Research Program Overview

Introduction

Wireless communications research has been given great impetus by the advent of cellular telephony, mobile satellite and portable personal communication services. The exponentially growing user demand for services, together with the increasing demands for higher speed transmission of large amounts of data, as well as customer requests for multi-media services, create the need for new technologies. In order to provide higher data transmission rates to more users without sacrificing the integrity of the received information, advances must be made in the transmission system designs and in the transmission system components.

The overall goal of the proposed research is higher capacity in broadband wireless communication systems at lower cost. We want to offer more wireless services with higher quality to more users. This is a significant challenge, scientifically and technologically, because the wireless spectrum (bandwidth) is a fixed resource. The primary thrust of this research is investigation into fundamental properties, limitations, and improvements in broadband wireless systems. Without new discoveries in engineering science, the available spectrum will become overloaded and unable to provide wireless services to the anticipated number of future users. A secondary thrust is the application of the research results to present and future systems. New engineering science must be put into action in the development of new technologies for real world systems. This two-pronged approach is consistent with the Chair's belief that strong fundamental research is vital to the understanding and improvement of technically challenging systems, while application of the fundamental research results is an important step in creating economic advantages for the supporting community.

Future wireless networks will allow people on the move to communicate with anyone, anywhere, and at any time using a range of multimedia services. Wireless communication will also enable a new class of intelligent home electronics that can interact with each other and with the Internet. Wireless video will support applications such as distance learning and remote medicine, and self configuring wireless networks will provide the baseline technology for widespread sensor networks and automated highways. These advances will require innovation at every stage from theory to implementation. Future wireless systems will require higher data rates with better coverage for a large number of users operating with a large variety of different systems. There are many technical challenges that must be overcome in order to make this vision a reality. In this process, wireless communication theory has a critical role in not only establishing a language and framework for thinking about solutions, but also in establishing fundamental limits to the capacities and quality of wireless communication systems.

The team is investigating a number of scientifically important and industrially relevant topics. A small number of "hot" topics under investigation in the iWCL are briefly described here to give an idea of the nature, relevancy and impact of the work.

Ultra-Wide Bandwidth (UWB) Communications Systems

UWB refers to new systems and technologies that are envisioned to provide short range, high data rate services to multiple users in an unlicensed transmission format. UWB systems are a hot research topic, attracting great scientific and industrial interest. These systems are unconvention-al in that they spread the information signal over an extremely wide bandwidth, occupying many gigaHertz of spectrum. Correspondingly, the signals have an extremely small power spectral density and appear as noise to existing users. As an emerging technology, UWB offers great potential as an area wherein one can make fundamental theoretical contributions and at the same time develop technology that can lead to socio-economic benefits for the supporting community.

Cooperative Diversity

Recently, researchers have considered utilizing the multiple protocol layers that exist in mobile wireless networks owing to the fact that there are many nodes (users), to solve some of the problems related to the physical implementation of multiple-element antenna arrays. In particular, a cross-layer network algorithm that has recently garnered considerable attention is cooperative diversity. Essentially, the idea of cooperative diversity is that multiple wireless devices benefit by relaying messages jointly as opposed to independently. Through cooperation, the broadcast nature of competing wireless signals, traditionally treated as interference, is used in a beneficial way. In fact, in certain wireless channels, cooperative diversity has been shown to provide full spatial diversity; that is, as if each individual transmitter had as many transmitter antennas as the entire set of cooperating transmitters. We are pursuing several projects related to cooperative diversity. We want to develop rigorous mathematical models that accurately model practical cooperative networks and that can be used to predict outage, coverage and error rate performance. These models will be invaluable for cooperative network design and optimization. The challenges in defining and solving realistic analytical models for cooperative networks are significant. In particular, the decentralization of the diversity operations and combining makes mathematical treatment complex.

Orthogonal Frequency Division Multiplexing (OFDM) Systems

Orthogonal frequency division multiplexing (OFDM) technology belongs to the class of multicarrier modulations where data are transmitted by modulating several parallel subchannels simultaneously. The application of OFDM modulation effectively converts a frequency-selective fading channel into several nearly flat-fading sub-channels reducing intersymbol interference caused by frequency-selective fading, and thereby maximizing data transmission capacity. The benefits of OFDM make it popular in today's broadband wireless communications industry, and new OFDM technologies are expected to be leading candidates for fourth generation wireless (4G) systems. For example, by utilizing OFDM in the physical layer, IEEE 802.11a is specified to achieve data rates up to 54 Mbps in the 5.2 GHz radio spectrum. This technology is also used in many systems proposed by the European Telecommunications Standard Institute (ETSI), such as, digital audio broadcasting (DAB), digital video broadcasting over terrestrial (DVB-T) and the HIPERLAN/2 standard. There is very strong and growing interest in using OFDM for the next generation of land mobile communication systems. One example is orthogonal frequency division multiple access (OFDMA), which implements multiple access by assigning distinct users to distinct subcarriers. Recently, OFDMA technology was adopted as a physical layer specification for 2-11 GHz broadband wireless access systems for metropolitan area networks in IEEE 802.16a.

