Quality Of Service In Manet Computer Science Essay

Published Date: 02 Nov 2017

CHAPTER 2

RELATED WORK

In the field of wireless technology, Mobile ad hoc network acts as a more prominent and great appeal of new innovation. Due to the absence of fixed infrastructure, this kind of networks operates, and facilitates ease construction of infrastructure at anytime and anywhere. Inadequate infrastructure in mobile ad hoc networks creates different challenges in the field of research and offer complexity to utilize the classical techniques for network services. Formal challenges include security problem, power constraints, routing mechanism and bandwidth constraints. The proposed framework in this thesis provides a better understanding of QoS (Quality of Service) issues in MANET. This chapter analyzes the core issues and basic problem of ad hoc networking by reviewing the related research background, which encompasses the application, characteristics, status, concept and difficulties of MANET.

2.1 BACKGROUND OF MANET

Autonomous system of mobile hosts without any centralized control or infrastructure is formed by Mobile ad hoc networks. The key benefit of deploying this network in an environment affords mobility support, robustness, high flexibility, effective deployment. It establishes direct communication between nodes or employs routers to accomplish communication through intermediate nodes. This kind of network plays a significant role in civilian forums like group conferences, digital and electronic class rooms. Early applications in ad hoc networking can be sketched to the DARPA (Defense Advanced Research Projects Agency). Due to the advancement in microelectronics technology and with the progress in time, has made it possible to integrate nodes and network devices into a single unit called Ad hoc node. Ad hoc Network Active research work on ad hoc networks started in 1995 in a conference session of Internet Engineering Task Force (IETF). However, the effective deployment of ad hoc network introduces many challenges and difficulties, which is distinct from classical wired and wireless technology. Samba S, et al. (2004) analyzed the issues generated by these challenges and a detailed overview of ad hoc networking have been discussed.

Maged Salah E.S, et al. (2007) proposed a novel adaptive routing protocol for MANET termed as AODVLRT (AODV with local Repair Trials), which is the modified version of AODV. This protocol customizes the local repair algorithm utilized for maintaining route of the AODV routing protocol. Original AODV local repair algorithm is generated as a result by considerably reduced the overhead in message routing. Enhancement in this adaptive routing protocol offers low latency and greatest throughput than AODV. Local repair in AODV is determined by only one trial to scrutinize the route repair in terms of broadcasting RREQ packet along with TTL, whereas route repair tracing in AODVLRT is determined by more than one trial. In AODVLRT, result of route failure increments the destination sequence number of upstream node by one and then initialization of first local pair trial is effectuated by broadcasting RREQ packet integrated with TTL = LR_TTL_START. To maximize the probability of examining a repair from the first local trial LR_TTL_START has been exploited by which low value of TTL will considerably reduce the overhead in message routing. During route discovery, there may be a possibility of upstream node again fails to receive RREP packet. In such situation, all the aforementioned process is iterated again until RREP packet is received. Similarly, when the value of TTL exceeds the fixed threshold LR_TTL_THRESHOLD, then final trial by RREQ packet broadcasting is created by upstream node along with TTL = LR_TTL_MAX.

Mobile networking has gained significant attention in collaboration with pervasive computing. Jun-Zhao S, et al. (2001) reviewed this pervasive computing technology. From the past few years, technological enhancement in both software and hardware offers wireless networking and mobile hosts as miscellaneous and common. Wireless mobile units can be communicated with each other by enabling either infrastructure or infrastructure less environment. Ad hoc networking is the infrastructure less network, which possesses widespread application with progressive improvement in wireless communication along with widespread use of portable devices. It achieve scalability in network communication by easily append and discard the mobile devices among the network. Military battlefield, Commercial sector, PAN (Personal Area Network) is the typical MANET applications. MANET permits the mobile host communicated through wireless links and offers the mobile host to move randomly. Compare to the classical infrastructure wireless network, MANET expose unique type of traffic, which encompasses Peer-to-Peer, Dynamic Traffic, and Remote-to-Remote.

2.2 QUALITY OF SERVICE IN MANET

A survey presented by Lei C, et al. (2007) exclusively studies the challenges and issues in QoS aware routing and also analyzed the classical QoS aware routing protocol. MANET plays a unique role than any other kinds of networks by comparing the physical characteristics, dynamic topology and organizational format. According to this survey, traditional MANET routing protocols lacks the network resources utilization and optimization by only examining possible route from source to destination. Many routing protocols, which supports QoS have been introduced by exploring the route discovery with maximum bandwidth, submit feedback report to the application regarding the estimation of route delay and affording call admission feature to discard the route requests when avails inadequate BW to accept the route request. This survey performed comparative analysis of all those routing protocols in terms of accomplishing QoS like routing overhead, network scalability, and mobility of nodes.

According to Buruhanudeen S (2007), reliability and efficiency of the protocol is determined by acquiring the proper routing metrics to be associated in a protocol. Successful implementation of MANET protocols rely on routing metrics. The author of this paper reviewed the most important routing metrics essential for the performance evaluation of the protocols in terms of efficiency and reliability. Routing functionality is the most important characteristics in MANET, which facilitates robust and efficient operation. Routing protocols in mobile ad hoc can be enumerated into table driven (proactive), on demand driven (reactive) and hybrid protocols. Commonly used table driven protocol includes Destination Sequenced Distance Vector (DSDV), Wireless Routing Protocol (WRP) and Cluster head Gateway Switching Routing Protocol (CGSR). Network routing information is continuously searched by the nodes of this kind of routing protocols. Reactive protocols establish route discovery within the network when entreat by source node. Signal Stability Routing (SSR), Dynamic Source Routing (DSR), Temporary Ordered Routing Algorithm (TORA), Ad hoc On Demand Distance Vector (AODV), Relative Distance Micro Diversity Routing (RDMAR), Associativity Based Routing Protocol (ABR) are the most important On Demand Driven Reactive Protocols. A hybrid protocol incorporates the key benefits of on demand and proactive protocols. Zone routing protocol comes under hybrid routing category in mobile ad hoc networks. Shortest path, density of node, load relay is the typical routing metrics. Minimum number of hops between the source and destination node is utilized for shortest path metrics. Success rate of repairing local route in mobile ad hoc networks is enhanced by node density metrics. Estimating the routing neighborhood node density and the node availability is key measure of node density.

