ExOR Opportunistic Multihop routing for Wireless Networks约会的机会多跳的无线网络路由.pptVIP

* Authors argue that performance benefits of long hops is because nodes could use asymmetric links * Results show that ExOR exploits long distance transmission. 25 % of Exor transmissions were 700 While 58% of traditional transmissions are very short distance ExOR: Opportunistic Multi-hop routing for Wireless Networks by; Sanjit Biswas and Robert Morris, MIT Presented by; Mahanth K Gowda Some pictures/graphs adopted from authors’ slides * OVERVIEW Traditional Routing ExOR: key intuitions and ideas ExOR: Realization Evaluation * TRADITIONAL ROUTING Links are abstracted as wires. A B C D E Destination Path E A D E C A D C * LINK TRANSMISSION IS A BROADCAST Probability of reception decreases with distance However, there is always a chance that data travels longer C E 90% 60% 10% * A D B EXOR EXPLOITS BROADCAST Best traditional route over 50% hops: 3(1/0.5) = 6 tx Throughput ? 1/# transmissions ExOR exploits lucky long receptions: 4 transmissions Assumes probability falls off gradually with distance src dst N1 N2 N3 N4 75% 50% N5 25% * EXOR EXPLOITS BROADCAST Traditional routing: 1/0.25 + 1 = 5 tx ExOR: 1/(1 – (1 – 0.25)4) + 1 = 2.5 transmissions Assumes independent losses N1 src dst N2 N3 N4 25% 25% 25% 25% 100% 100% 100% 100% * EXOR REALIZATION: BATCHING Packets are queued and sent in Batches A list of forwarders prioritized by their ETX values is included In the below topology --- Source: A, Destination: E Priority order : E C D B A In other words, if E C D B A receive packets, they should forward in that order Other nodes listen They forward packets only if a higher priority node has failed to do so * AN EXAMPLE A has transmitted a batch of 10 packets 1-10 E receives packets 1, 2 C receives 1 3 4 10 D receives 1 2 5 9 10 B receives 1 2 3 4 5 6 7 8 9 10 E received 1,2 1 2 3 4 5 6 7 8 9 10 Now C forwards 3, 4,10 1 2 3 4 5 6 7 8 9 10 D forwards 5,9 1 2 3 4 5 6 7 8 9 10 B forwards 6, 7, 8 1 2 3 4 5 6 7 8 9 10 * BATCHING