Coding and Wireless Networks

Recent error-correcting codes, which allow for transmission at rates almost equal to the fundamental limits of data transmission on some channels, are under intense study for various communication setups. All these codes are associated with iterative decoders. Application of these codes to wireless channels, modern communication techniques such as OFDM, and in the presence of channel estimation error is quite challenging. Moreover, existing analyses of these codes are based on assuming an infinite block length. In practice, however, finite-length codes must be used. The objectives of our coding research are therefore (1) to find universal codes which can provide good performance without knowledge of the channel. Such solutions have immediate applications in time-varying wireless channels; (2) to efficiently apply modern coding techniques to OFDM and discrete multi-tone systems; and (3) to extend existing analyses of modern codes to finite-length codes.


Research Projects

Many scientifically important and industrially relevant results are being achieved in the research program. Attention is limited to some of the highlights.

Ultra-Wide Bandwidth Wireless Systems

UWB wireless is a very hot topic in university and industrial research and has great commercial potentials. As the iWCL continues research into UWB wireless systems, we have the following results to report.

UWB Receiver Designs for Multiuser Interference Environments

Conventional UWB receivers use the matched filter or correlator receiver structure. However, this structure is not optimal for multiuser interference environments, as exist when more than one UWB device is operating in geographical proximity. Meanwhile, it is expected that as many as twenty UWB devices may operate in close proximity in multiuser applications. The design of an optimal UWB receiver for multiuser interference environments is extremely complex, owing to the complicated signal distributions arising from the interference. We have discovered five new UWB receivers, each of which has superior performance to conventional UWB receivers. Patents have been filed by the University of Alberta on all five of these novel designs. All of these designs have practical value, each fulfilling a different performance/cost criterion. We have also determined the theoretical performance of an optimal, minimum symbol error rate UWB receiver, although the structure of such an optimal receiver remains elusive. The iCORE Chair was recently invited to give a Distinguished Lecturer Seminar on these topics at the Harvard University School of Engineering and Applied Science. This talk was well attended by wireless researchers from Harvard, MIT, and MERL (Mitsubishi Electric Research Laboratories).

Improved Rake Receivers for Impulse Radio

State-of-the-art receivers for UWB employ a Rake structure to "comb" the wireless transmission medium for multiple transmitted signal rays. We have proposed new Rake receiver structures that outperform conventional Rake receivers. The new designs are based on using our new UWB receiver structures in the "fingers" of the Rake receiver. In future work, we will design new, improved Rake UWB receivers by deriving a new combining strategy for the outputs of the fingers based on the statistics of the multiple access interference.

A Best Pulse Shape for Multiple Access UWB Systems

We have derived a best pulse shape for UWB systems that directly accounts for restrictions placed on the transmitted pulse spectrum by the regulatory spectral emission mask. Our design incorporates the effects of the transmitter pulse-shaping filter and the receiver filter at the same time. We have demonstrated that our pulse design optimizes the signal-to-noise-plus-interference ratio of the UWB system when corrupted by multiple access interference.

Symbol Synchronization in UWB Systems

All UWB systems transmit information using extremely short pulses, which are very difficult to track in wireless channels. Meanwhile, the UWB receiver must recover the timing of these pulses in order to process them. Poor synchronization will seriously deteriorate the receiver performance. We have started rigorous theoretical investigations into the principles of synchronization for UWB transmission systems, and the search for practical synchronization schemes for UWB receivers. Two different approaches being pursued are maximum likelihood estimation and feature-extraction-based timing recovery. In preliminary results, we have disproved a common belief that zero-crossing-based synchronizers can outperform maximum likelihood synchronizers in large signal-to-noise ratio environments.

Efficient Semi-Analytical Techniques for UWB Systems

Direct simulation of the widely used IEEE UWB channel model is very time consuming, particularly for oversampled (more than one sample per bit) system studies. We propose to find more efficient simulation techniques for UWB systems by combining some semi-analytical models with the IEEE UWB channel model. In particular, the IEEE UWB model incorporates a number of sums of fading random variables. Recent work we have published reports simple and highly accurate analytical approximations to sums of lognormal, Rayleigh and Ricean random variables. We will use these results to construct more efficient semi-analytical simulation models for UWB systems.