A scenario described by Novatnack J, et al. (2005) illustrates the effect of routing protocols in MANET by achieving QoS. Three MANET routing protocols are evaluated in this study with an emphasis on how their routing behavior degrades the QoS. Some applications like multimedia and military require meeting the need of explicit performance metrics. A military application encompasses complex battlefield information, which needs to be guaranteed to deliver at the destination in an appropriate time. Multimedia applications are analyzed based on timing requirements that are essential to afford seamless audio/video streaming. The construction of effective QoS models that meet the need for aforementioned applications requires progress to classical QoS models. All the existing routing protocol holds distinguishing characteristics to accomplish metrics like Control Packet Overhead, packet delivery, end-to-end latency, average hop count. Above metrics directly impacts the requirement of QoS through guaranteed bandwidth/delivery/delay or latency. Moreover, aforementioned QoS metrics is impacted by the three aspects of proactive and reactive protocols. Each reactive and proactive protocol employs the mechanism for path selection (first aspect). Second aspect that impact the QoS metrics is the broken links detection and resolution, which is in accordance with the link layer. Final aspect focused on the effect of buffering in link breakage, which considerably affects the QoS based on transmission latency.

Typically, QoS routing in such networks is limited by the node or network failure due to either depleted energy of mobile nodes or mobility of nodes. A node disjoint multipath protocol was presented by Charu Gandhi, et al. (2009) to surmount the challenges in traffic distribution optimization and node or path breakage. Metrices like node stability and appropriate links is established to choose the route path. Routing strategy must attain the following objective to meet the QoS. Based on the policy constraints like selection of path, path cost, etc. feasible paths are determined dynamically. Another thing is to enhance the throughput of network and graceful degradation of performance during packet overhead to accomplish optimal resource utilization. Distributed routing, source routing and hierarchical routing are the classification of routing strategy on QoS. Due to the dynamic nature of MANET, it is very complex to maintain the precise link state information. This paper proposed solution to cater all the problems by defining a metrics that makes an attempt to stablize the balance between energy constraints and mobility in MANETs. All the routes that satisfy the stability and energy constraints are selected completely by exploiting this technique, thus attains fundamental objectives in QoS. The technical challenges that focus on providing desired quality of service to the users of wireless devices is also analyzed by Carla-Fabiana C, et al. (2007).

Among many application environment, MANET plays a wide role because of its independency in infrastructure support. MANET can be used to establish communication in smaller areas and constructs collaborative computing. Establishing communication in battlefields, disaster recovery areas, communication using sensor networks or float over water are some examples of application environments of MANET. Furthermore, a wide application of wireless devices extends the necessity and usage of MANET. Quality of Service (QoS) support in MANET requires the quantifying metrics along with complete knowledge regarding the issues and difficulties in provisioning QoS. The availabilty of resources, resource shortage may further append the demand for QoS. For instance, the features of these network increases the complexity of QoS support in networks. There are some difficulties and issues exists in network infrastructure at transport layer, application layer, medium access control layer, network layer, physical layer is encompassed by QoS support in MANET. The article described by Jian L, et al. (2003) provides a detailed survey of the issues involved in supporting QoS across all the protocol layers in MANETs. Classical wired infrastructure differs from emerging mobile multihop wireless networks comprises dissimilarity and distinct issues with respect to QoS in MANET. Node mobility, unlimited link properties, minimum battery life are the issues whereas route maintenance, hidden and exposed terminal problem and security are graded as consequences.

The service offered to the user by the network based on the performance level is termed as QoS. QoS ensures proper packet delivery and effective utilization of network resources to accomplish the ultimate goal of QoS. Possibly, any kind of services can be provided by service provider or network. Provisioning QoS support in MANET often requires call admission control, arbitration between network and host, priority packet scheduling and resource reservation. The quantifying parameters in provisioning QoS support vary from application to application. Consider, delay jitter, bandwidth and delay are the essential QoS parameters in multimedia applications. On the other hand, stringent security is required for military applications. Some other applications like handling rescue operations and emergency service, the key component is the network availability. Applications like group communication in a conference hall consumes less energy when require transmission among nodes. Research area gains attention by enabling QoS support in AWN (Ad hoc Wireless Network). Subsequently, some unique features in wireless network pose complexity in provisioning QoS. Elaborated discussion regarding the characteristics of Ad hoc wireless which consequently degrade the QoS was described by Bheemarjuna R, et al. (2006). a) Network topology dynamic variation: Lack of mobility constraints in ad hoc wireless network enables dynamic changes of network topology. Through consecutive path breaks, a QoS session needs to reestablish the new paths to accomplish QoS support. This reestablishment in construction of new path may incur delay, which leads some packets belonging to that session to miss their delay deadlines/target. b) Hidden Terminal Problem: This problem commonly occurs in AWN when two or more sender nodes initiate transmission with the common receiver node, thus collides each other while transmission of packets. This in turn necessitates packet retransmission, which is not acceptable as per stringent QoS requirements. c) Unassured medium: Broadcast nature of wireless medium ensures insecure communication through wireless channel. It reveals that the security is an important issue in AWN, particularly for tactical and military applications. Insufficiency in security mechanism does not guarantee secure communication. This article focussed a way for QoS provisioning by means of Hard state vs soft state reservation of resources, Soft QoS vs Hard QoS, Stateful vs Stateless approach.

AODV (MNRB-AODV), a novel approach developed by Gabriel I, et al. (2008), which supports QoS. The proposed approach dynamically distributes the traffic and acts as a backbone for mobile routing over AODV. Dynamic distribution of traffic among the network selects the best route that supports QoS between source and destination. For instance, a mobile ad hoc network comprises the characteristics of heterogeneity. According to those characteristics, nodes in the MANET are classified by the presented solution that establishes simple routing nodes in network to forward the packets without rely on special service provisions and transceiver nodes. Examining the routing capability of all those nodes, mobile routing backbone is then created. These types of routing nodes have forwarding capabilities to avoid starvation of best-effort packets. Maximum utilization of available bandwidth by dynamically distribute the traffic within the network is the key benefit of QMRB-AODV. Moreover, this approach minimize the message overhead when establish communication between source and destination nodes.

