1.

Wireless Technologies

– A case study for IEEE 802.11g OFDM system

Arpan Pal

Center of Excellence for Embedded Systems

TCS, Kolkataarpan_pal@tcscal.co.in

2.

Introduction

Cellular Wireless Systems & Wireless Networks

802.11g OFDM PHY developmentSystem DescriptionDevelopment FrameworkSimulation IssuesReceiver AlgorithmsImplementation IssuesSimulation ResultsSecurity Algorithms

Areas of future - Convergence to 4G

Agenda

3.

Wireless is the next giant leap in information services. The new paradigm for connectivity enables business to operate

• faster• better• more cost effectively• and more profitably

through the use of • always on, • always connected, and • always available content and applications.

With the tremendous increase in wireless LANs, Mobile phones, PDAs, and other mobile devices, the merging of computation and telecommunication technologies is a fundamental part of modern society.

Can broadly be classified into two types – Cellular Wireless Systems and Wireless Networks

Introduction

4.

Cellular Wireless Systems Roadmap

9.6 k

64 k

384 k

1000 k

2000 k

5000 k

1995 2000 2005

AMPSTACS NMT

IS136GSM

IS95A

IS95BGSM-GPRS

1980

CDMA2000W-CDMA

4G3G2.5G2G1G

???

5.

Wireless Networks PAN – IEEE 802.15

LAN – IEEE 802.11

MAN – IEEE 802.16

6.

OFDM Overview

• Multi-Carrier Modulation Technique

• Carrier spacing kept minimum maintaining orthogonality

• Multi-path robustness due to multi-carrier

• Reduction of ISI through Guard Bands

• Robust against narrow-band interference

• Efficient FFT based receiver structures

• Simpler Frequency Domain Channel Equalization

• Simpler Receiver Synchronization Techniques

• High peak-to-average power levels

• Susceptible to RF Front-end non-linearity

• Susceptible to LO frequency offset / drift

7.

802.11g OFDM PHY

Assemble frame

Scrambler

Convolution

Block Interleaver

Bit Mapper IFFT Add Guard Interval

Window

MAC Layer

DA

C

RF Transmitter

Transmit

Remove Guard

Interval

FFT Channel / PhaseCorrection

De-mapper

De-interleaver Viterbi Decoder Descrambler Disassemble

Frame

Channel Estimator

Frequency/Phase Correction

MAC Layer

AD

C

RF Receiver

AGC

AFC

Receive

Management Entity

PLCP

PMD

PMD

PLCP

FrameSync &Coarse Frequ.Correct

Receiver Sync

PLME

Preamble & Pilot Insertion

8.

Development Framework

Simulation Environment

Fixed point Model

Floating point Model

MATLAB model

Analysis

Result

C System level model

Analysis

Result

VHDL / Verilog

Simulation

Synthesis

Rest of the process

9.

Simulation Issues

• Proper selection of channel models with various delay-spreads and Doppler shifts.

• Proper simulation of Sampling Clock error.

• Proper simulations of Phase and Frequency Error.

• Proper modeling of Phase Noise.

• Proper modeling of I-Q imbalance.

• AGC

• Proper modeling of the LNA and anti-aliasing filter.

• AWGN noise.

10.

Receiver Algorithms• Time Synchronization

Packet Detection

Energy Based

Frame Synchronization

Short Training Sequence Cross-Correlation Based

False Alarm Reduction

Short Training Sequence Auto-Correlation Based

11.

Receiver Algorithms• Frequency Synchronization

Frequency Offset Estimation

Coarse estimate based on Short Training Sequence Cross-Correlation

Fine estimate based on Long Training Sequence Cross-Correlation

Frequency Offset Correction

Time domain (pre-FFT) rotation based on estimated offset

Carrier Phase Tracking

Phase offset due to residual frequency offset and sampling clock error

Pilot based estimation for phase

Each OFDM symbol contains 4 pilots

Frequency domain (post FFT) rotation for phase correction

12.

Receiver Algorithms

• Channel Estimation

Assumes quasi-stationary channels (does not change within a packet)

Channel Transfer function estimated from long training sequence (LTS)

Estimated Channel Transfer Function

= FFT(Received LTS) / FFT(ideal LTS)

Takes care of indoor channels along with gain variation

Channel compensation done post FFT by dividing with estimated Channel Transfer Function

13.

Implementation Issues

• Use of Radix-22 FFT /IFFT algorithm instead of a Radix-2 or a Radix-4 implementation.

• Use of CORDIC (COordinate Rotation DIgital Calculation) for performing complex multiplication and division.

• Determining the scaling factor to be used after every stage of the FFT block.

• Deciding on the number of iterations to be used for implementing CORDIC.

• Approximating all sqrt(x2 + y2) with (|x| + |y|) for hardware simplification and altering the various threshold values accordingly.

• LUT implementation of various mathematical calculations.

• Deciding on the number of bits to be used for ADC

• Fixing the number of bits to be used for implementing FFT/IFFT.

14.

Simulation Results2

3

4

5

15.

Simulation Results

Channel Model 3 used

16.

Security Algorithms

• “An Alternative Approach for Enhancing Security of Wireless Networks using Physical Layer Encryption” – patent filed

• Provides enhanced security against- Data Privacy- Data Forgery- Denial of Service

• Two KEY Security method (secure KEY delivery assumed)• KEYS used to encrypt/modify physical layer parameters like

- Error Control Coding Rate- Type of Modulation / Constellation Mapping- Length of Packet- Interleaving Pattern- Phase offset

• Contribution presented in IEEE 802.20 Standard body meeting in September, 2003

17.

Areas of future – Convergence to 4G

• 4G is whatever that is beyond 3G

• To be used for Real-Time Video Delivery and similar applications

• QOS is important

• Calls for more efficient Modulation, ECC and Equalization

• Incorporation of Mobility and stringent Multipath including non- LOS scenario into Wireless Networks can lead to 4G

• OFDM and UWB likely candidates for 4G PHY

• Space-Time Diversity, Smart Antenna Processing and Multiple-Input-Multiple-Output (MIMO) systems are also likely to be used

• CDMA with its multi-user capabilities can provide the access mechanism

• Calls for implementation of a MAC layer that seamlessly integrates all the above features of PHY