multicast-enabled landmark (m-lanmar) : implementation and scalability yunjung yi, mario gerla, js...
Post on 19-Dec-2015
213 views
TRANSCRIPT
Multicast-Enabled Landmark Multicast-Enabled Landmark (M-LANMAR) : (M-LANMAR) :
Implementation and scalabilityImplementation and scalability
YunJung Yi, Mario Gerla, JS Park, Yeng Lee, SW LeeYunJung Yi, Mario Gerla, JS Park, Yeng Lee, SW Lee
Computer Science DeptComputer Science DeptUniversity of California, Los AngelesUniversity of California, Los Angeles
The AINS ScenarioThe AINS Scenario
FLIRFLIR
LANMARLANMAR Key insight: Key insight: nodes move in teams/swarmsnodes move in teams/swarms Each team is mapped into aEach team is mapped into a logical subnet logical subnet IP-like Node addressIP-like Node address = <subnet, host> = <subnet, host> Address compatible with IPv6Address compatible with IPv6 Team leaderTeam leader (Landmark) (Landmark) elected in each group elected in each group
Logical SubnetLogical Subnet
LandmarkLandmark
Scalable Ad hoc Scalable Ad hoc multicastingmulticasting
Multicast (ie, transmit same message to all Multicast (ie, transmit same message to all member of a group) critical in battlefieldmember of a group) critical in battlefield
““Multiple unicast” does not scaleMultiple unicast” does not scale Current ad hoc Current ad hoc multicast solutions: multicast solutions: inappropriateinappropriate
They do not exploit affinity team modelThey do not exploit affinity team model multicast tree approach is “fragile” to mobility;multicast tree approach is “fragile” to mobility; no congestion control; no reliable end to end deliveryno congestion control; no reliable end to end delivery
Proposed approach:Proposed approach: TEAM MulticastTEAM Multicast
swarmswarm
Command postCommand post
Swarm Swarm LeaderLeader
UAVs: - equipped with video, chemical sensors - read data from ground sensors - “fuse” sensor data inputs - multicast fused data to other teams
Team Multicasting
Two-tier team multicast: M-Two-tier team multicast: M-LANMARLANMAR
Extension of LANMAR enabling Extension of LANMAR enabling multicastmulticast Inter-teamInter-team communication: communication: unicastunicast
tunneling from the source to the tunneling from the source to the representative of each subscribed team representative of each subscribed team
Intra-teamIntra-team communication: scoped communication: scoped floodingflooding within a team within a team
Source node
LM2
LM3
Subscribed Teams
LM4
Tunneling to landmarks
Flooding
Flooding
Scope = 2
Scope = 2
M-LANMAR
Testbed configurationTestbed configuration 3 teams (= 3 IPv4 subnets), 1 sender, 3 receivers3 teams (= 3 IPv4 subnets), 1 sender, 3 receivers
Dell P4 laptop with Lucent Orinoco 802.11b pcmcia cardDell P4 laptop with Lucent Orinoco 802.11b pcmcia card CBR traffic (512B/packet, 5~15 packets/sec)CBR traffic (512B/packet, 5~15 packets/sec) Protocols:Protocols: ODMRP; M-LANMARODMRP; M-LANMAR
Sender
Experimental Results: Experimental Results: Delivery Ratio and Control Delivery Ratio and Control
OverheadOverhead
0.0
0.2
0.4
0.6
0.8
1.0
1.2
5 10 15Sending rate (packet/sec)
Del
iver
y ra
tio
M-LANMAR delivery ratio ODMRP delivery ratio
M-LANMAR has higher Delivery Ratio than ODMRP: unicast tunneling helps reliable data delivery as it incorporates RTS/CTS/ACK)
M-LANMAR has higher control overhead
0.0
0.5
1.0
1.5
2.0
2.5
5 10 15Sending rate (packet/sec)
Con
trol
Ove
rhea
d
M-LANMAR control overhead
ODMRP control overhead
ScalabilityScalability Objective: test M-LANMAR scalabilityObjective: test M-LANMAR scalability Compared withCompared with
ODMRPODMRP Flooding Flooding
Simulation EnvironmentSimulation Environment QualNetQualNet 1000 nodes 1000 nodes forming forming 36 teams36 teams on on 6000 x 6000 m6000 x 6000 m22 field field CBR traffic (CBR traffic (512 bytes512 bytes/packet, 1packet/sec)/packet, 1packet/sec)
Simulation ResultsSimulation Results
As the number of multicast groups increasesAs the number of multicast groups increases ODMRP suffers from large control overhead and collisionsODMRP suffers from large control overhead and collisions M-LANMAR achieves high delivery ratio (by unicast tunneling M-LANMAR achieves high delivery ratio (by unicast tunneling
and flooding)and flooding)
Reliable Multicast Reliable Multicast SupportSupport
Reliable Adaptive Lightweight Multicast (RALM)Reliable Adaptive Lightweight Multicast (RALM) Source continually monitors the channel Source continually monitors the channel
conditioncondition No congestionNo congestion: the source transmits at “native” rate: the source transmits at “native” rate CongestionCongestion detecteddetected (i.e., packet loss feedback via (i.e., packet loss feedback via
NACK): the source falls back to “send-and-wait” NACK): the source falls back to “send-and-wait” mechanism (source stops upon receiving a NACK; it mechanism (source stops upon receiving a NACK; it resumes when it receives an ACK )resumes when it receives an ACK )
Combining with M-LANMARCombining with M-LANMAR Only landmarks return feedback (e.g. NACK/ACK) to Only landmarks return feedback (e.g. NACK/ACK) to
the sourcethe source Prevents unnecessary feedback implosionPrevents unnecessary feedback implosion
Simulation Results with RALMSimulation Results with RALM“Reliable Multicast”“Reliable Multicast”
(1000 nodes, 3 teams for each group, 5 (1000 nodes, 3 teams for each group, 5 multicast groups)multicast groups)
Delivery Ratio
0
0.2
0.4
0.6
0.8
1
1.2
512 1024 1280 1689.6 2560 5120
Offered Load (Bytes/sec)
Deliv
ery
Ratio
M- LANMAR w/ UDP ODMRP w/ UDPM- LANMAR w/ RALM ODMRP w/ RALM
ODMRP suffers from feedback implosion; congestion is unacceptable