Advanced Wireless Communication Systems and Signal Processing

Professor Pentti Leppänen, Telecommunication Laboratory,
Department of Electrical Engineering, University of Oulu
Professor Matti Latva-aho, Director of the Centre for Wireless Communications
Research Professor Aarne Mämmelä and Research Professor Petri Mähönen, VTT Electronics
pentti.leppanenee.oulu.fi, matti.latva-ahoee.oulu.fi,
aarne.mammelavtt.fi, petri.mahonenee.oulu.fi
http://www.infotech.oulu.fi/awc

Background and Mission

The demand for high-quality wireless communication services exceeded all expectations in the beginning of the 1990's. The demand will continue to grow in the new century. In addition to wireless telephony, wireless data and video transmission will also be required, which clearly will increase the data rate to several megabits per second. In wireless local area networks, the expected data rate is more than 100 Mbit/s. The research activities of the "Advanced Wireless Communication Systems and Signal Processing" (AWC) group are focused on advanced wireless communication systems and on the advanced signal processing methods and architectures needed in the implementation of the above-mentioned systems. The group consists of 4 sub-groups:

1. The group at the University of Oulu's Telecommunication Laboratory, directed by Professor Pentti Leppänen (TL),

2. The group at the University of Oulu's Centre for Wireless Communications, directed by Professor Matti Latva-aho (CWC),

3. The group at VTT Electronics, directed by Research Professor Aarne Mämmelä (VTT),

4. The Internet group at VTT Electronics, directed by Research Professor Petri Mähönen (VTT).

The activities of the sub-groups together cover completely the lower layers of a generic communication system model, from the network and transmission system level down to the micro-architecture and logic gate levels.

The main goals of the group are

• To support the development of the third generation mobile phone system (UMTS/IMT-2000), and future wireless local area networks and other systems to be used for fast wireless data and multimedia transmission.

• To analyze and design communication networks and protocols.

• To develop channel measurement systems and channel models based on measurements.

• To optimize communication systems, methods and algorithms based on the developed channel models. The channel models, and therefore systems, are developed for various different channels, for example, channels with intentional interference (military applications), channels with multiple users, and indoor-outdoor static channels.

• To develop floating-point and bit-true models of receiver algorithms for different systems. The models include, for example, multiuser detectors, channel decoders, synchronization algorithms, and cancellation of intentional interference.

• To develop VHDL architecture models for the most important algorithms.

• To develop experimental systems to verify the theoretical results.

• To produce international publications, doctoral theses and licentiate theses.

The groups at the University of Oulu (TL and CWC) focus on system and algorithm level studies of wireless communication networks and methods. The interest is in broadband transmission schemes (spread-spectrum, CDMA and multicarrier techniques), and their application to future wireless communication systems, as well as implementation of such systems by modern digital signal processing methods. Potential application areas are, for example, 3^rd and 4^th generation cellular systems, military communication on VHF/UHF bands and indoor wireless communications, especially wireless local area networks (WLANs).

The mission of the CWC and TL groups is scientific research and development of wireless communication methods and applications, and the transfer of the knowledge gained for use in industry, defence forces and society. CWC focuses on commercial and civilian applications, whereas TL specializes in military systems. The groups work in close co-operation, sharing the information and know-how relevant to both groups.

The main effort of the VTT group is in the research and development of algorithms and implementation architectures for digital transmitters and receivers that are both typically adaptive. The group's research and the development knowledge has in the past been applied in several projects, in which digital receiver algorithms and their architectures have been designed for mobile phones and base stations, a fast wireless WLAN system, a power line DS spread-spectrum system, and a fast microwave radio.

A great deal of emphasis is put on the system design in telecommunication systems. For example, the design of implementation architectures is performed jointly with the design of algorithms. The work is focused on architectures suitable for ASIC implementation. For a given set of algorithms, the most promising architectures are selected and modelled using VHDL language. VHDL architecture models are synthesized to gate level descriptions from which accurate estimates of the complexity (in number of gates) and the throughput (speed) of the ASIC can be achieved.

Scientific Progress

ABRAS is a project lasting several years studying the implementation of advanced receiver algorithms for the WCDMA system. These receivers, suitable to the downlink environment, will be developed and implemented in the Texas Instruments C6x DSP environment.

The aim of the AWICS project is to develop advanced receiver algorithms and coding/decoding schemes (combined with new diversity techniques) for future wireless communication systems. Emphasis will be on the development of general theory and techniques that can be used in various applications. A particular emphasis will be on issues related to counteracting the detrimental effects caused by fading and the various sources of interference. The main contribution is to utilize the 2-dimensional channel, i.e., spatial signal processing is applied, both in CDMA and multicarrier applications. The third and fourth generation mobile communication systems, as well as new military communication networks, will be the main applications.

