Announcements

Project progress reports due today.

Homework 2 ready later today – due 6/2 (next Friday)

Graded HW 1 and solutions ready shortly.

Third paper summary on ad-hoc networks due next Wednesday.

Ad-Hoc Wireless Networks

Main Characteristics Each node generates independent data Any node can communicate with any

other. No centralized controller (self-

configuring) Data transmitted in (short) packets Links typically symmetric. Nodes may be mobile and/or power

constrained. Typically a large number of nodes

Applications Battlefield communications Wireless LANs Emergency infrastructures Short-term networks (e.g.

convention) Sensor networks

Medical applications (on-body) Buildings Wide area

Cellular phone evolution

Communication infrastructure for automated vehicles Automobiles Airplanes

Widely different channel characteristics, distances, mobility, and rate requirements.

Design Issues

Link Layer design

Channel sharing (MAC/reuse)

Reliability/QOS

Routing

Network topology

Network management/controlMust exploit synergies

between design layers

Link Layer Issues

Modulation and Coding Robustness Rate requirements Performance Adaptive techniques

Rate, power, BER, code, framing.

Bandwidth requirements Control and communication

requirements

Power control Typically distributed

Antenna design Smart antennas Multipath mitigation Multiuser detection

Connectivity Binary or adaptive.

Channel Access

Frequency-Division

Time-Division

DS Spread Spectrum

FH Spread Spectrum

Frequency reuseBandwidth efficientDistributed allocationDynamic channel

allocation hard for packet data

Frequency Division

Fixed allocation inefficientHard to implement when

node locations dynamically change

Distributed dynamic channel allocation hard to do

FD typically only used to create hierarchical networks

Time-Division

Fixed allocation inefficient and impractical (as in FD)

Aloha InefficientNo capture

Carrier sensingHidden nodes degrade

performanceBusy tone may interfere with

transmission to other nodes.

Busy Tone

Spread Spectrum Code

Assignment

Common spreading code for all nodesCollisions occur whenever receiver

can “hear” two or more transmissions.

Near-far effect improves capture.Broadcasting easy

Receiver-orientedEach receiver assigned a spreading

sequence.All transmissions to that receiver

use the sequence.Collisions occur if 2 signals

destined for same receiver arrive at same time.

Can randomize transmission time.

Little time needed to synchronize. Transmitters must know code of

destination receiver Complicates route discovery. Multiple transmissions for broadcasting.

Transmitter-orientedEach transmitter uses a unique

spreading sequenceNo collisionsReceiver must determine sequence

of incoming packet Complicates route discovery. Good broadcasting properties

Poor acquisition performance

Preamble vs. Data assignmentPreamble may use common code

that contains information about data code

Data may use specific codeAdvantages of common and

specific codes: Easy acquisition of preamble Few collisions on short preamble New transmissions don’t interfere with

the data block

Data link control

Packet acknowledgements neededMay be lost on reverse linkShould negative ACKs be used.

Combined ARQ and codingRetransmissions cause delayCoding may reduce data rateBalance may be adaptive

Hop-by-hop acknowledgementsExplicit acknowledgementsEcho acknowledgements

Transmitter listens for forwarded packet Not possible with directive antennas. Large delays in FIFO queues. More likely to experience collisions than

a short acknowledgement.Hop-by-hop or end-to-end or both.

Connectivity

Determining connectivitySNR measurementsBit/Packet error rate

Connectivity controlLink can adapt to maintain

connectivity (adapt rate, power,…)

Interaction with routing protocol.

Power increase may affect other nodes (Bambos technique).

How many connected nodes constitute a networkOr, take what you can get.

Routing (1987)

FloodingBroadcast packet to all neighbors InefficientRobust for fast changing

topologies.Little explicit overhead

Point-to-point routingRoutes follow a sequence of linksConnection-oriented

Explicit end-to-end connection Less overhead/less randomness Hard to maintain under rapid

dynamics.Connectionless

Packets forwarded towards destination

Local adaptation

Route dessemination

Route computed at centralized nodeMost efficient route computation.Can’t adapt to fast topology changes.

Distributed route computationEach node transmits connectivity

information to other nodes.Nodes determine end-to-end route

based on this local information.Adapts locally but not globally.

Nodes exchange local routing tablesNode determines next hop based on

some metric.Deals well with connectivity dynamics.Routing loops common.

Routing (1999*)

Table-drivenDestination-sequenced

distance-vectorClusterhead gateway switch

routingWireless routing protocol

On-Demand RoutingOn-demand distance vector

routingDynamic source routingTemporally ordered routingAssociativity-based routingSignal stability routing *”A review of current routing protocols for ad hoc mobile

wireless networks,” Royer and Toh, IEEE Personal Communications Magzine, April 1999.

Packet Forwarding

Overhead informationRouting informationPacket identifiersPriority/delay

informationTradeoffs in overhead

size

Synergies of routing and packet forwarding with link layer.

Other Network Issues

Network CapacityAdmission ControlInterface with wired

networksSecurityUpgrades

Software changesSoftware radios

Network Capacity

Capacity limits of ad-hoc 3D networks. Data rates per nodeNumber of nodes

Assumptions N users uniformly distributed over the

interior of a sphere. Each user communicates with another

user randomly chosen among all users. Signal power decays based on free

space path loss. All users transmit at the same power. No channel separation or diversity. Interference acts as additive white

Gaussian noise

Capacity Bounds

The total number of bits that may be transmitted by all users, per second, is approximately

Lower BoundBased on deterministic routing

scheme.

Upper BoundSimilar formulaUses convexity

C K N 3

Proportional to the cube root of N

Lower Bound Proof Sketch

Estimate the effects of interference in the limit of large N.

Construct a series of cell tessellations with useful properties.

Use the weak law of large numbers to prove the existence of one user in each cell.

Specify a routing and transmitting scheme using time sharing.

Determine the capacity of this scheme, which lower bounds the capacity of the best scheme.

What has changed since

1985?

Signal processing is better, cheaper, and lower power.

More powerful channel codes.

Multiuser detection and smart antennas.

Signal strength measuring techniques available in radios.

How would we leverage these developments to make better ad-hoc networks?

Sensor Networks

Sensor Networks Data highly correlated in time and space. Low homogeneous rates. Links typically asymmetric. Data flows to centralized location. Energy is the driving constraint. 1000-100,000 Nodes Have a common mission. Very different from typical ad-hoc

networks