Advanced multicarrier systems for wireless communications

Multicarrier communication techniques like orthogonal frequency division multiplexing (OFDM) and discrete multitone (DMT) have become increasingly popular over the past few years. They are currently being used in several standards (xDSL, DAB, DVB, HIPERLAN/2, etc.) and are promising candidates for future (fourth- generation) cellular communications systems. The main advantages of multicarrier modulation are high flexibility, excellent performance even in difficult environments, and the existence of efficient implementations. For application in future (fixed and mobile) wireless communication systems, multicarrier techniques will have to satisfy high demands regarding data rates, bit error rates, mobility, coverage, flexibility, and quality of service. The development of advanced multicarrier techniques whose performance is substantially superior to current techniques is thus both practically important and theoretically challenging. The research project described in this grant proposal will aim at achieving improvements of multicarrier techniques for time-varying environments. Time-varying channels are encountered especially in mobile wireless communications. Specific research topics we plan to address within the proposed project include the following: predictive equalization of time-varying mobile radio channels without the use of pilot symbols; application of channel prediction to multicarrier systems employing space-time processing and adaptive modulation; generalized multicarrier systems using pulse shaping, biorthogonal pulse design, and/or multipulse modulation; modeling and simulation of time-varying channels for wireless multicarrier systems. The project will be based on our previous work on channel prediction, statistical signal processing, and time- frequency methods.

The results of this project are a contribution to the development of advanced wireless telecommunication systems such as improved cellular mobile telephony systems and wireless local area networks (WLANs). They will thus contribute to future technological innovation in a key area of research and development. Wireless telecommunication systems are faced with detrimental effects of the radio propagation channel like multipath propagation and fading. In this respect, a particularly attractive communication scheme is multicarrier modulation which uses multiple radio frequencies in parallel for data transmission. The most prominent multicarrier system, orthogonal frequency division multiplexing (OFDM), is used in important standards for wireless communications like digital audio and video broadcasting, WLANs, and wireless broadband access networks. In the course of this project, we performed fundamental research on various aspects of wireless multicarrier systems. We devised algorithms for tracking the radio channel at the receiver in a fully adaptive mode and without the need for transmitting training data (i.e., dedicated pseudodata that allows the receiver to measure the channel). We designed improved pulse shapes (the signals carrying the data) and introduced generalized multicarrier systems using several different pulses in parallel. We also calculated the maximum achievable data rates of OFDM systems for the large bandwidths that will be required by tomorrow`s ultrafast wireless data links. Further results of the project concern a recent major development in wireless communications known as multiple- input multiple-output (MIMO) wireless systems. MIMO systems use multiple antennas at the transmitter and receiver to achieve significantly increased data rates and substantially improved link reliability. We developed efficient techniques for tracking MIMO channels and for recovering the transmitted data at the receiver. Furthermore, we devised multiple-antenna receiver techniques for a special variant of the UMTS cellular telephony standard. On a more theoretical level, we introduced realistic statistical models for wireless channels and novel techniques for the modeling and processing of nonstationary random signals.