The progressive need for MANET focused on the applications of real time multimedia applications. Many of the current routing protocols in MANET support only best effort routing. To enhance the QoS on MANET, application like multimedia services that depends on MANET needs to be improved. One such extended version is proposed by Ferreira M, et al. (2006). The routing protocol AODV is extended and implemented. The quantifying effect of this enhancement is evaluated in terms of accomplishing QoS in a test grid with wireless infrastructure. The evaluated performance is then compared with the DYMO routing protocol and simulate the changes in routing metrics. End to end delay is measured, and changes occur from hop count to estimated end to end delay in routing metrics. Through OPNET simulation, the effectiveness of transforming the routing metrics to estimated end to end delay from hop count is measured. Simulation in OPNET environment acquires voice and video services (delay sensitive applications) for route selection. This extended work simply enhances the QoS level in MANET.

2.3 PROVISION FOR IEEE 802.11 MAC PROTOCOL

The problem of node mobility, disruption in service, node or link failure, low power, limited network bandwidth and security problem degrade the QoS accomplishment and possesses delay in allocation of multimedia resources. In a network, the capability of affording required resources without any critical issues and challenges in represented by the term QoS. The most significant concern in WLAN is delivering QoS based on time critical applications. Numerous techniques are devoted in literature to achieve QoS in WLAN. MAC layer service categories with tunable parameter have been designed by Puthal (2008) for better QoS in WLAN. A cross layer design approach can be employed to achieve better throughput. According to this paper, employing only DCF (Distributed Coordination Function) encompassed by IEEE 802.11 is not capable of achieving or satisfying the necessity of voice and video applications. To enhance the MAC DCF functionality, novel QoS schemes was proposed in this paper. The Quality of service can be established by modifying MAC DCF. The way in which QoS is provided is summarized in terms of three aspects. First thing is to offer prioritization technique by which classes with distinct traffic are prioritized. Many of the available approaches deploy different Inter Frame Spacing (IFS) or different Contention Windows (CW) or sometimes both. Second aspect is based on resource allocation by which data are prioritized. To accomplish these criteria Weighted Fair Queuing (WFQ) is employed. On the basis of model admission control and measurement, QoS is accumulated in case of third aspect.

Another extended work of MAC 802.11 has been designed by Guillermo A.P, et al (2007) to enhance the MANET performance. In wireless ad hoc networks, variety of protocols probably handles the problems like noise of the network, range of transmission, routing information, etc. Also a possibility of misinterpretation among the protocols happen due to only one slice of data gathered by one protocol is received by some other protocol. To address this issue, MAC 802.11 protocol is transformed, which then resolve the unwanted launching operations in the DSR (Dynamic Source Routing). This design modification enhance the network performance by providing less routing overhead, low packet collision, less change in routing, less route error rate, increased throughput and less MAC error. The strength of the node signal is tracked by this extended approach, which inform the amount of signal strength to the routing layer and skipping the DSR launched routing error.

Sunil Kumar, et al. (2006) presented a broad overview of the research work conducted in the field of ad hoc wireless networks with respect to MAC protocols. MAC layer which is also referred to as a sublayer of the Data Link layer, involves the functions and procedures necessary to transfer data between two or more nodes of the network. This layer is responsible for resolving conflicts among different nodes for channel access. The design of a MAC protocol should address issues caused by mobility of nodes and an unreliable time varying channel. Various MAC schemes developed for wireless ad hoc networks can be classified as contention free schemes and contention based schemes. TDMA, CDMA, FDMA come under contention free schemes by which certain assignments should be used to avoid contention whereas contention based schemes are aware of the risk of collisions of transmitted data. The survey fully focused on contention based MAC schemes since contention free MAC schemes are more applicable to static networks or the network with centralized control. CSMA

in contention based schemes reduce the possibility of packet collisions and improve the throughput. One distinguishing factor for MAC protocols is whether they rely on the sender initiating the data transfer or the receiver requesting the same. Another classification is based on the number of channels used for data transmission. Some class of MAC protocols uses directional antennas. MAC protocols for ad hoc networks typically assume the use of omni-directional antennas, which then transmit radio signals to and receives them from all directions. The IEEE 802.11 specifies two modes of MAC protocol as distributed coordination function (DCF) mode (for ad hoc networks) and point coordination function (PCF) mode (for centrally coordinated infrastructure-based networks). Moreover IEEE 802.11 DCF is designed to provide a channel access with equal probabilities to all the contending nodes in a distributed manner. Survey result shows that Most of the existing MAC schemes focus on only a subset of QoS features with a simple network topology, while ignoring the issues of end to end packet delay in multi-hop networks, node mobility, power control, channel errors.

A novel MAC layer for Wireless LAN proposed by Matteo Cesana, et al. (2003) which is able to extend the capabilities of the basic IEEE 802.11 DCF (Distributed Coordination Function) to environments with high interference, both in ad hoc and in infra structured mode. Even if 802.11 is the official standard for Wireless LAN???s both in infrastructure and in ad hoc mode, originally it was devised only for a single Access Point scenario where all the nodes were within the transmitting range of one another. For this reason, many problems arise when use 802.11 both in a pure ad hoc mode and in a infra structured cellular-like scenario which have different APs and mobile nodes moving around. The basic idea is to propose minimal modifications to the IEEE 802.11 MAC DCF in order to make feasible multiple parallel communications in an interference scenario. IA-MAC (Interference Aware in MAC) can be useful both in ad hoc mode and in infrastructure one. IA-MAC permits parallel 802.11 communications and solves partially the problem of the exposed terminal in the ad hoc mode.

Bianchi model was proposed by Bianchi G, et al. (2000) for the performance analysis of IEEE 802.11 Distributed coordination function. It is the analytical calculation of saturation throughput in a closed-form expression. The model also calculates the probability of a packet transmission failure due to collision. It assumes that the channel is in ideal conditions, i.e., there is no hidden terminal and capture effect. IEEE 802.11 is the de facto standard for WLANs. It specifies both the medium access control and the physical layers for WLANs. The scope of IEEE 802.11 working groups (WGs) is to propose and develop MAC and PHY layer specifications for WLAN to handle mobile and portable stations. The distributed coordination function (DCF) is the basic medium access mechanism of IEEE 802.11, and uses a carrier sense multiple access with collision avoidance (CSMA/CA) algorithm to mediate the access to the shared medium. The standard also describes centralized, polling-based access mechanism, the point coordination function (PCF) which is very rarely used in practice. The DCF protocol in IEEE 802.11 standard defines how the medium is shared among stations. It includes a basic access method and an optional channel access method with request-to-send (RTS) and clear-to-send (CTS).

