1 / 14

Optimized Multicast

Optimized Multicast. Cho, song yean Samsung Electronics. What we trying to do ?. Communication Service inside the building. Testbed 30 nodes 4 gateway 802.11b Barriers Barrier per 20 m Barrier per 15 m. gateway. Services Push to Talk (PTT) VoIP Alarm & Notification Messenger.

moral
Télécharger la présentation

Optimized Multicast

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Optimized Multicast Cho, song yean Samsung Electronics

  2. What we trying to do ? • Communication Service inside the building Testbed • 30 nodes • 4 gateway • 802.11b • Barriers • Barrier per 20 m • Barrier per 15 m gateway Services • Push to Talk (PTT) • VoIP • Alarm & Notification • Messenger gateway 30m 80m

  3. What functions We need? PTT Agent Messenger VoIP Agent Voice CODEC LQ-OLSR For v6 OOM For v6 SIP RTP/RTCP SDP GSM Speex ACELP G.723.1 TCP/UDP IPv6 Auto Config Mobile IP Virtual Mac MACOM Kernel 802.11 MAC 802.11 MAC 802.11 MAC NIC NIC NIC

  4. Requirements for OOM • Minimizing Overhead • Utilizing Information I have already  Use Information (MPR, Routing Table) of OLSR • Making Multicast Delivery Path only for Real multicast data  building multicast delivery paths when Multicast Client try to send data • Maximizing group member reachability • Delivering Data to All Group Members  building redundant paths

  5. Assume what we have already? • Info from OLSR • Which Node is Group? • Shortest Hop Count to all Group members • Mechanism to Build Multicast Paths • Source Tree • Multicast Mesh • Stateless

  6. Tree vs. Mesh • Source Based or Shared Tree • Multicast Mesh • Stateless Source Based Mesh Based • Build Mesh • Mesh : multiple paths between source and group members • Outperforms tree-based multicasting when node moves frequently and fast • Alternative Path  Higher Reachability • Less Optimal • Ex) ODMPR, CAMP,FGMP  • Build Multicast Tree • Root: source • Leaves : group members • Combining shortest path from source to all group members • Optimal  Efficient • Fragile to Mobility Ex) AMRIS, MAODV,MOLSR

  7. Basic Idea • On-Demand : (Group Members, Shortest Hop Count) in Packet • Multiple Paths to Group Members • Node N is forwarder? Condition Effect D(S) + D(GM) <= Shortest Hop Count Use All Shortest Paths to Group Members Mesh • D(S) + D(GM) <= Shortest Hop Count • && MPRSelector(N) ∈ Sender Use some Shortest Paths connecting MPRS to Group Members Mesh +MPR • (D(S) + D(GM) <= Shortest Hop Count • && MPRSelector(N) ∈ Sender) or • (D(S) + D(GM) <= Shortest Hop Count +extra • && MPRSelector(N) ∈ Sender) Use some Shortest Paths connecting MPRS to Group Members including paths without MPRs Mesh +MPR +extra

  8. Simulation Model • Testbed Model • Random Unit graph G • Graph obtained by systematically linking pairs nodes whose distance is smaller or equal to the unit length-r • N : nodes distributed uniformly on square of size L x L unit lengths (L>0) • Mobility • 100 * (r / R) • r :distance to move at a time • R:distance to transfer the data

  9. Comparison Mechanism • MPR flooding with radius • delivering data through MPRs to all directions until it reaches the furtherest group members • Mesh • delivering data through all shortest paths from source to group members • Mesh+MPR • delivering data through some shortest paths connecting MPRs from source to group members • Mesh+MPR+extra • delivering data through some shortest paths connecting MPRs from source to group members and extending the paths to extra hops OOM for static topology OOM for dynamic topology

  10. Comparison Result • Overhead • Minimizing the Cost to transmit Multicast Data? • the number of forwarders • Group Reachability • Maximizing the Reachability of Group Members • (group members to receive the packet #/total node#) * 100

  11. Overhead without mobility Node # = 200

  12. Group Reachability with mobility Node #= 200 , Group member #= 10

  13. Overhead with mobility Node #= 200 , Group member #= 10

  14. Topology info MPR Selector Info Implementation Design on Linux OLSR Daemon Multicast Daemon Group Membership info Multicast Routing Info Group membership Info Multicast Data Forwarder Multicast Client Netfilter TCP/UDP IP

More Related