An Improved Qos Multipath Routing Using Bandwidth Estimation and Rate Adaptation

Mobile Adhoc Networks (MANETs) are composed of nodes which communicate with one another without network infrastructure. Their advantage being that they can be used in isolation or along with wired infrastructure, usually via a gateway node to ensure traffic relay for both networks. Quality of Service (QoS) is harder to ensure in ad hoc networks than in other network types, as wireless bandwidth is shared by adjacent nodes with network topology changing as nodes move. Most QoS protocols are implemented for specific scenarios and consider parameters such as network topologies, bandwidth, mobility, security and so on. This work proposes a novel multipath routing protocol which is an extension of AOMDV by discovering routes based on available bandwidth and rate adaptation. The method with Hello message box is used to calculate available bandwidth for a route. Relative Fairness and Optimized Throughput is an approach for rate adaptation in this paper which is to ensure fairness and allow nodes to adapt transmission rates and contention windows to channel quality. In sequence this is determined by calculating the access probability of a channel for each node in a distributed manner approximating successful and failed transmissions.


INTRODUCTION
In order to effectively adapt the transmission rate, network congestion has to be reliably detected. In particular, among all kinds of packet losses, the congestion loss probability needs to be estimated.
Treating all losses as congestion loss leads to undesirable rate adaptations (Fu et al. 2003). To effectively support best-effort multimedia streaming, dramatic rate variations are highly undesirable. To offer bandwidthguaranteed QoS, the available end-to-end bandwidth along a route from the source to the destination must be known (Chen & Heinzelman 2005). Available bandwidth is the amount of bandwidth left over after the cross traffic. It can be determined by finding the time period for which the link is not utilized for transmitting data. In recent years, the main focused research area is ad hoc network. In ad hoc network, available in the node to transmit packets over the network. Whole channel will not be used for packet transmission (Banu 2010).

METHODOLOGY
As seen in the previous chapter, the performance of network is improved when only rate adaptation and bandwidth estimation is considered during routing. In this chapter, the rate adaptation and bandwidth estimation proposed in the previous chapters are combined to propose a Rate adaptation, Bandwidth estimation AOMDV Routing. The proposed approach can be applied when the node needs to transmit the data with high capacity as well as dynamic rate adaptation.
The values of rate for transmission and available bandwidth are considered when selecting a path. The route with the maximum value is chosen as the path between source and destination.
QoS parameter estimation is through normalization of the values and final value is computed as follows: where , are constants with + = 1 and in this study, and are assigned the value of 0.5 . The route with the maximum P QoS is used to transmit the packets.   Rate adaptation is computed based on the improved technique implemented in chapter 3 using RREG. Similarly Bandwidth is also estimated using the modified HELLO packet. Larger rate adaptation improves the Packet Delivery Ratio but can also increase the congestion when the number of hops is high which can be identified using Bandwidth estimation. This work uses a combination of the normalized value of Rate adaptation and Bandwidth estimation to achieve better QOS.

RESULTS AND DISCUSSION
In this work, a QoS aware routing for AOMDV is proposed. This

Figure 4.4 End to End Delay
The performance metric End to End Delay is shown in Table 4

Figure 4.5 Remaining Energy in joules
The Table 4.4 gives the remaining energy levels in joules of proposed approach. The Figure 4.5 shows that the remaining energy level of proposed approach AOMDV with Bandwidth Estimation and Rate Adaptation is 0.53% higher than the AOMDV with Bandwidth Estimation at mobility speed of 10.8Kmph. But it increases to 2.83% when the mobility speed increases to 90Kmph.

CONCLUSION
This study proposes a QoS aware routing for AOMDV. QoS routing needs to locate a route from source to destination by considering the QoS constraints bandwidth and rate for transmission. The proposed method is an integration of QoS constraints for routing. Experimental results show that the AOMDV with Bandwidth Estimation and Rate Adaptation obtains high packet delivery ratio. Also the performance of the integrated approach is compared with all metrics and proved that the combination of the two constraints improved the QoS in AOMDV routing.