Sky DSP logo

Ph.D. Thesis


Home
Publications
Links
Glossary


Adaptive Techniques for Multiuser OFDM

Print Friendly Version in PDF (4.8 MB), 327pages

Thesis submitted by Eric Lawrey in Dec 2001 in partial fulfilment of the requirements for the degree of Doctor of Philosophy, School of Engineering at James Cook University.

Abstract

The demand for high-speed mobile wireless communications is rapidly growing. OFDM technology promises to be a key technique for achieving the high data capacity and spectral efficiency requirements for wireless communication systems of the near future.

This thesis presents an investigation into methods for maximising the spectral efficiency of Orthogonal Frequency Division Multiplexing (OFDM) systems. As part of this, an investigation of detrimental effects on OFDM is presented, showing the effect of: band pass filtering, the use of a raised cosine guard period, clipping distortion, Additive White Gaussian Noise (AWGN) on modulation BER rate, time synchronisation error, and frequency offset errors.

An investigation of two adaptive techniques is also presented. These techniques utilise knowledge obtained by dynamically tracking the radio channel response, to optimise the user frequency, and subcarrier modulation. Adaptive modulation independently optimises the modulation scheme applied to each subcarrier so that the spectral efficiency is maximised, while maintaining a target Bit Error Rate (BER). For a fading channel, adaptive modulation results in an improvement of 12 - 16 dB in the Signal to Noise Ratio (SNR) required to maintain a given BER, as compared with fixed modulation. Adaptive user allocation exploits the difference in frequency selective fading between users, to optimise user subcarrier allocation. In a multipath environment the fading experienced on each subcarrier varies from user to user, thus by utilising user/subcarrier combinations that suffer the least fading, the overall performance is maximised. Adaptive user allocation results in an additional average signal power improvement of 3 - 5 dB.

The simulated performance of the adaptive techniques is presented using a set of measured wide bandwidth (70 MHz) frequency response measurements taken at 1 GHz. These measurements show the changes in the frequency selective fading with small increments in space, allowing the effects of different tracking rates for the adaptive allocations schemes to be investigated.

This thesis also presents a method for maximising the signal strength within buildings, by using transmission repeaters. This is similar to the Single Frequency Networks used in DAB and DVB systems, except applied to small-scale bi-directional communications. Using multiple repeaters causes multipath problems in most conventional systems, however OFDM has a sufficiently high multipath tolerance to combine the multipath energy. It was found to decrease the path loss by 7 dB for an indoor system with two repeaters, and up to 20 dB for eight repeaters.

In addition, two techniques are presented for reducing the Crest Factor (peak to average power ratio of the RF signal envelope) of OFDM signals. The first technique is a phasing scheme for OFDM pilot symbols, which uses genetic algorithms to optimise the phase angle of each subcarrier. This technique achieves a lower CF than any previously published techniques, obtaining a CF as low as 0.65 dB, which is 2 dB lower than commonly used techniques. The second technique reduces the CF of data carrying symbols, by including additional subcarriers that are optimised in amplitude and phase to cancel out the peaks in the overall OFDM symbol. This was found to produce a net improvement of 4 dB to the worst-case symbol CF.

Code

The matlab code developed using version 5.3 and is provided as is with no support. The code is split into two parts depending on what area you are interested in.

  • Part 1: (2 MB) Main code, everything except diagrams in Chapter 4 Multiuser OFDM
  • Part 2: (7.4 MB) Code for Chapter 4 including fading measurement data.
Copyright 2009 Eric Lawrey, Last updated: Jan 6, 2009