2.4 CHALLENGES IN MANET

The major challenges faced by the internet architecture in incorporating emerging wireless network elements such as mobile terminals, ad-hoc routers and embedded sensors and to provide end-to-end service abstractions that facilitate application development in imposed by Kavita T, et al.. The major challenges faced by the internet architecture can be broadly classified as:

a) Incorporating emerging wireless network elements such as MDs, ad-hoc routers and embedded sensors in the existing protocol framework.

b) To provide end-to-end service abstractions that facilitates application development.

These challenges are posed by a broad range of environments such as cellular data services, WiFi hot-spots, Info stations, mobile peer-to-peer, Adhoc mesh networks for broadband access, vehicular networks, sensor networks and pervasive systems. These wireless application scenarios lead to a diverse set of service requirements for the future Internet as summarized as:

1. Naming and addressing flexibility.

2. Mobility support for dynamic migration of end-users and network devices.

3. Location services that provide information on geographic position.

4. Self-organization and discovery for distributed control of network topology.

5. Security and privacy considerations for mobile nodes and open wireless channels.

6. Decentralized management for remote monitoring and control.

7. Cross-layer support for optimization of protocol performance.

8. Sensor network features such as aggregation, content routing and in-network Processing.

9. Cognitive radio support for networks with physical layer adaptation.

10. Economic incentives to encourage efficient sharing of resources.

A deep insight into three mechanisms to differentiate among traffic categories, i.e., differentiating the minimum contention window size, the Inter-Frame Spacing (IFS) and the length of the packet payload according to the priority of different traffic categories have been analyzed by Bo Li, et al. (2003). For that purpose an analysis model is proposed to compute the throughput and packet transmission delays. In order to gain a deeper insight into the modified IEEE 802.11 MAC with service differentiation support, system modeling and performance analysis are needed. The basic DCF method is not appropriate for handling multimedia traffic requiring guarantees about throughput and delay. For the quality of service of real-time multimedia it is important to know the time that a packet must wait for transmission over the IEEE 802.11 MAC. Because of this weakness, task group E of the IEEE 802.11 working group is currently working on an enhanced version of the standard called IEEE 802.11e. The goal of the extension is to provide a distributed access mechanism capable of service differentiation. By adopting the scheme of differentiating minimum contention window sizes and packet payloads of different traffic types, simple relationships exist among throughput and packet delays, which is helpful to simplify the design of the whole system. On the other hand, in order to make the system as simple as possible, one should limit the number of parameters that can be adjusted. On the whole, by using the analysis model proposed in this paper, one can obtain deeper insight, which is important and helpful to the design of real systems.

The IEEE 802.11 technology is a good platform to implement single-hop ad hoc networks because of its extreme simplicity. Single-hop means that stations must be within the same transmission radium (say 100-200 meters) to be able to communicate. This limitation can be overcome by multi-hop ad hoc networking. This requires the addition of routing mechanisms at stations so that they can forward packets towards the intended destination, thus extending the range of the ad hoc network beyond the transmission radium of the source station. Routing solutions designed for wired networks (e.g., the Internet) are not suitable for the ad hoc environment, primarily due to the dynamic topology of ad hoc networks. Even though large-scale multi-hop ad hoc networks will not be available in the near future, on smaller scales, mobile ad hoc networks are starting to appear thus extending the range of the IEEE 802.11 technology over multiple radio hops. Most of the existing IEEE 802.11-based ad hoc networks have been developed in the academic environment. The characteristics of the wireless medium and the dynamic nature of ad hoc networks make (IEEE 802.11) multi-hop networks fundamentally different from wired networks. Furthermore, the behavior of an ad hoc network that relies upon a carrier-sensing random access protocol, such as the IEEE 802.11, is further complicated by the presence of hidden stations, exposed stations. The interactions between all these phenomena make the behavior of IEEE 802.11 ad hoc networks very complex to predict. Extensive literature related to the performance analysis of the 802.11 MAC protocol in the ad hoc environment have been imposed by Giuseppe A, et al. (2003). The IEEE 802.11b extended version, also known as Wi-Fi, is the reference technology for ad hoc networking. IEEE 802.11 performance with an extensive set of measurements based on IEEE 802.11b products was conducted on a real testbed. The measurements were done in an outdoor environment, by considering different traffic types (i.e., TCP and UDP traffics). Experimental results indicate that transmission ranges are much shorter than previous version.

An analytical model to evaluate the IEEE 802.11 Distributed Coordination Function (DCF) saturation throughput performance was proposed by Bianchi G, et al. (2005). This approach relies on elementary conditional probability arguments rather than bidimensional Markov chains (as proposed in previous models), and can be easily extended to account for backoff operation more general than DCF???s one.

Wireless Local Area Networks (WLANs) has gained significant attention and revolutionized all over the world, because of its flexibility and does not require any cables for the implementation. IEEE 802.11 is the dominating protocol promoted by WLAN, which exemplifies specification of Medium Access Control (MAC) and physical layer (PHY). Based on the working principle of carrier sense multiple accesses with collision avoidance (CSMA/CA) scheme, IEEE 802.11 MAC protocol includes two modes like Distributed Coordination Function as mandatory and Point Coordination Function as optional for accessing channel. DCF lacks the differentiation mechanism between delay sensitive/insensitive real time traffic. Voice over IP and video come under delay sensitive real time traffic whereas FTP and e-main comes under delay insensitive traffic. On the other hand, PCF works on the fundamental principle of polling based contention free access scheme. This scheme employs access point as centralized control of point coordinator, which is mainly designed to reinforce applications like multimedia. Two techniques like two way handshaking (DATA-ACK), which provides basic access, and four way handshaking (RTS-CTS-DATA-ACK) that accommodates RTS/CTS data access is described by DCF for packet transmission. When the data transmission is initiated by basic access scheme, sender transmits the packet then the receiver responds the packet request and acknowledges the receiver by sending an ACK. In case of RTS/CTS scheme, short frame between transmitter and receiver is exchanged to reserve the channel and make the medium available for transferring short data packet.

