Routing Protocol Comparison - University of California ...Routing Protocol Comparison Josh Broch,...
Transcript of Routing Protocol Comparison - University of California ...Routing Protocol Comparison Josh Broch,...
Routing Protocol Comparison
Josh Broch, David A. Maltz, David B. Johnson, Yih-Chun Hu, and Jorjeta Jetcheva.
A Performance Comparison of Multi-Hop Wireless Ad Hoc Network Routing Protocols.
In Proceedings of the Fourth Annual International Conference on Mobile Computing and Networking (MobiCom'98), ACM, Dallas,
TX, October 1998.
-Shaan Mahbubani
Overview
● Focus
● Protocols
● Simulation
● Paper Evaluation
● Discussion
Focus
● “infrastructureless networking”
● Ad hoc networks
● Multiple hops
● Routing protocols
● Performance comparison
Protocols
● Destination-Sequence Distance Vector
● Temporal-Ordered Routing Algorithm
● Dynamic Source Routing
● Ad Hoc On-Demand Distance Vector
Destination-Sequence Distance Vector
● Hop by hop Distance Vector● Bellman Ford● Loop freedom
● Per node routing table– Destination– Next hop– Number of hops– Sequence number
Temporally Ordered Routing Algorithm
● Link reversal● Minimize communication● Avoid overhead rather than shortest path
● Query, Update● Internet MANET Encapsulation (IMEP)
– Reliable in order delivery– Link status: BEACON, HELLO
Dynamic Source Routing
● Source routing● Each packet: complete route
● Route Discovery● Route Maintenance● Caching optimizations
Ad Hoc On-Demand Distance Vector
● Like DSR– on demand route discovery, maintenance
● Like DSDV– hop by hop routing, route sequence numbers
● Route Request● Route Reply
– Number of hops to destination● Maintenance
– Periodic HELLO
Simulation
● Methodology
● Results
● Other observations
● Related work
Methodology
● Goal: "Measure the ability of the routing protocols to react to network topology changes while continuing to successfully deliver data packets to their destinations"
● Scenarios
● Movement: Random Waypoint Model
● Communication: constant bit rate– no TCP
● Metrics– Packet Delivery Ratio– Routing Overhead– Path Optimality
Results: Packet Delivery
● Best to worst:– DSR, AODV, TORA,
DSDV
● DSR, AODV: load independent
● DSDV: stale routing
● TORA: routing loops from link reversal
Packet Delivery Details
Results: Routing Overhead
● Least to most overhead:– DSR, DSDV, AODV,
TORA
● DSR: decreasing incremental overhead
● AODV: less caching
● DSDV: constant overhead
● TORA: positive feedback loop -> congestive collapse
Routing Overhead details
Other Observations
● Overhead: DSR worse than AODV when measured in bytes
● DSDV updates: upon new sequence or metric number– Sequence: more overhead, better packet
delivery
● Broadcast reliability
● ARP/on demand protocol interaction:– Does not cache enough waiting packets
Related Work
● Park & Corson– TORA– Over simplified simulation: no node mobility
● Johnson & Maltz– DSR– Lacked radio propagation, MAC
● Freisleben & Jenson– DSDV, DSR– Simulation deficiencies (synchronization)
Paper Evaluation
● Strengths
● Weaknesses
● Further discussion points
Evaluation: Strengths
● Realistic Model– 802.11 MAC– Physical layer
● Detailed Comparison– 210 scenarios– Varied pause time, # of data sources
● Effective metrics– Packet delivery ratio– Overhead
● Detailed considerations– Considered implementation specific optimizations– Proposed protocol specific reasons for performance differences
Evaluation: Weaknesses
● Comparing very specific aspect of protocol– Only changes to network topology– No load/stress, normal case overhead
● Not real enough– Flat space, no obstructions– Too much movement– No device locality– Constant bit rate– Scalability– Effect of TCP
Evaluation: Overall
● Achieved what they said they would
● Other information needed for a more complete comparison
Discussion
● Implementation– Which protocol is easier/better?
● Node behavior– Malicious/lazy nodes?
● Power consumption
● Scalability
● Security