DOCENT is a project to study methods of increasing the capacity of WCDMA downlink connection. The project´s main area of interest is the Transmit Diversity techniques already proposed within the 3GPP standardisation organisation. Special attention will also be given to the models showing how the WCDMA system basic mechanisms (transmit power control, data rate, user distribution and transceiver structure) contribute to the total capacity of a WCDMA system. The research methods include radio access network and physical layer simulations.

The FIXWIRE project investigates and develops novel transmission and transceiver technologies for future fixed and low-mobility broadband wireless communication systems. The main focus of the project is on space-time processing, adaptiveness, multicarrier modulation and ad-hoc radio networks. The project will last for several years.

FUBS project (commencing 1999) is focused on licence free, ultra wideband (UWB) communication systems. The aim of this project is to study different technologies and system concepts to exploit UWB signals. The idea of licence free systems is based on extremely large bandwidth and very low power spectral density of the transmitted signal. FUBS is planned last until the summer of 2003.

The KOMA project is a joint project between CWC and VTT (Electronics Laboratory). The role of CWC has been mainly in defining multiuser interference cancellation algorithms for a WCDMA system. The transmitter _ receiver was implemented by VTT on a Pentek multiprocessor DSP card. During 1999, the TX-RX link was completely tested and used to perform base station receiver simulations in a WCDMA system. The KOMA project ends in 2001.

The MURSU project studies the algorithm optimization and implementation of multiuser receivers for the WCDMA system uplink.

WINNER is direct continuation from WINGIP. It is a 2 year joint project between the Centre for Wireless Communications at the University of Oulu and Wireless Internet Laboratory at VTT Electronics. The project is developing wireless packet-based communication in hybrid wireless networks. Usage of IPv6 in heterogeneous wireless networks will be in a specific focus. The project educates researchers for this field and IP-networking overall. Another objective is to contribute to the IETF standardization from the areas of research. The testbed will be an important tool for evaluating research results.

Software defined radio (SDR) has been recognized as a credible option for current and future radio systems. In the commercial sector, software radio provides a future-proof platform, which can be used to implement current second generation cellular systems as well as evolving third generation systems with their possible new features and services.

Software radio development aims at wideband RF access and software partitioning for plug-and-play type of use. There are many problems in implementing software radio. These problems include high quality wideband RF access and high performance analog to digital and digital to analog conversion due to the extreme demands placed upon them. Because of the lack of suitable methods for performance evaluation it is also difficult to estimate the signal processing requirement for a software radio, when implementing several systems in the same platform. There are also no methods for estimating the processing capacity of a software radio platform consisting of multiple DSP processors, CPUs, memory modules and low speed/high speed buses. On top of this, there are no universal design tools that could be used all the way from system level specification to actual implementation, but instead different CAD-tools with different kinds of options must be used in the design process. Due to finite processing capacity, novel and computationally efficient DSP-architectures must be defined for the time critical functions of the SDR.

The aim of this research work is to study new methods to design and estimate requirements of a software defined multi-band multi-mode (MBMM) radio. Different architectures and technical challenges for a software defined radio have been studied. Methods and simulation environments to define the impact of nonlinearities and A/D-conversion have been developed. Performance requirements for MBMM-radio have been estimated, and methods for hardware capacity estimation have been studied. These research results and simulation environments have been and will still be used in defining a suitable architecture for flexible and reconfigurable software defined radio systems.

The goal of Future Military Radio Communications projects in the TL group is to apply ideas and results based on intensive research made for next generation wireless communications to future military communication systems, like tactical systems, field radio links, wireless LANs, radiolocation systems. The applicability of the software defined radio concept is also under study.

New results for future tactical military communication systems based on spread spectrum methods have been reported. Different architectures and technical challenges for a software defined radio have been studied. Methods and simulation environments to define the impact of nonlinearities and A/D-conversion have been developed. Performance requirements for a multi-band multi-mode (MBMM) radio has been estimated and methods for hardware capacity estimation have been studied.

Preliminary study has been done for a positioning system which is robust against jamming, and against other radio signals. This includes a study of present positioning systems like GPS and GLONASS and their immunity against interference. It also includes studies for methods that can be used to realize a robust positioning system.

In the VTT group, a strategic, budget-funded program (1999-2001) referred to as Telecommunication Electronics (TELEC) is going on within the Electronic Systems (ELS) field at VTT Electronics. Physical layer algorithms, architectures and the design of future wireline and wireless wideband access to information service providers are considered. We use orthogonal frequency division multiplexing (OFDM) as a common system model for our research. During the second year (2000) of the TELEC program, a group of four doctoral students was working for TELEC. Our seminar system was significantly improved. Two doctoral theses are in preparation on channel estimation and synchronization algorithms. We use the expectation maximization (EM) algorithm in symbol and carrier phase synchronization, and a matched filter bank for channel estimation. The results will be published soon in review and original papers. In one Master's thesis, adaptive transmission was considered. Adaptive bit rate and power control was analyzed and simulated in an OFDM system over Rayleigh fading multipath channels. We have also many activities on turbo codes, for example, on systems where the constituent decoders use Berrou SOVA algorithm, which sets the scene for low complexity implementation.