The DCF scheme implemented by wireless station obtains the request for data transmission from upper layer. Based on the busy or idle status of the channel medium, station proceeds the transmission and defer of data packets. This strategy avoids the probability of collision. A random backoff timer is generated by the station when the channel is busy. The transmission is deferred until idle DIFS is encountered, the timer value minimized in terms of slot time. If the counter reaches zero, then the station successfully transmits the packet. When the packet is delivered successfully at the destination node, it immediately waits for response of ACK and SIFS.

An analysis for probability of delay distribution in back off timer under saturation conditions is presented by Raptis P, et al. (2009). It is understandable that backoff timer elapse the time once after the successful transmission of packets. Author of this paper analyze the back off delay distribution and reported that all the stations should have packet for data transmission in WLAN under saturation conditions. The distribution of end to end delay algorithm is presented in a mixed scenario applicable to voice and data traffic. Another algorithm called admission control guarantees end to end delay and affords voice traffic. With the given probability distribution, algorithm for voice packets accomplishes end to end delay when the threshold value reaches low than the certain thresholds. Employing these algorithms assures low delay in most of the packet transmission.

IEEE 802.11 access method of DCF works on the basis of CSMA or CA protocol. Any of the station wants to establish the packet transmission, it first senses the channel. When the sensed channel is free then transmit the packet for a DIFS time otherwise station waits until the medium becomes idle. The channel starts the backoff process in terms of initializing the back off timer by producing a random value acquired from the uniform probability distribution in the parameter range of (0, CW-1). The obtained CW (Contention Window) value depends on the number of failed transmissions for the packet. The value of CWmin is used to set the CW value when attempting initial transmission and the value is doubled to the greatest of CWmax once after the transmission is unsuccessful. The values of CWmin and CWmax are fixed by the standard. When the proper delivery of packet to the destination is attained, ACK frame is send from the receiving station. Assumption of collision occurrence is made if the ACK frame is not delivered by the receiving station within an ACK timeout time. Once the collision is encountered, rescheduling of packet transmission is desired based on acquired backoff rules. Packet is discarded when the transmission count reaches beyond the limit of retry.

In IEEE 802.11, the mechanism of Request to Send or Clear to send is optional. The utilization of this option under the low range of back off timer, receiving station receives the RTS frame from the transmitting station. The receiving station responds the transmitting station by the CTS frame along with ACK frame at SIFS time.

2.5 RED CONGESTION CONTROL

For the accommodation of transient congestion in high speed network, gateways and large delay BW products are designed to manipulate greatest queues. In the current network scenario, TCP protocol encounters congestion when the packet is discarded at the gateway. To accomplish congestion avoidance an effective mechanism is employed at the gateway called RED (Random Early Detection), which supports transport protocols. One such algorithm was proposed by Sally Floyd, et al. (1993) that accomplishes the goal of avoiding bias against bursty traffic. In network, connection is established along the range of burstiness against traffic and Drop tail/Random drop gateways. In this paper, goal is twofold: One is to avoid bursty traffic and the other is deciding the connection that reduces the synchronization and avoids congestion. To surmount the problem of congestion avoidance and synchronization, distinct algorithm is used by gateways. Randomization technique is deployed by gateways to notify the packet arrival, which increases the probability of packet arrival rate of particular connection notification. This method is implemented in an efficient manner without manipulating the gateway connection state. Congestion avoidance gateway comprises the capability of reducing the average queue size. This method works by dropping the arrived packets when the queue size exceeds certain threshold value.

Akintola AA, et al. (2005) modified the original RED algorithm termed Dynamic Random Earlier Detection (DRED) algorithm based on a parameter named warning line, which was introduced by them. DRED was used to measure the burstiness of the traffic that was coming in, which was carried through computing the average queue size that was adjusted dynamically. Here, the scholars were highly interested and utilized the gateway in order to avoid congestions. In case there was a difficulty in obtaining the feedbacks from the gateway then the protocol that was implemented in the transport layer deduce the congestion through the following ways.

The computed service time of bottleneck

Changes in the throughput

Modification in end-to-end delay

Drop of packets

Queuing behavior of the packets in the nodes of the network can be viewed unfriendly only through the gateway. Moreover, gateway was shared through many connections that were active with delay tolerance, throughput requirement, roundtrip time, etc. The gateway was focused mainly to enhance the performance of end-to-end congestion control approach. Figure 1 shows the architecture of the proposed DRED mechanism.

Fig 1: Architecture of DRED

The results that were derived from various experiments showed that the DRED responded quickly as the number of packets has been increased at the gateway. It also represented that the performance of the DRED has improved on comparing with the standard RED congestion mechanism. This also helped to evade the overflow of buffers at the gateways.

Another queue based technique was proposed by Sharma V, et al. (2002) to modify the RED technique. The authors have studied the queue dynamics and provided stability, packet loss probability, rate of convergence, and waiting time distribution. These results were extended to a case where two traffic classes were present for each stream renewal. With this concept it was very difficult and computationally too expensive to measure the performance for a system. To overcome this, authors focused on approximating the dynamics preset in the average length of the queue through an ODE. Here they have considered a transient where the empty system began at time zero to receive the input stream. It was also used to explain a case where the input process differs in time. A special case was considered for the packets with exponential size distribution and arrival process in terms of Poisson that allowed deriving specific and accurate results.

Apart from that RED has a disadvantage towards parameter settings. This made it to perform poorly in the non-optimum parameter setting strategies. The dynamic changes in the network require the RED mechanism to optimize the parameter settings and dynamically updated. To overcome this, Vaidya, et al. (2006) proposed an optimized technique for optimizing the parameter settings. In addition to that, scholars also described a method that was model-free technique designed for tuning the parameters of RED, which do not need the knowledge about the conditions of a network, also it preserves the efficiency of model based optimized technique. The proposed technique was considered as a black-box. They have developed a gradient based two-timescale simultaneous robust stochastic approximation algorithm along with the deterministic perturbation sequence. Therefore, this approach does not require any explicit simulation of the network during analyzing through experiments. Furthermore, the experimental results presented in this article have determined that the optimized RED performed much better than the algorithm that were modified the standard RED mechanism for congestion control.

