The Research Of DSR-iETT Routing Protocol Based On Geographical Location Prediction

A Mobile Ad Hoc Network is a distributed wireless multi-hop network, which doesn’t depend on any fixed infrastructure. It has characteristics of non-center, self-organized, the dynamic topology and limited resources of the mobile node, etc. In view of these characteristics, many routing protocols have been proposed, for example, typical AODV and DSR routing protocols choose the shortest path as routing metric. However, the unreliability and sharing of wireless link makes the protocols hardly acquire the actual performance of the physical link, which needs to choose the appropriate routing metric for routing. So far, various route metrics have been proposed, such as ETX, ETT, WCETT and iETT, etc. iETT is the improvement of ETT in these routing metrics, which considers other two factors including loss rates differences in links and MAC overhead affecting the whole path performance on the basis of ETT.The acquisition of iETT mainly includes the acquisition of df and dr, which stand for the link forward and reverse direction success rates respectively, and data rate in every link along the path. The calculation of df and dr are according to the link probe packet broadcasted periodically by each node in the network. The two adjacent nodes of a link remember the number of probes from the two directions they receive to calculate the df and dr, which are used to calculate the value of the link loss rate P and ETX. The calculation of the link data rate was based on the distance between nodes calculated through location information of adjacent nodes, which should be compared with the transmission ranges of different rates in order to determine the data rate of the link between nodes. Based on the analysis, DSR-iETT routing protocol was proposed, and its implementation was described in the NS2Simulator. But evaluation of the quality of the link by iETT metric used could only reflect the loss rate of the network in the past period. The iETT value wasn’t able to reflect real status of the network. In order to solve this problem, a geographical location prediction algorithm based on ARIMA was proposed and which can predict the node’s location during next interval. In this way, we could judge that whether the nodes would still be in each other’s communication coverage during next interval. Thus, the predicted result would act as the pre-condition of the iETT.DSR-iETT routing protocol based on geographical location prediction (DSR-L-iETT) and DSR routing protocol were simulated and analyzed on the NS2simulation platform by changing the three scene parameters including the residence time nodes stay and movement speed respectively. The simulation results showed that DSR-L-iETT routing protocol was able to increase the network throughput, reduce the average end-to-end delay, and improve the network overall performance compared with the common DSR routing protocol.