A wireless Ad hoc network made of wireless nodes which communicate without the need for a centralized control. A set of autonomous nodes or terminals communicating with each other by means of a multi-hop wireless network which maintain connectivity in a decentralized way is called an Ad hoc network. There is no fixed infrastructure for such network, like server or a base station. These Ad hoc radio networks create, organize and administer their own temporary routes. Communication among Ad hoc networking nodes is called cooperative communication. This is because all these nodes cooperate among themselves in maintaining data traffic of the network and for this each node has its own router.
Thus, these networks exist merely in the form of interactions within their fellow smart mobile nodes. These interactions can be used in ingenious way to achieve different paradigms of wireless network performances useful for todays' applications. Ad hoc networks have very wide range of applications moreover there is lot of scope for their performance improvement.
On the contrary, conventional networks which are employed in local and wide area networks, has well defined structure and hierarchy. Also their topology is fixed or changes at very slow rate. Hence, the mechanisms for discovery and maintenance of route can possibly be very lethargic in their response times. Due to this manual reconfiguration and maintenance of the routing topology is usually done.
An Ad hoc network is certainly not compatible with this basic and conventional model, as it is highly time varying topology. Rather it makes users free from bondage of rigid topology of conventional fixed networks. For example, in case of today's' cellular mobile networking topology, mobile handsets are constrained to be present in the radiation coverage area of central tower of base station of the corresponding cell. Moreover the base station is connected to fixed network infrastructure of public telephone network. Thus, in future, cellular mobile networking based on well designed Ad hoc network philosophy will provide global connectivity. However the connectivity may only be restrained by data transfer capability of Ad hoc routes among nodes (link capacity) and routing delays of the Ad hoc networks thus formed. Although such networks are not implemented in the practice, their creation is certainly possible and within the realm of todays technology. Routing protocols has very important role to play in such networks.
1.2 General Issues in Ad hoc Routing
In conventional network infrastructure, IEEE 802.11 standard is widely adapted. Interfacing cards and simulation models based on this standard are widely available. Hence researchers of Ad hoc networks assume the IEEE 802.11 standard for underlying wireless technology [1]. This standard provides an Ad hoc mode, allowing mobiles to communicate directly. As the communication range is limited by regulations, a distributed routing protocol is required to allow long distance communications much beyond the range of today's limited area of cell in the cellular mobile networks. Moreover, communication is in Ad hoc networks the communication is not direct but occurs through multiple hops. Hence the IEEE 802.11 standard is not suitable for such networks and there is a lot of scope for invention and innovation.
Routing in Ad hoc network is technologically challenging. First challenge is to maintain sufficient throughput or data transfer rate. The throughput corresponds to effective share of bandwidth given by network to an application running on the network. There are two reasons behind variation of the network throughput or data transfer rate. The first reason is variable latency of the network. As already mentioned communication in Ad hoc networks being researched is in multi hops. Latency or delay goes on adding up per hop. Moreover, the topology of such networks changes, frequently and in unpredictable manner. Hence, the number of hops between source and destination node will also go on changing. Ultimately latency of such networks is not only high but also variable. Second reason for variation the throughput is the decrease in the signal-to-noise ratio (SNR) as the distance between two wireless nodes increases according to Shanon's information theory. Hence, link bit rate decreases. Therefore, as the signal fades away, the range of applications that can be run on the networks decreases. Moreover bit error rate increases. This alteration of throughput with respect to time is commonly termed as "fading".
Currently, very few protocols for MAC layer exist which can adjust their data transfer rate in response to fluctuations in link performance that occur due to variation of channel parameters. In IEEE 802.11 wireless Ethernet, link quality degradation results in decreased link bit rate. There is a need of genuinely robust protocol which dynamically associates the bit rate of the channel to achieve a required balance of throughput and bit error rate. In wireless networking there are lot of constraints on power and bandwidth. So far the greatest problem of present research in Ad hoc networking is routing in a scenario continuously changing network topology and that also nonuniformly distributed.
Conventional static networks generally use either Link State (LS) or Distance Vector (DV) algorithms for routing. Unfortunately, none of these algorithms are perfectly appropriate for frequently changing topology which is very distinctive feature of Ad hoc network's topology. Network nodes will often be battery powered, which limits the capacity of CPU, memory, and bandwidth [2]. This will require network functions that are resource effective. Furthermore, the wireless (radio) media will also affect the behavior of the network due to fluctuating link bandwidths resulting from relatively high error rates. These unique features pose several new challenges in the design of wireless Ad hoc networking protocols. Network functions such as routing, address allocation, authentication, and authorization must be designed to cope with a dynamic and volatile network topology. In a vigorously dynamic wireless network, protocols of this type encounter number of difficulties:
Network topologies may be highly inconsequential, where some nodes are surrounded by large number of nodes while some are in derelict condition i.e. have very few neighbors.
Bandwidth is very limited hence required to be used in judicious manner.
Battery power on portable devices is very scarce resource that needs to be used properly.
