EE 359: Wireless Communications Advanced Topics in Wireless.

EE 359: Wireless Communications

Advanced Topics in Wireless

Topics

Future wireless networksSoftware-defined and low-

complexity radiosAdvanced design of cellular

systemsWireless network convergence

and SDNAd-hoc and sensor networksCognitive and software-defined

radiosEnergy Constraints in Wireless

SystemsControl over WirelessNeuroscience applications of

wireless

EE360

EE360

EE360

EE360

Challenges

Network ChallengesScarce spectrumDemanding applicationsReliabilityUbiquitous coverageSeamless indoor/outdoor operation

Device ChallengesSize, Power, CostMIMO in SiliconMultiradio Integration Coexistance

Cellular

AppsProcessor

BT

MediaProcessor

GPS

WLAN

Wimax

DVB-H

FM/XM

Software-Defined Radio

Multiband antennas and wideband A/Ds span BW of desired signals

The DSP is programmed to process the desired signal based on carrier frequency, signal shape, etc.

Avoids specialized hardware

Cellular

AppsProcessor

BT

MediaProcessor

GPS

WLAN

Wimax

DVB-H

FM/XM A/D

A/D

DSPA/D

A/D

Compressed sensing may be a solution for sparse signals

Today, this is not cost, size, or power efficient

Compressed Sensing Basic premise is that signals with some

sparse structure can be sampled below their Nyquist rate

Signal can be perfectly reconstructed from these samples by exploiting signal sparsityThis significantly reduces the burden on the

front-end A/D converter, as well as the DSP and storage

Might be key enabler for SD and low-energy radiosOnly for incoming signals “sparse” in time,

freq., space, etc.

Reduced-Dimension Communication System Design

Compressed sensing ideas have found widespread application in signal processing and other areas.

Basic premise of CS: exploit sparsity to approximate a high-dimensional system/signal in a few dimensions.

Can sparsity be exploited to reduce the complexity of communication system design?

Sparsity: where art thou?

Sparse state space: e.g reduced-dimension network control

Sparse users: e.g. reduced-dimension multiuser detection

To exploit sparsity, we need to find communication systems where it exists

Sparse samples: e.g. sub-Nyquist sampling

Sparse signals: e.g. white-space detection

Sparse Samples

For a given sampling mechanism (i.e. a “new” channel) What is the optimal input signal? What is the tradeoff between capacity and sampling

rate? What known sampling methods lead to highest capacity?

What is the optimal sampling mechanism? Among all possible (known and unknown) sampling

schemes

h(t)

SamplingMechanism

(rate fs)

New Channel

Capacity under Sub-Nyquist SamplingTheorem 1:

Theorem 2:Bank of Modulator+FilterSingle Branch Filter Bank

Theorem 3:

Optimal among all time-preserving nonuniform sampling techniques of rate fs (ISIT’12; Arxiv)

zzzzzzzzzz

)(ts ][ny

)(1 ts

)(tsi

)(tsm

)( smTnt

)( smTnt

)( smTnt

][1 ny

][nyi

][nym

equals

zzzzzzzzzz

Selects the m branches with m highest SNRExample (Bank of 2 branches)

Joint Optimization of Input and Filter Bank

highest SNR

2nd highest SNR

low SNR

skffX 2

fX

skffX

skffX

)( skffH

)( fH

)( skffH

)( skffN

)( fN

)( skffN

)( skffS

)( fS

)( skffS

)2( skffH

)2( skffN )2( skffS

low SNR

fY1

fY2

Capacity monotonic in fs

How does this translate to practical modulation and coding schemes

Reduced-Dimension Network Design

Random State Evolution

Samplingand

Inference

StochasticControl

Resource Management

Reduced-Dimension

Representation

Scarce Wireless Spectrum

and Expensive

$$$

Spectral ReuseDue to its scarcity, spectrum

is reused

BS

In licensed bands

Cellular, Wimax Wifi, BT, UWB,…

and unlicensed bands

Reuse introduces interference

Careful what you wish for…

Growth in mobile data, massive spectrum deficit and stagnant revenues require technical and political breakthroughs for ongoing success of cellular

“Sorry, America: Your wireless airwaves are full”CNNMoneyTech – Feb. 2012

The “Spectrum Crunch”

Are we at the Shannon limit of the Physical Layer?

