Font Size: a A A

Research On An Improved ZRP Algorithm Based On Relative Velocity

Posted on:2020-01-20Degree:MasterType:Thesis
Country:ChinaCandidate:W H ZhongFull Text:PDF
GTID:2428330602952375Subject:Engineering
Abstract/Summary:PDF Full Text Request
A Mobile Ad-Hoc Network(MANET)is a group of mobile nodes building a temporary network with no need of any central access points,fixed infrastructures,or central administrations.In a MANET,information transfer between two nodes,maybe requires the relay on some nodes due to the limited transmission coverage.Furthermore,the high mobility of nodes makes the network even more complicated.According to the characteristics of MANETs,the Zone Routing Protocol(ZRP)was proposed.ZRP is a hybrid routing protocol in which each node proactively maintains the routes within its routing zone.On the other hand,in the ZRP,awareness of routing zone topology is to improve the efficiency of the on-demand routing a mechanism.The radius of zone is an important factor in influencing the performance of the network.However,the strategy with fixed routing zone in ZRP has a defect of the slow response to dynamic topology changes and parameters in improved ZRP protocols cannot be regulated precisely.Since the traditional ZRP protocol and improved protocols cannot adapt to topology changes in MANET,a new efficient improved ZRP protocol that is sensitive to topology changes,RVBAZRP,is proposed.It can make use of node relative velocity to adapt to topology changes.The algorithm transforms the zone radius adjustment task into the classification task in machine learning,namely increasing the radius,maintaining the same and reducing the radius.The multi-class logistic regression model is used for prediction,and the radius is adjusted by the classification result.The algorithm mainly has the following characteristics: Firstly,the algorithm can improve the sensitivity of the network by adjusting the lifetime of the routing entries;Secondly,the algorithm regulates weights of overhead in different time intervals to provide more accurate eigenvectors for radius regulators.Finally,The radius adaptive regulator is designed by the multi-class logistic regression model to predict the radius adjustment and improve the network packet delivery rate.This RVBAZRP protocol is simulated on the OPNET platform,compared with the traditional ZRP and ZRP-DV protocol.The simulation result shows that RVBAZRP can improve nodes' sensitivity and the packet delivery rate.However,overhead may be higher in some scenarios.RVBAZRP can fully utilize the bandwidth resources when the topology changes rapidly,and get higher packet delivery rate,lower end-to-end delay and routing overhead.
Keywords/Search Tags:ZRP, logistic regression, radius adjustment, relative velocity, feature vector
PDF Full Text Request
Related items