CGGC Cached Greedy GeocastCGGC缓存的贪婪算法.ppt

Evaluation results 2010/2/4 * Evaluation results 2010/2/4 * Evaluation results 2010/2/4 * Summary and outlook We have presented cached greedy geocast, a mechanism to improve greedy forwarding that is especially suitable for ad hoc networks with high mobility and high node velocities and low node densities. 2010/2/4 * Summary and outlook The work shows that both alternatives, size restricted and time restricted caching, can significantly increase the messages successfully delivered to a geographic destination region. 2010/2/4 * As long as a packet cannot be forwarded to a node closer to the destination, it surrounds the network gap on a tangential way. * This method is only used if no closer neighbor is available. * * CGGC: Cached Greedy Geocast 指導教授：許子衡 教授 報告學生：馬敏修 2010/2/4 * Outline Introduction Discussion of related work Overview of Cached Greedy Geocast Detailed problem description Enhanced greedy forwarding Cached geocast approach Implementation description Simulation results Environment and setup parameters Evaluation results Summary and outlook 2010/2/4 * Introduction Divide the overall delivery process in exactly these two substantial parts, which on the one hand is called line–forwarding as long as the initial packet hasn’t reached its destination region and on the other hand area–forwarding inside of this region. 2010/2/4 * Introduction The most obvious decision scheme in this context is greedy forwarding. A message is always forwarded to that neighbor, whose position is nearest to the packet’s destination, which is primarily defined through a position vector. 2010/2/4 * Introduction If a node tries to forward a message while no neighbor is present that is closer to the destination than the relaying node himself. We call it a local maximum or dead end — this node is not able to continue the greedy forwarding at this point in time. 2010/2/4 * Introduction 2010/2/4 * Discussion of related work GPSR(Greedy perimeter stateless routing) is a position–based