To analyze the drop behaviour of RED Wang Y, et al. (2004) proposed a complicated discrete-time queueing model. A matrix-analytic approach is applied to analyse both the long-term and the short-term drop behaviour of a router with an RED scheme. The bursty nature of packet drop is examined by means of conditional statistics with respect to alternating congested and non-congested periods. The performance measures are derived by conditional statistics, including the longterm drop probability, and the three short-term measures comprising average length of a congested period, average length of a non-congested period, and the conditional packet drop probability during a congested period. The packet stream is considered to follow a discrete-time batch Markovian arrival process (D-BMAP), and the queuing model of the router with an RED scheme can be modeled as D-BMAP/D/1/K. With a threshold level set in the RED scheme, the drop behavior of the D-BMAP/D/1/K queuing system with the RED scheme is characterized by examining the conditional statistics in a congested period and in a non-congested period through two hypothesized discrete-time absorbing Markov chains. Formulas are derived to explore the distributions of the lengths of a congested period and a non-congested period. In addition, the distribution of the number of packets dropped during a congested period and the long-term packet drop rate are evaluated.

2.6 RWM MOBILITY VARIATIONS

Mobility model governs the changes in the moving direction and speed of terminals according to a deterministic approach or a random process. For sometimes, movement path of terminals can be restricted to predetermined paths. In case of ad hoc environments, such mobility models are impractical since wireless ad hoc networks are created ???on the fly???, and collecting data to generate the paths for all situations can be very complicated. Thus, a mobility model that dictates the movement of hosts due to a random process, that is, random mobility model, is more appropriate for the performance evaluation of these networks. Random mobility models formulate the movement pattern of mobile hosts by consecutive random length intervals called movement epochs. During each epoch, mobile terminal moves at a constant speed, and at a constant direction for a random amount of time. The speed and direction choice for each epoch may or may not be correlated with their values in the previous epochs, and mobility characteristics of other terminals. Denizhan N, et al. (2006) proposed a generalized random mobility model that is general enough to capture the major characteristics of a realistic movement profile, and yet is simple enough to mathematically formulate its long-run behavior with analytical expressions. The mobility pattern of a terminal that moves according to this generalized model is composed consecutive movement epochs in a closed region, and it is independent with the movement behavior of other terminals. During each movement epoch, mobile terminal firstly moves on the finite line segment joining the starting and destination points of the epoch at a random speed and then it pauses at the destination for a random amount of time. This generalized approach has a number of advantages. First, since destinations are selected from a general distribution, a movement scenario in which terminals selects some specific locations. Second, the generic approach for determining speed provides

a unique opportunity to select speed according to the distance that is going to be traveled, and also a method to model variable speed during movement epochs. Clearly, if the speed of the terminal can vary during moving, then this model can even be used to capture different acceleration characteristics of vehicles. This proposed model is capable of capturing several mobility scenarios and gives a mathematical framework for its exact analysis over one-dimensional mobility terrains.

To study and evaluate the performance of ad hoc routing protocol, Random way point model is the widely used synthetic model. According to this model, all the mobile nodes connected in a network moves along zig zag line, segmented straight line in the direction is termed as leg. Before moving along zig zag direction, node may retain constant for a random think time. Camp T, et al. (2002) focused on particular spatial properties of the RWP model, namely the node distribution and mean length of a leg. According to this property, the mean distance among two arbitrary turning points is estimated. Moreover, analytical expression is derived in this model to compute the node distribution and mean length of the leg. Node distribution application requires the combined transverse movement of probable initial and destination points. This four dimensional integration create significant complexity through analytical expression. To accomplish the distribution at constant normal, four dimensional integral is reduced into one dimensional integral by employing density expression.

Currently there are two types of mobility models used in the simulation of networks: traces and synthetic models. Traces are those mobility patterns that are observed in real life systems. Traces provide accurate information, especially when they involve a large number of participants and an appropriately long observation period. However, new network environments (e.g. ad hoc networks) are not easily modeled if traces have not yet been created. In this type of situation it is necessary to use synthetic models. Synthetic models attempt to realistically represent the behaviors of MNs without the use of traces. A survey of several mobility models have been proposed for the performance evaluation of ad hoc network protocols. A mobility model should attempt to mimic the movements of real MNs. Changes in speed and direction must occur and they must occur in reasonable time slots.

Fig 2. RWM model

Fig 2 illustrate example of a topography showing the movement of nodes for Random Mobility Model. The Random Waypoint Mobility Model includes pause times between changes in direction and/or speed. An MN begins by staying in one location for a certain period of time (i.e., a pause time). Once this time expires, the MN chooses a random destination in the simulation area and a speed that is uniformly distributed between [minspeed,

maxspeed]. The MN then travels toward the newly chosen destination at the selected speed. Upon arrival, the MN pauses for a specified time period before starting the process again. In most of the performance investigations that use the Random Waypoint Mobility Model, the MNs are initially distributed randomly around the simulation area. This initial random distribution of MNs is not representative of the manner in which nodes distribute themselves when moving.

2.7 ROUTING PROTOCOLS IN MANET

2.7.1 DYNAMIC MANET ON-DEMAND PROTOCOL [DYMO]:

According to Chakeres I.D, et al. (2009) DYnamic MANET On-demand (DYMO) protocol is a reactive routing protocol being developed within IETF's MANET working group. The recently developed protocol was DYnamic MANET On-demand DYMO which has the combined advantages of reactive protocol, AODV and DSR. DYMO is the extension of AODV. It is a new field in research and many works are carrying out in DYMO protocol. Typically, all reactive routing protocols rely on the quick propagation of route request packets throughout the MANET to find routes between source and destination. While this process typically relies on broadcasting, the route reply messages that are returned to the source rely on unicasting, DYMO is basically an improved version of AODV protocol every node records the next hop to send a packet to a specific destination.

Fig 3. Message processing in DYMO

In DYMO, route discovery can be performed by using RREQ and RREP messages. These messages contain information about the originator of the message along with the target node, hop count between nodes and a sequence number. This information is used to maintain a routing table held by each node. Notice that the routing table is updated correctly to avoid old routing information and the routes with possible loops. Fig 3 shows the processing of DYMO messages. DYMO protocol has a mechanism to notify nodes about a broken nodes by sending a Route Error (RERR).