Lot of overhead due to frequent updates of highly dynamic topology.
Over and all it can be said that the challenge of Ad hoc network routing can be divided into to two parts i.e. discovery of routes and maintenance of routes. Moreover, there is a need of an elaborate maintenance of routing information.
1.3 Wireless Network Protocols
As of date, wireless communication is one of the most demanding areas of research within networking, with many proposed, but unverified protocols. The success of the proposed protocols depends on the availability of robust implementations that enable both real-time test beds and non-real time simulations.
The idea of such networking is to ensure proper and efficient execution of real time operations in mobile wireless networks by embedding comprehensive routing functionality into the Ad hoc network design. A user can move anytime in an Ad hoc scenario and, as a result, such a network needs to have routing protocols which can adopt dynamically changing topology. Hence, as mentioned previously, routing in wireless Ad hoc networks is challenging because of its highly dynamic topology.
To accomplish this, specially configured routing protocols are engaged for the networks where multiple hops are needed rather than single hop to establish communication routes. The peculiar feature of these protocols is their capability of tracing routes even in the presence of its dynamic topology. In past few years several routing protocols aimed at mobile Ad hoc networks are being invented. Few examples of such protocols are DSR, DSDV and AODV routing [2]. Out of which AODV is basic protocol & shows better performance over the other two on demand protocols.
1.3.1 Destination Sequenced Distance Vector (DSDV) Protocol:
DSDV is a multihop distance vector routing protocol. DSDV is a Proactive routing protocol. This implies that each network node maintains a routing table that contains the next-hop for and number of hops to all reachable destinations. Periodical broadcasts of routing updates attempt to keep the routing table completely updated at all times. To guarantee loop-freedom, DSDV uses a concept of sequence numbers to indicate the freshness of a route. The sequence number for a route is initiated by the destination node and increased by one for every new originating route advertisement. When a node along a path detects a broken route to a destination D, it advertises its route to D with an illimitable hop-count and a sequence number is added by one. Route loops can occur when incorrect routing information is present in the network after a change in the network topology, e.g., a broken link. DSDV uses triggered route updates when the topology changes. The transmission of updates is delayed to introduce a damping effect when the topology is changing rapidly.
1.3.2 Dynamic Source Routing ( DSR) Protocol:
Dynamic Source Routing is a reactive routing protocol which uses source routing to deliver data packets. Headers of data packets carry the sequence of nodes within which the packet must travel. This means that intermediate nodes only need to keep track of their immediate neighbors in order to forward data packets. The source, on the other hand, needs to know the complete hop sequence to the destination. The route acquisition procedure in DSR requests a route by flooding a Route Request packet. A node receiving a Route Request packet searches its route cache, where all its known routes are stored, for a route to the requested destination. If no route is found, it forwards the Route Request packet further on after having added its own address to the hop sequence stored in the Route Request packet.
The Route Request packet travels through the network till it reaches either its proposed destination or to a node on the route to the destination. If a route is found, a Route Reply packet containing the proper hop sequence for reaching the destination is unicasted back to the source node. DSR does not rely on bi-directional links since the Route Reply packet is sent to the source node either according to a route already stored in the route cache of the replying node, or by being piggybacked on a Route Request packet for the source node. However, bi-directional links are assumed throughout this study. Then the reverse path in the Route Request packet can be used by the Route Reply message. The DSR protocol has the advantage of being able to learn routes from the source routes in received packets. To avoid unnecessarily flooding the network with Route Request messages, the route acquisition procedure first queries the neighboring nodes to see if a route is available in the immediate neighborhood. This is done by sending a first Route Request message with the hop limit set to zero, thus it will not be forwarded by the neighbors. If no response is obtained by this initial request, a new Route Request message is flooded over the entire network.
1.3.3 On Demand Distance Vector (AODV) Protocol:
AODV is the protocol proposed by Perkins. It combines features of the DSR and DSDV protocols, which uses DSR's responsive route discovery & maintenance models, along with destination sequence number, update at regular time characteristics of DSDV protocol.
AODV is a reactive routing protocol. That is, AODV requests a route only when needed and doesn't need nodes to keep routes to destinations that are not transferring data at that time. The process of finding routes is referred to as the route acquisition. AODV uses sequence numbers in a way similar to DSDV to avoid routing loops and to indicate the freshness of a route. Whenever a node needs to find a route to other node it transmits a Route Request (RREQ) message to its entire neighbor. The RREQ message is flooded through the network until it reaches the destination or a node with a fresh route to the destination. On its way through the network, the RREQ message initiates creation of temporary route table entries for the reverse route in the nodes it passes. If the destination, or a route to it, is found, the route is made available by sending back a Route Reply (RREP) message to the source along the time being reverse path of the received RREQ message. On its way back to the source, the RREP message initiates creation of routing table entries for the destination in intermediate nodes. Routing table entries expire after a certain time-out period. Neighbors are detected by periodic HELLO messages (a special RREP message). If a node X does not receive HELLO messages from a neighbor Y through which it sends traffic, that link is deemed broken and a link failure indication (a triggered RREP message) is sent to its active neighbors. The latter refers to the neighbors of X that were using the broken link between X and Y. When the link failure messages eventually reach the affected sources, these can choose to either stop sending data or to request a new route by sending out new RREQ messages.
