Lecture 18:Mesh Networking

CSE 222A: Computer Communication NetworksAlex C. Snoeren

Thanks:Sanjit Biswas, Lili Qiu

Lecture 18 Overview

● Wireless mesh networks

● ExOR

22CSE 222A – Lecture 18: Mesh Networking

● Dense 802.11-based mesh● Goal is high-throughput and capacity

1 kilometer

MIT Roofnet

33CSE 222A – Lecture 18: Mesh Networking

packet

packet

packet

● Identify a route, forward over links● Abstract radio to look like a wired link

src

A B

dst

C

Multi-hop routing

44CSE 222A – Lecture 18: Mesh Networking

Radios aren’t wires

● Every packet is broadcast● Reception is probabilistic

123456123 63 51 42345612 456 src

A B

dst

C

55CSE 222A – Lecture 18: Mesh Networking

packet

packetpacketpacketpacketpacketsrc

A B

dst

C

packetpacketpacket

● Decide who forwards after reception● Goal: only closest receiver should forward● Challenge: agree efficiently and avoid duplicate transmissions

ExOR

66CSE 222A – Lecture 18: Mesh Networking

● 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%

Example scenarios

77CSE 222A – Lecture 18: Mesh Networking

● 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 dstN1 N2 N3 N4

75%50%

N5

25%

Example scenarios

88CSE 222A – Lecture 18: Mesh Networking

Issues to Address● What we want: an effective protocol with low overhead● How often should ExOR run?

u Per packet is expensiveu Use batches

● Who should participate the forwarding? u Too many participants cause large overhead

● When should each participant forward?u Avoid simultaneous transmissions

● What should each participant forward?u Avoid duplicate transmissions

99CSE 222A – Lecture 18: Mesh Networking

Who should participate? ● A background process collects ETX

information via periodic link-state flooding.

● The source chooses the participants (forwarder list) using ETX-like metric.u Only consider forward delivery rate

» Why?u The source runs a simulation and selects only the

nodes which transmit at least 10% of the total transmission in a batch.

1010CSE 222A – Lecture 18: Mesh Networking

● Forwarders are prioritized by ETX-like metric to the destination

● The highest priority forwarder transmits when the batch ends

● The remaining forwarders transmit in prioritized order

● Question: How does each forwarder know it is its turn to transmit?u Assume other higher priority nodes send for five

packet durations if not hearing anything from them

When to forward?

1111CSE 222A – Lecture 18: Mesh Networking

ExOR batching

● Source estimates ETX between each node and the destination

● Source decides on a list of forwarders and prioritizes the list. Let the list be (dst, N4, N3, N2, N1)

● Node closest to the dst sends the overheard packet firstu Other nodes listen, send remaining packets in turn

src

N3

dstN4

tx: 23

tx: 57 -23@ 34

tx: @ 31

tx: 100

rx: 23

rx: 57

rx: 88

rx: 0

rx: 0tx: 0tx: @ 12

N1

N2rx: 45

rx: 18

1212CSE 222A – Lecture 18: Mesh Networking

Which packets: Batch maps● Batch map indicates, for each packet in a batch,

the highest-priority node known to have received a copy of that packet.

N0

N1

N2

N3

1st round Tx: 1, 2, 3, 4, 5, 6, 7, 8

Forwarder list:N3(dst), N2, N1, N0 (src)

Rx:2,5,8

Rx: 2,4

Rx: 1,2,7,8 Batch map: 03030000

Batch map: 03032002

Batch map: 13032012

Tx: batch map only

Tx: 5,8

Tx: 1,7

2nd round Tx: 3,6Batch map: 13032012

1313CSE 222A – Lecture 18: Mesh Networking

End Game● A nodes stops sending the remaining packets in the

batch if its batch map indicates over 90% of this batch has been received by higher priority nodes.

● The remaining packets transferred with traditional routing.

1414CSE 222A – Lecture 18: Mesh Networking

Example

Forwarder list: N24(dst), N20, N18, N11, N8, N17, N13, N5(src)

1515CSE 222A – Lecture 18: Mesh Networking

Using ExOR with TCP

NodeProxy

ExOR

GatewayWeb Proxy

Client PC Web ServerTCP TCP

ExOR Batches (not TCP)

● Batching requires more packets than typical TCP window

1616CSE 222A – Lecture 18: Mesh Networking

Evaluation on Roofnet

1 kilometer

1717CSE 222A – Lecture 18: Mesh Networking

● 65 Node pairs● 1.0-MB file transfer● 1 Mbit/s 802.11 bit rate● 1-KB packets

Traditional Routing ExOR

802.11 unicast with link-level retransmissionsHop-by-hop batchingUDP, sending as MAC allows

802.11 broadcasts100 packet batch size

Evaluation setup

Throughput improves

● Median throughputs: 240 Kbits/sec for ExOR, 121 Kbits/sec for Traditional

Throughput (Kbits/sec)

1.0

0.8

0.6

0.4

0.2

00 200 400 600 800

Cum

ulat

ive

Frac

tion

of N

ode

Pairs

ExORTraditional

1919CSE 222A – Lecture 18: Mesh Networking

25 Highest throughput pairs

Node Pair

Thro

ughp

ut (K

bits

/sec

)

0

200

400

600

800

1000 ExORTraditional Routing

1 Traditional Hop

1.14x

2 Traditional Hops1.7x

3 Traditional Hops2.3x

2020CSE 222A – Lecture 18: Mesh Networking

25 Lowest throughput pairs

Node Pair

4 Traditional Hops3.3x

Longer Routes

Thro

ughp

ut (K

bits

/sec

)

0

200

400

600

800

1000 ExORTraditional Routing

2121CSE 222A – Lecture 18: Mesh Networking

ExOR uses links in parallel

Traditional Routing3 forwarders

4 links

ExOR7 forwarders

18 links

2222CSE 222A – Lecture 18: Mesh Networking

ExOR moves packets farther

● ExOR average: 422 meters/transmission● Traditional Routing average: 205 meters/tx

Frac

tion

of Tr

ansm

issio

ns

0

0.1

0.2

0.6 ExORTraditional Routing

0 100 200 300 400 500 600 700 800 900 1000

Distance (meters)

25% of ExOR transmissions

58% of Traditional Routing transmissions

2323CSE 222A – Lecture 18: Mesh Networking

For Next Class…

● Review for quiz Wednesday

2424CSE 222A – Lecture 18: Mesh Networking