230 likes | 392 Views
A Probabilistic Routing Protocol for Mobile Ad Hoc Networks. Abdallah Jabbour • James Psota • Alexey Radul {ajabbour, psota, axch}@mit.edu. presentation notes (delete this slide).
E N D
A Probabilistic Routing Protocol for Mobile Ad Hoc Networks Abdallah Jabbour • James Psota • Alexey Radul {ajabbour, psota, axch}@mit.edu 6.829 Final Project
presentation notes (delete this slide) • slides are currently not formatted; formatting will occur after content is more finalized; please don’t worry about that for now • total presentation time: ~15 minutes (5 min questions) 10-15 slides total • should give protocol a name! • need to add figures, but will talk to teammates first • may need to consolidate/merge slides… • I’d like to add an animation/diagram of the protocol in action 6.829 Final Project
Routing in Ad Hoc Networks • Most routing protocols… • Use fixed route to send all packets from a given source to a given destination • Send along path with minimum hop count • Use two main types of packets • Data packets • Control (routing) packets • <one reason why these are bad> • We think we can do better! 6.829 Final Project
Outline • Related Routing Protocols • DSDV, DSR, AODV • Probabilistic Routing Protocols • Shortcomings of Related Protocols • Protocol Description • Simulation Overview and Results • Conclusion 6.829 Final Project
Related Routing Protocols • Destination-Sequenced Distance Vector • Loop-free, hop-by-hop distance vector • Routes prioritized by sequence numbers • Dynamic Source Routing • Employs source routing • Floods route requests • Maintains routes by link breakage notification • Ad Hoc On-Demand Distance Vector • Combines sequence numbers from DSDV and Route Discovery from DSR 6.829 Final Project
Probabilistic Routing Protocols • Control and routing information (“ants”) sent randomly • Data forwarded deterministically to path with best metric • Examples • Ant-Based Control (ABC) • AntNet • Ant-Colony-Based Routing Algorithm (ARA) 6.829 Final Project
Drawbacks and Limitations of Above Protocols • Routing packets hinder performance • Decreases available bandwidth • Increases transmission latency • High recovery latency due to static routes • DSDV, DSR, AODV • Probabilistic protocols incorrectly assume symmetric traffic • Above protocols use shortest hop routes • Tends to choose routes with less capacity than optimal ones • Tends to use marginal links 6.829 Final Project
Our Protocol’s Design Goals • Minimize routing packets • Especially those interfering with traffic • Use a better measure of network state than minimum hop count to make forwarding decisions • Better cope with link outages 6.829 Final Project
Protocol Overview • Minimize control packets by prepending protocol-level headers onto all data packets • Both when originating and forwarding a packet • Nodes promiscuously listen to all traffic, using protocol headers to update state • Base routing decisions on link loss ratios • Modular in both choice of metric and metric to probability mapping • Use ETX instead of minimum hop count 6.829 Final Project
Random Routing • Nodes forward probabilistically to neighbor ni with probability pi n1 routingtable p1 = 0.1 s n2 d p1 = 0.4 p3 = 0.5 n3 • Route is not fixed, so packets can still reach destination immediately upon link breakage 6.829 Final Project
Random Routing • Nodes forward probabilistically to neighbor ni with probability pi n1 routingtable p1 = 0.3 x s n2 d x x x p1 = 0.4 link breaks! p3 = 0.7 n3 • Update forwarding probability upon link breakage (nodes see infinite loss ratio on link) 6.829 Final Project
Deterministic Routing • Forward ALL packets along path with best metric • Our flexible infrastructure allowed simulation of both • First to compare Random to Deterministic Routing 6.829 Final Project
Node State • Nodes maintain the following state • Dynamically-updated set of neighbors • Loss ratios to and from each neighbor • Routing state • Metric values for each destination and each neighbor-destination pair • Probability of forwarding to a certain neighbor in order to reach a desired destination • Requests and fulfillments thereof for information about destinations 6.829 Final Project
Protocol Header Contents • Each outgoing packet contains a protocol-level header 6.829 Final Project
State Update • Nodes update state • Upon sending • Upon receiving • Periodically • Refresh stale state and alert neighbors that you’re still alive • Probability Distribution Updates • Probability distribution and metric values updated along with other node state • Values evolve in response to nodes entering and leaving the system and changes in link quality 6.829 Final Project
Simulation Environment • 50 mobile nodes in 1500m x 300m area • Random waypoint movement model • 900s simulation time • We investigated… • Pause times • Node speeds • Connection patterns • Packet sizes • Used 30 CBR UDP sources • Avoid TCP because YYY 6.829 Final Project
Evaluation Measures • Packet delivery ratio • Packet delivery latency • Goodput • Number of bytes of overhead • Path length optimality • Transmission overhead • Route acquisition latency 6.829 Final Project
Simulation Results 6.829 Final Project
Conclusions 6.829 Final Project
insert title slide 6.829 Final Project
Backup Slides 6.829 Final Project
Implementation Overview • ns-2 with Monarch mobility extensions used to model mobile ad hoc network • Implemented protocol in C++ • Data generated by 30 CBR UDP sources • Metrics 6.829 Final Project
Routing and Distribution Updates • Random Routing • Each node sends probabilistically to neighbor i with probability pi for destination d • Deterministic Routing • Each node sends exclusively to neighbor i when probability pi for destination d is highest • Our flexible infrastructure allowed simulation of both 6.829 Final Project