WIND-FLEX is a large project (2000-2003) partially funded by the European Commission. A high bit rate flexible and configurable modem architecture is considered, which works in single-hop, ad-hoc networks and provides a wireless access to the Internet in an indoor environment. The best possible performance with reasonable complexity is attained by using a jointly optimized run time adaptive system. The research reports will be a good basis for defining new standards. During the first year (2000), the system was being specified and plans for the first prototype were made. At the end of the project a demonstrator will be implemented.

In the SCIFI project, one licentiate thesis is in preparation on the applications of per-survivor processing (PSP) in the Viterbi algorithm in the GSM/EDGE system with 8-PSK modulation. The aim is to equalize the fast fading channel when the terminal is in an express train at speeds approaching several hundred kilometers per hour. Decision feedback was used to reduce the high complexity of the Viterbi algorithm and other possibilities are being considered. The final aim is to have an idea of how an intellectual property (IP) block for algorithms is created.

In our LALAMO project, we are considering wideband wireless modems. This system is also based on OFDM, and it is quite similar to the system described in the HIPERLAN/2 and IEEE802.11a standards. We are transmitting packets of several thousand bits at a bit rate to the order of 10-50 Mbit/s in the 5 GHz frequency band. We have analyzed the performance and developed a COSSAP model for the most important synchronization and estimation algorithms, including time of arrival estimation, time and frequency synchronization and frequency-domain equalization/estimation. We also made bit-true models of several blocks to show the quantization effects and the results were published in a Master's thesis. The CORDIC algorithm was used in some of these bit-true blocks, designed for efficient ASIC implementation in future projects.

Exploitation of Results

The results of the ABRAS project can be utilized when designing a high-performance receiver for a UMTS system.

In the AWICS project, the main applications for the results are found both in the third generation mobile cellular and future communication systems as well as in new military communication networks. The research supports industrial partners in designing new system concepts as well standards, terminals and base stations for them. The basic research carried out will find potential applications in a wide range of future mobile systems, both commercial and military.

The results of the DOCENT project can be exploited in radio network planning and capacity optimization. They yield also information to be utilized in UMTS standardization.

The results of the FIXWIRE project can be applied to the design of system concepts offering improved performance in fixed and low-mobility wireless communication.

The results of the FUBS project are used when evaluating the usefulness of future unlicenced wireless radio systems.

The simulation environment and DSP demonstrator developed in the KOMA project are used for testing the performance and implementation complexity of the receiver algorithms in the WCDMA system. Parallel implementation of the PIC algorithm can be studied to set initial requirements for the hardware used in commercial implementation.

The results of many Future Military Radio Communications projects have been used in specifying and developing new tactical systems by Finnish industry. Research results and simulation environments are now used in defining suitable architecture for flexible and reconfigurable software defined radios for military applications.

In the VTT group, the results were exploited in various joint and contract research and development projects. Joint TEKES projects on fast wireless modems, multi-standard radios and multi-user CDMA detectors have been going on for some time. Third generation mobile phone systems, digital subscriber line modems, digital microwave links and robust wireless modems were considered in contract research and development projects.

Future Goals

The goal of CWC research during the year 2000 was to open up new research directions into wireless systems and application that will be used in 10 - 15 years time. Those research areas include wireless networks and related protocols (Ad-Hoc Networks, mobile IPv6) and future digital transmission schemes (multicarrier modulation, ultrawideband systems). At the same time, the core competence of CWC (cellular CDMA systems) will be utilized in near-future applications, such as WCDMA. Research topics during 2001 include some implementation issues of practical baseband algorithms for WCDMA, WCDMA radio network traffic analysis, development of flexible air-interface solutions for future high data rate wireless systems, wireless packet data systems and unlicenced wideband radios.

The practical goal of many Future Military Radio Communications projects will be the development of adaptive air interfaces and Ad Hoc mobile networks for new tactical communication systems with Finnish industries.

At VTT, a future goal is to improve our system level understanding and integrate the work of different groups even better. This should be possible since in our groups most OSI layers are covered. Detailed trade-off between the network capacity and the link capacity will be important to make future systems more efficient. In addition, since diverse standards will exist also in the future, we must be prepared to work with many different system models, including, for example, conventional and spread spectrum single-carrier systems and also multi-carrier (OFDM) systems.

Personnel

External Funding