Time-varying channels with memory/feedback.

We don’t know the Shannon capacity of most wireless channels

Channels with interference or relays.Uplink and downlink channels with

frequency reuse, i.e. cellular systems.Channels with delay/energy/$$$ constraints.

Rethinking “Cells” in Cellular

Traditional cellular design “interference-limited”MIMO/multiuser detection can remove interferenceCooperating BSs form a MIMO array: what is a cell?Relays change cell shape and boundariesDistributed antennas move BS towards cell boundarySmall cells create a cell within a cellMobile cooperation via relaying, virtual MIMO, analog network

coding.

SmallCell

Relay

DAS

Coop MIMO

How should cellularsystems be designed?

Will gains in practice bebig or incremental; incapacity or coverage?

Are small cells the solution to increase cellular system capacity?

Yes, with reuse one and adaptive techniques (Alouini/Goldsmith 1999)

A=.25D2p Area Spectral Efficiency

S/I increases with reuse distance (increases link capacity).Tradeoff between reuse distance and link spectral efficiency (bps/Hz).Area Spectral Efficiency: Ae=SRi/(.25D2p) bps/Hz/Km2.

The Future Cellular Network: Hierarchical Architecture

Future systems require Self-Organization (SON) and WiFi Offload

10x Lower COST/Mbps

10x CAPACITY Improvement

Near 100%COVERAGE

(more with WiFi Offload)

MACRO: solving initial coverage issue, existing networkPICO: solving street, enterprise & home coverage/capacity issue

Macrocell Picocell Femtocell

Today’s architecture• 3M Macrocells serving 5 billion users• Anticipated 1M small cells per year

SON Premise and Architecture

Node Installation

Initial Measurements

Self Optimization

SelfHealing

Self Configurati

on

Measurement

SON Server

SoNServer

Macrocell BS

Mobile GatewayOr Cloud

Small cell BS

X2

X2X2

X2

IP Network

SWAgent

SON is part of 3GPP/LTE standard

LTE.11

Convergence of Cellular and WiFi

- Seamless handoff between networks- Load-balancing of air interface and backbone- Carrier-grade performance on both networks

Wireless networks are everywhere, yet…

- Connectivity is fragmented- Capacity is limited (spectrum crunch and interference)- Roaming between networks is ad hoc

TV White Space &Cognitive Radio

SDWN Architecture

WiFi APFemto Cell

Pico Cell Cognitive Radio

Freq.Allocation

PowerContr

ol

SelfHealing

ICIC QoSOpt.

CSThreshol

d

UNIFIED CONTROL PLANE

Commodity HW

SW layer

App layerVideo SecurityVehicularNetworks HealthM2M

Cognitive Radio Paradigms

UnderlayCognitive radios constrained to

cause minimal interference to noncognitive radios

InterweaveCognitive radios find and exploit

spectral holes to avoid interfering with noncognitive radios

OverlayCognitive radios overhear and

enhance noncognitive radio transmissions

Knowledge

andComplexi

ty

Underlay Systems Cognitive radios determine the

interference their transmission causes to noncognitive nodesTransmit if interference below a given

threshold

The interference constraint may be metVia wideband signalling to maintain

interference below the noise floor (spread spectrum or UWB)

Via multiple antennas and beamforming

NCR

IP

NCRCR CR

Interweave Systems Measurements indicate that even

crowded spectrum is not used across all time, space, and frequenciesOriginal motivation for “cognitive” radios

(Mitola’00)