To improve the performance in MANET environment, Marga N, et al. (2007) enhanced DYMO to support multipath traffic dispersion. Traffic dispersion is a technique that can help to prevent the threat of eavesdropping, to do load balancing or to minimize the energy consumed by nodes. In this paper, author analyzed the impacts of multipath routing with DYMO on both UDP and TCP traffic.

BoomaraniMalany A, et al. (2009) experiments the Quality of service in DYMO protocol with some implements of parameter change. It is the first experiment in variation of parameters with mobility. Dymo is the most recent ad hoc networking protocol proposed by the Manet working group especially for mobile adhoc network. In this paper, random waypoint mobility model was used for examination. The Random Waypoint model is the most commonly used mobility model in research community. At every instant, a node randomly chooses a destination and moves towards it with a velocity chosen randomly from a uniform distribution [0, V_max], where V_max is the maximum allowable velocity for every mobile node. After reaching the destination, the node stops for a duration defined by the 'pause time' parameter. After this duration, it again chooses a random destination and repeats the whole process until the simulation ends. The parameter Qualnet is used for evaluating mobility speeds in Dynamic Manet On Demand routing protocol. QualNet is a discrete event simulator developed by Scalable Networks. It is extremely scalable, accommodating high fidelity models of networks of 10???s of thousands of nodes. QualNet makes good use of computational resources and models large-scale networks with heavy traffic and mobility, in reasonable simulation times.

Fig 4. QualNet Simulator

The state diagram of a typical QualNet protocol is shown in Figure 4. The simulation results gives very high throughput and small delay in quality of service DYMO protocol of mobile adhoc network.

2.7.2 FISHEYE STATE ROUTING (FSR)

Chun-chun Y, et al. (2007) stated that Fisheye state routing protocol resembles the Fisheye which is used to reduce the data required to the graphical area. Fisheye state routing is based on global state routing [GSR] protocol both are based on the link state routing protocol. The scope of the fisheye at the centre is circled with red. The scope is defined as but the number of hops needed to reach the destination node. Each node does not contain the whole network information; instead, information about closer nodes is exchanged regularly rather than it is done about farther nodes, thus reducing the update message size. Each node divides the network into a number of scopes. Each scope i, a period Ti s is assigned which decides how often node may transmit information about the nodes in scope i to its neighbors. Depending upon the table the updated message size is reduced by using different exchange periods in the table. The entries corresponding to nodes within the smaller scope are propagated to the neighbors with the highest frequency. FSR is functionally similar to Link State (LS) Routing in that it maintains a topology map at each node. The key difference is the way in which routing information is disseminated. The reduction of routing update overhead in FSR is obtained by using different exchange periods for different entries in routing table. More precisely, entries corresponding to nodes within the smaller scope are propagated to the neighbors with the highest frequency. FSR produces timely updates from near stations, but creates large latencies from stations afar. However, the imprecise knowledge of the best path to a distant destination is compensated by the fact that the route becomes progressively more accurate as the packet gets closer to destination. The idea of fisheye leads to a multi-level routing zone structure in FZRP, in which different link state update rates are associated with different levels. By suppressing the link state entries the message size is reduced. Fisheye Zone Routing Protocol (FZRP) is an extension of Zone Routing Protocol (ZRP) adopting the concept of Fisheye State Routing (FSR).

According to Natarajan Meghanathan, et al. (2009), Fish-eye State Routing (FSR) protocol is a type of link-state based proactive routing protocol proposed to lower the traditionally observed higher control overhead with the proactive protocols. In FSR, a node exchanges its link-state updates more frequently with nearby nodes, and less frequently with nodes that are farther away. The number of nodes with which the link-state information is exchanged more frequently is controlled by the Scope parameter while the frequency of updating the neighbors outside the scope is controlled by the Time Period of Update (TPU) parameter. The operation of FSR is basically controlled by these two parameters. As a result, a node maintains accurate distance and path information to its nearby nodes, with progressively less accurate detail about the path to nodes that are farther away. This is also the basic principle behind the vision system for fishes and hence the routing protocol is named after this principle. A scope value of 1 and a larger TPU value typically results in a lower control overhead at the cost of a higher hop count path (a suboptimal path) between any two nodes. On the other hand, a scope value equal to the diameter of the network and a smaller TPU value basically transform FSR such as the control overhead (FSR to OLSR, resulting in higher control overhead with the advantage of being able to use the minimum hop path between any two nodes. However, in this research, the FSR can be normally operated with a smaller scope, typically hop, because even with larger TPU values, a data packet is more likely to get forwarded on a better path towards the destination as the packet approaches the destination. Our contributions in this paper are as follows: Given that the scope parameter is normally set to hop, the critical performance metrics for number of link-state messages exchanged), the hop count of the paths and energy consumption are heavily dependent on the TPU parameter.

2.7.3 ZONE ROUTING PROTOCOL (ZRP)

The Zone Routing Protocol (ZRP) combines the advantages of the proactive and reactive approaches by maintaining an up-to-date topological map of a zone centered on each node. Within the zone, routes are immediately available. For destinations outside the zone, ZRP employs a route discovery procedure, which can benefit from the local routing information of the zones. Zygmunt J, et al. (2002) presented Zone Routing Protocol and discussed the problem of routing in ad hoc networks. A routing zone is defined for each node separately, and the zones of neighboring nodes overlap. The nodes of a zone are divided into peripheral nodes and interior nodes. ZRP refers to the locally proactive routing component as the IntrA-zone Routing Protocol (IARP). The globally reactive routing component is named IntEr-zone Routing Protocol (IERP). IERP and IARP are not specific routing protocols. Instead, IARP is a family of limited-depth, proactive link-state routing protocols. IARP maintains routing information for nodes that are within the routing zone of the node. Correspondingly, IERP is a family of reactive routing protocols that offer enhanced route discovery and route maintenance services based on local connectivity monitored by IARP. ZRP performs better than any single proactive or reactive protocol. This is especially true when taken into account that almost any pure proactive and reactive protocol can be adapted as an IARP or IERP component of ZRP. However, the cost of ZRP is increasing complexity, and in the cases where ZRP performs only slightly better than the pure protocol components, one can speculate whether the cost of added complexity outweigh the performance improvement.

