology, Jetalpur, Ahmedabad, Gujarat, India

Abstract— Mobile communication has enjoyed an incredible rise in quality throughout the last decade. Network dependability is most important concern in wireless Ad-hoc network. a serious challenge that lies in MANET (Mobile Ad-hoc network) is that the unlimited mobility and lots of frequent failure because of link breakage. Standard routing algorithms are insufficient for Ad-hoc networks. as a results of major drawback in MANET is limited power provide, dynamic networking. In MANET each node works as a router and autonomously performs mobile practicality. The link connectivity changes ceaselessly because of mobility to reflect this routing information additionally needs to get changed ceaselessly. AODV protocol is projected for this extraordinarily mobile network. In ancient AODV if any node fails in middle of transmission the method starts from the source node but in our propose scheme the transmission starts from the closest neighbor node therefore shows very important reduction in delay and improvement in packet delivery ratio are achieved. It also reduces the routing overhead by reducing the frequency of route discovery process.

Key words: MANET, Wireless, Routing Protocol, AODV, DSR, ZRP


A mobile ad hoc network (MANET) is a collection of mobile devices that are connected by wireless links while not the utilization of any fixed infrastructures or centralized access points. In MANET, every node acts not only as a host however also as a router to forward messages for different nodes that don't seem to be among the same direct wireless transmission range[1]. Every device in a very MANET is free to move independently in any direction, and can thus modification its links to different devices frequently. MANETs are much more vulnerable and are susceptible to numerous kinds of security attacks [2].

These nodes have routing capabilities which permit them to make multihop paths connecting node that aren't within radio range. The routing protocols are roughly divided into three categories: proactive (table driven routing protocols), reactive (on- demand routing protocols), and hybrid. The first goal of such an ad hoc network routing protocol is to supply correct and efficient route establishment between pair of nodes, in order that messages could also be delivered in time. In proactive, every node maintains a routing table, containing routing data on reaching each different node within the network. In reactive, once a node wishes to send packet to a selected destination, it initiates the route discovery process, in order to find the destination [3].

Proactive protocols are known as table driven protocols like DSDV and OLSR during which, every node maintains the routing information of all nodes in routing table. This type of protocol is appropriate for limited number of nodes. in contrast to proactive protocols, reactive protocols are known as on demand protocols like AODV, DSR [7] and establish the route to the destination whenever communication is required. Hybrid protocols like ZRP are the combination of proactive and reactive protocols.

Ad-Hoc On demand Distance Vector Routing protocol (AODV) is wide used for the route discovery within the MANET. The AODV routing protocol comes beneath the class of reactive routing protocol, which means that it discover the route after receiving the Route Request (RREQ) from the source node [4]. AODV doesn't permit keeping additional routing that isn't in use [5]. There are three AODV messages i.e. Route Request (RREQs), Route Replies (RREPs), and Route Errors (RERRs) [6].

illustration not visible in this excerpt

Fig. 1: RREQ and RREP messages in MANET using AODV

When the source node needs to make a new route to the destination, the requesting node broadcast an RREQ message within the network [8]. in the figure one the RREQ message is broadcasted from source node A to the destination node B. The RREQ message is shown by the black line from source node A to several directions. The source node A broadcasts the RREQ message within the neighbor nodes [9]. Once the neighbor nodes receive the RREQ message it creates a reverse route to the source node A. This neighbor node is that the next hop to the source node A. The hop count of the RREQ is incremented by one. The neighbor node can check if it's an active route to the destination or not. If it's a route, then it'll forward a RREP to the source node A. If it doesn't have an active route to the destination, then it'll broadcast the RREQ message within the network once more with an incremented hop count value. The figure one shows the procedure for locating the destination node B [9]. The RREQ message is flooded within the network in searching for finding the destination node B. The intermediate nodes can reply to the RREQ message only if they have the destination sequence number (DSN) adequate to or larger than the number contained within the packet header of RREQ [10].

illustration not visible in this excerpt

Fig. 2: Working of AODV protocol Neighboring nodes sporadically exchange hello message, Absence of hello message is used as a sign of link failure.


illustration not visible in this excerpt

Table 1: Comparison of Protocol


Reactive protocols establish the route once source has data send to destination. Examples for Reactive protocols are AODV and DSR routing protocol. During this paper, our work is to enhance performance of AODV routing protocol. During this section, we'll see the function of AODV routing protocol and literature survey of various routing protocols that optimizes the performance of AODV. AODV (Ad hoc On demand Distance Vector) Routing Protocol was proposed [8] and it's designed for nodes that are in mobile fashion to line up an ad hoc Network. As a reactive routing protocol, AODV protocol maintains the routing information of all destinations within the routing table

In AODV every node have latest information available about the sequence number for the IP address of destination node for which the route table entry is maintain, this sequence number call “destination sequence number“to enable Fresh route and to avoid loop free mechanism AODV use destination sequence numbers.AODV is hop by hop routing protocol that sends and receives data and maintain active route using Route Request (RREQ), Route Reply (RREP), Route Error (RERR) and HELLO message..

A node can change the sequence number in entry of routing table of destination when the node is destination node itself and offers new routes itself, or it receive the new information of sequence number of destination node or path of Destination node expire or break. A destination node can increments its own sequence number in two ways , when a node originate a route discovery or a destination node originate route reply in response to route request in this circumstances it must update its own sequence number to the maximum of current sequence number and destination sequence number in RREQ packet.


illustration not visible in this excerpt

Figure 3: Route Request Frame Format

In Route Discovery phase source node want to send data to destination node but if it not have route to destination it broadcast the RREQ with incrementing the broadcasting id and destination sequence number. Intermediate node receives RREQ and send RREP if it has active route available for destination otherwise it broadcast again RREQ. Route request packet travels entire in network, intermediate node receive the packet and add node address in routing table from which it has received for reversed route if it does not has route. Intermediate node or destination node can send a RREP. If intermediate node has route to destination then it generate RREP and send it to source node using reverse route with new sequence number in a unicast manner.


illustration not visible in this excerpt

Fig. 4: Route Reply Frame Format

After receiving a route reply, an intermediate node establish forward route to destination. If intermediate node has an active route to the destination, the destination sequence number within the node’s existing route table entry for the destination is valid and bigger than or equal to the Destination Sequence number of the RREQ. The intermediate node additionally updates its route table entry for the node originating the RREQ by inserting the next hop towards the destination for the reverse route entry. If the destination node sends RREP, it must increment its own sequence number by one if the sequence number within the RREQ packet is equal to that incremented value.



ISBN (eBook)
File size
598 KB
Catalog Number
reinforcement learning network novel adaptive routing algorithm wireless ad-hoc




Title: A Reinforcement Learning Network based Novel Adaptive Routing Algorithm for Wireless Ad-Hoc Network