These holes can be used for communicationInterweave CRs periodically monitor

spectrum for holesHole location must be agreed upon between

TX and RXHole is then used for opportunistic

communication with minimal interference to noncognitive users

Overlay Systems

Cognitive user has knowledge of other user’s message and/or encoding strategyUsed to help noncognitive

transmissionUsed to presubtract noncognitive

interferenceRX1

RX2NCR

CR

Performance Gains from Cognitive Encoding

Only the CRtransmits

outer bound

our schemeprior schemes

Ad-Hoc Networks

Peer-to-peer communications. No backbone infrastructure. Routing can be multihop. Topology is dynamic. Fully connected with different link

SINRs

Cooperation in Wireless Networks

Many possible cooperation strategies:Virtual MIMO, relaying (DF, CF, AF), one-

shot/iterative conferencing, and network coding

Nodes can use orthogonal or non-orthogonal channels.

Many practice and theoretical challengesNew full duplex relays can be exploited

Design Issues

Ad-hoc networks provide a flexible network infrastructure for many emerging applications.

The capacity of such networks is generally unknown.

Transmission, access, and routing strategies for ad-hoc networks are generally ad-hoc.

Crosslayer design critical and very challenging.

Energy constraints impose interesting design tradeoffs for communication and networking.

General Relay Strategies

Can forward message and/or interference Relay can forward all or part of the

messages Much room for innovation

Relay can forward interference To help subtract it out

TX1

TX2

relay

RX2

RX1X1

X2

Y3=X1+X2+Z3

Y4=X1+X2+X3+Z4

Y5=X1+X2+X3+Z5

X3= f(Y3)

Beneficial to forward bothinterference and message

• For large powers, this strategy approaches capacity

Wireless Sensor NetworksData Collection and Distributed Control

Energy (transmit and processing) is the driving constraint

Data flows to centralized location (joint compression) Low per-node rates but tens to thousands of nodes Intelligence is in the network rather than in the

devices

• Smart homes/buildings• Smart structures• Search and rescue• Homeland security• Event detection• Battlefield surveillance

Energy-Constrained Nodes

Each node can only send a finite number of bits.Transmit energy minimized by maximizing

bit timeCircuit energy consumption increases with

bit timeIntroduces a delay versus energy tradeoff

for each bit Short-range networks must consider

transmit, circuit, and processing energy.Sophisticated techniques not necessarily

energy-efficient. Sleep modes save energy but complicate

networking.

Changes everything about the network design:Bit allocation must be optimized across all

protocols.Delay vs. throughput vs. node/network

lifetime tradeoffs.Optimization of node cooperation.

Channel Coding Channel coding used to reduce error probability

Metric is typically the coding gain Also measured against Shannon limit

Block and Convolutional Codes Induce bandwidth expansion, reduce data rate Easy to implement; well understood (decades of

research) Coded Modulation (Trellis/Lattice Codes; BICM)

Joint design of code and modulation mapper Provides coding gain without bandwidth expansion Can be directly applied to adaptive modulation

Turbo Codes (within a fraction of a dB of Shannon limit) Clever encoding produces pseudo “random” codes easily Decoder fairly complex

Low-density parity check (LDPC) Codes State-of-the-art codes (802.11n, LTE) Invented by Gallager in 1950s, reinvented in 1990s Low density parity check matrix Encoder complex; decoder uses belief propagation

techniques

Coded

Uncoded

Coding Gain

SNR

Pb

ShannonLimit

Green Codes (for short distances)

Is Shannon-capacity still a good metric for system design?

Coding for minimum total power

Is Shannon-capacity still a good metric for system design?

Extends early work of El Gamal et. al.’84 and Thompson’80

ComputationalNodes On-chip

interconnects

Fundamental area-time-performance tradeoffs

For encoding/decoding “good” codes,

Stay away from capacity!Close to capacity we have

Large chip-area, more time/power

Area occupied by wires Encoding/decoding clock cycles

Regular LDPCs closer to bound than capacity-approaching LDPCs!Need novel code designs with short wires, good performance

Green” Cellular Networks

Minimize energy at both the mobile and base station via New Infrastuctures: cell size, BS placement,

DAS, Picos, relays New Protocols: Cell Zooming, Coop MIMO,

RRM, Scheduling, Sleeping, Relaying Low-Power (Green) Radios: Radio

Architectures, Modulation, coding, MIMO

Pico/Femto

Relay

DAS

Coop MIMO

How should cellularsystems be redesignedfor minimum energy?