Zone Routing Protocol (ZRP) adopting the concept of Genetic Algorithm (GA) designed by Sateesh Kumar P, et al. (2009). GZRP is studied for its performance compared to ZRP in many folds like scalability for packet delivery and proved with improved results. ZRP is the destination node is within the routing zone of the source node. The route to the destination is available in the routing table of the source node which is produced due to IARP. However, if the destination node is not found in the routing table of the source node, it initiates the route discovery process by sending Route Request (RREQ) packets with the help of IERP. These RREQ packets are broadcasted by BRP. Every border node searches for the destination node within its routing table. When a route to the destination is found, a Route Reply (RREP) packet is sent back to the source node. The GZRP makes use of GA at each border node and generates possible alternative paths which may be optimal or sub-optimal. These alternative paths are stored at the border nodes for two basic reasons: they can utilize these routes as the alternative routes in case of the existing route fails or node fails. (Fault tolerance) and they can distribute the packets on multiple alternative routes to reduce the congestion and as well to balance the network (load balancing). At each border node, instead of broadcasting the RREQ packets on a primary path alone, they can be broadcasted on many routes. Even though, GA produces many possible alternative paths, we make use of the limited number of alternative routes which are either optimal or near optimal.

The Zone Routing Protocol (ZRP) it reduces considerably the average end-to-end delay and control overhead. The results indicate that GZRP is well balanced protocol compared to ZRP due to the mobility of the nodes and number of the nodes in a network is concerned.

2.7.4 SHARP ROUTING PROTOCOLS

Zygmunt J, et al. (2003) introduced the Sharp Hybrid Adaptive Routing Protocol (SHARP), which automatically finds the balance point between proactive and reactive routing by adjusting the degree to which route information is propagated proactively versus the degree to which it needs to be discovered reactively. This protocol utilizes the fundamental tradeoff between proactive versus reactive routing to find a good balance between route information propagated proactively and route information that is left up to on demand discovery. SHARP utilizes both a proactive and a reactive protocol to perform routing. Each SHARP node determines the network neighborhood, called proactive zone, in which routing information pertaining to itself is disseminated proactively. SHARP relies on a novel proactive routing algorithm that is both efficient and analytically tractable. However, SHARP can use any reactive routing algorithm whose costs can be characterized analytically; our current implementation uses off-the-shelf AODV. SHARP finds the ???sweet spot??? between the two routing regimes by dynamically adjusting the extent of proactive and reactive routing. This boundary is determined by an analytical model and guided by dynamically-performed empirical measurements from the physical network. An adaptive hybrid routing protocol requires the following three properties for successful deployment. Protocol is said to be adaptive when it should be applicable to a wide range of network characteristics. It should automatically adjust its behavior to achieve target goals in the face of changes in traffic patterns, node mobility and other network characteristics. Flexibility is achieved when protocol should enable applications to optimize for different application-specific metrics at the routing layer. These optimization goals should not be set by the network designer, but be placed under the control of the network participants. The protocol achieves more efficiency when it has better performance than pure, non-hybrid, strategies without invoking costly low-level primitives such as those for distributed agreement or reliable broadcast. Through a combination of protocol design, model, and mechanisms, SHARP???s hybridization approach provides all of these properties. SHARP achieves its performance through low cost mechanisms to determine zone sizes and control the extent of proactive routing.

2.8 ON DEMAND ROUTING PROTOCOLS

A central challenge in the design of ad hoc networks is the development of dynamic routing protocols that can efficiently find routes between two communicating nodes. In the literature, Das S.R, et al. (2000) implemented a systematic performance study of two dynamic routing protocols like DSR (Dynamic Source Routing) and AODV (Ad hoc On demand Distance Vector) for ad hoc networks have been analyzed on the basis of ???on demand???. While DSR and AODV share the on-demand behavior and initiate the routing activities only in the presence of data packets in need of route. DSR does not rely on any time based activities but uses source routing, however AODV uses a table driven routing framework and destination sequence numbers. In DSR sender knows the complete hop-by-hop route to the destination. It makes very aggressive use of source routing and route caching. AODV adopts different routing mechanism to maintain routing information. Since it maintain a routing table of one entry per destination. It uses sequence number maintained at each destination to determine the freshness of routing information and to prevent routing loops. Regarding utilization of individual routing table entries, AODV maintains timer based states in each node. The two on demand protocols comparatively share certain salient characteristics in case of route discovery only in the presence of data packets in the need for a route to a destination. There rise several important differences in the dynamics of these two protocols leads to significant performance differentials. When compared to AODV, more significantly greater amount of routing information has been accessed by DSR. AODV gather only a very limited amount of routing information due to the absence of source routing and promiscuous listening. Moreover AODV causes significant network overhead because of route learning limited only to the source of any routing packets being forwarded. DSR aggressively making the use of route caching thus replies to all requests reaching a destination from a single request cycle which will be useful in the case of primary route failure. But there is a possibility of a route reply flood. In case of AODV, destination replies only once to the request arriving first and ignores the rest. AODV has a much more conservative approach than DSR. Then finally, performance metrics are evaluated with key metrices like Packet delivery fraction, average end-to-end delay, and normalized routing load.

Another extended work imposed by Lakshmi M, et al. (2006) compares the performance of three protocols in an Ad hoc environment. It compares the Ad hoc On Demand Distance Vector protocol (AODV) and the Dynamic Source Routing protocol (DSR) with the Source Tree Adaptive Routing Protocol (STAR). It is the representative table driven protocol for ad hoc networking environment while AODV and DSR are the two most popular on demand protocols to date. The basic mechanisms in STAR include the detection of neighbors and exchange of topology information among nodes. Each node discovers and maintains topology information of the network to build the shortest path tree and prefer the suitable path to destination. STAR greatly reduces control overhead in ad hoc network environment by adopting the Least-Overhead Routing Approach (LORA). LORA rules are further defined for the case when the underlying MAC protocol does not support reliable transmission. The basic information unit in STAR is the representation of a link used by a node in its preferred paths to destinations form the source tree of the node. All three routing protocols use the network layer service to communicate control messages with neighbors. It is not simple to determine which of the three protocols under comparison has the best performance for ad hoc networks. A good criterion to choose a protocol might be the size and expected traffic load in the target network. The comparison is made in terms of control overhead, data delivery, and average latency. A simulated result shows that STAR gains more performance in small networks whereas AODV delivers more packets than the other two protocols.

2.9 SUMMARY

Many routing protocols are reviewed and proposed in literature for MANET. But our propose