1.3.4 Performance Comparison of the Protocols
The performance comparison of these protocols considering all the characteristics that need to be available with routing protocols is the fundamental step towards the invention of new routing protocol. For performance comparison, I have referred many papers dealing with detailed comparative analysis of routing protocols.
The performance comparison done by [2] using ns-2 simulator providing all protocols with identical traffic load and mobility patterns. Special mobility metric was used in this simulation which was proposed to acquire and figure out the kind of node motion relevant for an Ad hoc routing protocol. It was observed that AODV performs well among its fellow protocols for a particular scenario where mobility is high, nodes are densely populated, area is wide, the extent of traffic is high and network is for elongated period. Comparative study of protocols with multiple mobility models [3] also indicate that AODV gives excellent performance in presence of high or low mobility, high or low traffic, hence superior as compared to other protocols.
1.4 QoS Issues in Ad hoc Networks
QoS is a term widely used in the last recent years in the area of wireless networks. QoS stands for Quality of Service. Quality of service (QoS) is usually defined as a set of service requirements that needs to be fulfilled by the network while delivering a stream of packets from a source to its destination.
Ad hoc networks are popularly called as MANets (Mobile Ad hoc Networks). In the early stages of MANet development, QoS provision did not attract much attention and thereby most routing protocols operated on a best effort model. However, with the growing popularity of time-sensitive applications, QoS support becomes much more important than it was, leading to a shift of research interest from best effort routing to QoS provision routing. QoS routing protocols try to find for routes with enough resources for satisfying the QoS requirements of data flow. The QoS routing protocol should find the path that consume minimum resources. However, providing QoS guarantees in MANETs is quite challenging due to the dynamic topology, limited bandwidth and energy constraint.
Support for QoS is integral to the design of Ad hoc networks. Frequent fluctuations in channel quality affect the QoS parameters on each established link or the over span of all nodes participating in the route. Moreover, the interference from transmission of distant nodes affects the link quality. Hence, QoS is an essential component of Ad hoc networks. This thesis presents solutions and approaches for supporting QoS in Ad hoc networks and describes future challenges that need to be addressed to design a QoS enabled Ad hoc network.
The QoS requirements arise at the application layer in the form of restrictions on values of certain QoS metrics. The most commonly studied QoS parameters are available bandwidth, delay in packet deivery and jitter. I have chosen Bandwidth as is the QoS metric, it is the most important metric that need to be researched most for contemporary and futuristic Ad hoc networking applications.
This thesis focuses on enhancement in performance of normal AODV protocol by improving the QoS. The various QoS parameters can be stated as bandwidth, cost, end-to-end delay, delay variation (jitter), throughput, probability of packet loss, battery charge, processing power etc. Research is going on towards Performance Improvement by emphasizing any of these parameters. This thesis considers the Bandwidth parameters Hence as to improve QoS. Various Performance metrics are to be studied for Performance evaluation of QoS-enabled AODV protocol.
1.5 Problem Statement
The aim of this research is to improve and evaluate performance of the normal AODV protocol by means of QOS-enabled routing strategies. Depending on the application involved, the QoS constraints could be available bandwidth, cost, end-to-end delay, delay variation (jitter), throughput, probability of packet loss, battery charge, processing power, buffer space and so on.
This thesis presents a QoS-Enabled AODV Routing Protocol & its comparison with normal AODV protocol. In a distributed Ad hoc network, an available bandwidth is not only decided by the raw channel bandwidth of a node, but also by its neighbor's bandwidth usage and interference caused by other sources, each of which reduces that nodes available bandwidth for transmitting data. Therefore, applications dependent on such network for their communication cannot properly optimize their coding rate without knowledge of the status of the entire network. Thus, bandwidth estimation is a fundamental function that is needed to provide QoS. However, bandwidth estimation is extremely difficult, because each host has imprecise knowledge of the network status and communication links among nodes change dynamically. Therefore, an effective bandwidth estimation scheme is highly desirable to achieve QoS. The scheme needs to be evaluated in terms of comparison with its predecessor.
This thesis is organized as follows. This Chapter gives introductory information about issues in research of Ad hoc network protocols. Second chapter give comprehensive review of past work. Third chapter deals with routing protocols of MANets. Fourth chapter itemizes some characteristics of AODV protocol. Fifth chapter gives definition of QoS and its relation with network protocols. Sixth chapter gives comprehensive account of ns2 which is used for performance evaluation of AODV protocol with quality of service. Seventh chapter outlines research methodology and performance evaluation techniques used in this research, especially modifications made in AODV protocol for getting better QoS (Quality of Service). Chapter on results and discussions i.e. chapter no. 8 give account of research outcome with conclusion and future scope.