Research indicates thatsignificant savings is possible

Why Green, why now? The energy consumption of cellular networks is

growing rapidly with increasing data rates and numbers of users

Operators are experiencing increasing and volatile costs of energy to run their networks

There is a push for “green” innovation in most sectors of information and communication technology (ICT)

There is a wave of consortia and government programs focused on green wireless

Distributed Control over Wireless

Interdisciplinary design approach

• Control requires fast, accurate, and reliable feedback.

• Wireless networks introduce delay and loss

• Need reliable networks and robust controllers

• Mostly open problems

Automated Vehicles - Cars - Airplanes/UAVs - Insect flyers

: Many design challenges

The Smart Grid:Fusion of Sensing, Control, Communications

carbonmetrics.eu

Much progress in some areas

Smart meteringGreen buildings and structuresOptimization Demand responseModeling and simulationIncentives and economics

But many of the hard researchquestions have not yet been asked

Current work: sensor placement, fault Detection and control with sparse sensors

Applications in Health, Biomedicine and

Neuroscience

Recovery fromNerve Damage

Doctor-on-a-chip

WirelessNetwork

Neuro/Bioscience- EKG signal

reception/modeling- Brain information theory- Nerve network

(re)configuration- Implants to

monitor/generate signals- In-brain sensor networks

Body-Area Networks

scan

Gene Expression Profiling

tumor tissue

RNA extraction labeling hybridization

70 genes

Gene expression profiling predicts clinical outcome of breast cancer (Van’t Veer et al., Nature 2002.) Immune cell infiltration into tumors good

prognosis.

Gene expression measurements: a mix of many cell types

Gene Signatures

Cell-typeProportion

Cell Types

Looks like CDMA “despreading”

• Many gene expression deconvolution algorithms exist Shen-Orr et al., Nature methods 2010 known “C” and “k” Lu et al. , PNAS 2003 known “G” and “k” Vennet et al., Bioinformatics 2001 known “k”

Large databases exist where these parameters are unknown Can we apply signal processing methods to blindly separate

gene expression?We adapt techniques from hyperspectroscopy (Piper et

al, AMOS 2004) assuming “C”, “G” and “k” unknown

Beat existing techniques, even nonblind ones

B A

C D

E

𝐼 ( 𝑋𝑛 →𝑌 𝑛)=𝐻 (𝑌 𝑛) −∑𝑖=1

𝑛

𝐻 (𝑌 𝑖∨𝑌 𝑖− 1 , 𝑋 𝑖)

Direct information (DI) inference

𝐼 ( 𝑋𝑛 →𝑌 𝑛)=𝐻 (𝑌 𝑛) −∑𝑖=1

𝑛

𝐻 (𝑌 𝑖∨𝑌 𝑖− 1 , 𝑋 𝑖− 𝐷− 𝑁𝑖− 𝐷 )

Constrained DI inference

B A

C D

E

Neuron layoutB A

C D

E

𝐼 ( 𝑋𝑛 →𝑌 𝑛)=𝐻 (𝑌 𝑛) −∑𝑖=1

𝑛

𝐻 (𝑌 𝑖∨𝑌 𝑖− 1 , 𝑋 𝑖− 𝐷 )

DI inference with delay lower bound

B A

C D

E

Pathways through the brain

Other Applications of Communications and Signal Processing to Brain Science

EpilepsyEpileptic fits caused by an

oscillating signal moving from one region to another.

When enough regions oscillate, a fit occurs

Working with epilepsy expert (MD) to understand how signal travels between regions.

Has sensors directly implanted in brain: can read signals and inject them.

Can we use drugs to block propagation or signal injection to cancel signals

ParkinsonsCreates 20 MHz noise in brain

region120 MHz square wave injection

mitigates the symptoms

The EndThanks!!!Have a great winter break

Unless you are studying for quals – if so, good luck!