Space-Time Coding and MIMO Systems

Multiple input multiple output (MIMO) antenna systems can in some cases increase the capacity of a wireless network by a factor of ten. Space-time coding is a MIMO technique that spreads the transmitted signal across multiple transmitter antennas, using space (multiple transmitter antennas) as another dimension, in addition to time, to achieve huge capacity gains. iWCL space time codes have been licensed to an American WiMAX company, and adopted in the IEEE 802.16e industry standard, the WiMax Standard. This intellectual property, which generates revenue, is assigned to the university and is the subject of a patent pending.

In other MIMO research, we have developed reduced-complexity antenna combining schemes for MIMO receivers. These schemes are unique and noteworthy in that they perform equally well as existing reduced-complexity schemes that employ channel state estimation, but without requiring any channel estimation. Multiple international patents on this novel technology have been filed by the university.

Cooperative Wireless Networks

The use of multiple antennas can be an effective solution to combat the signal fading and shadowing that causes performance degradation in wireless systems. However, in many applications there may not be adequate room to accommodate multiple antennas spaced far enough apart to experience independent fading. This problem is further highlighted by the decreasing size of wireless communications devices. In cooperative wireless networks, users cooperate by receiving and retransmitting (relaying) the signals of other users. In this way, the cooperating users create a multiple antenna array. These innovative wireless networks aim to increase the channel capacity and reliability of wireless communications.

Building on results reported in last year's Annual Report, we have derived new mathematical solutions that can be used to predict in advance how often a cooperative wireless network that employs multiple antennas at the cooperating users (receiver diversity) will be in outage (that is the signal or call is dropped). These new solutions for outage probability indicate that cost-effective selection diversity can be deployed at the network nodes with little performance loss relative to expensive maximal ratio diversity combining when the decode-and-forward (DF) protocol is used, but not when the amplify-and-forward (AF) protocol is used.

In other related work, we have derived simple and explicit solutions for the outage and error rate of multi-hop and multi-branch cooperative wireless networks employing AF relaying. The solutions are general and lead to a test condition which dictates when a multi-hop system will outperform a single hop system in terms of outage and error rate performances under an appropriate power constraint. The effect of noise amplification of the relays has been investigated and quantified in detail. On the more theoretical side, we have found exact, explicit solutions for the ergodic capacities of both DF and AF wireless relaying systems.

In other work on wireless relaying networks, we have proposed and studied two new protocols for rateless coded opportunistic relaying. The two new protocols have been compared to each other and to three previously known protocols in a power-fair regime using newly formulated fairness criteria for the power constraints. The novel proposed protocols are based on the water-filling principle and power optimization and take advantage of the opportunistic relaying concept. They have been shown to outperform the known protocols.

Wireless Systems with Correlated Antennas

It is well known that adding independent, additional receiver antennas will yield substantial improvements to the performance of wireless systems. Much recent research has further shown that adding correlated, additional antennas will also give worthwhile performance gains, although the gains are less than those achieved for independent antennas. Clearly, there must be a limit to how many additional, correlated antennas can be added to a wireless receiver if one requires that each additional antenna must further improve the system performance by a worthwhile amount, and that all the antennas must occupy a fixed, finite volume of space.

Fundamental Limits

In research published in the reporting year, we have shown that there is a maximum number of receiver antennas that should be deployed in a fixed volume of space. Beyond this number, little or no improvement is achieved, and performance may in fact deteriorate. We have determined a simple engineering "rule of thumb" that will be of great utility to wireless system designers. The results have been interpreted in terms of sampling the cochannel-interference-plus-noise space.

Decorrelator for Practical Dual Antenna Systems

We have discovered a decorrelating pre-processor that takes as input two signals from two, correlated antennas and gives as output two independent signals. The pre-processor circuitry is remarkably simple and the pre-processor works for many fading conditions and arbitrary correlation between antennas with no or minor adjustment. The pre-processor can provide as much as 2.1 dB gain in some applications (a 2.1 dB gain is equivalent to a 62% increase in transmitter power). A patent application on the pre-processor has been filed by the university.

Indoor Position Location based on UWB Signalling

The primary objective is to investigate precision indoor position based on UWB signalling. Several time synchronized receivers located throughout the building premise will simultaneously demodulate the UWB signal emanating from a mobile transmitter. The superior multipath resolution capabilities afforded by the large UWB bandwidth coupled with the synthetic array processing of the moving mobile provides significant diverse observation data for position estimation and tracking. An initial goal of this research is to develop implementable processing algorithms for mobile position estimation and tracking. Theoretical performance assessment of these algorithms will be made followed by experimental verification.