Investigation of Multicarrier Schemes for Cognitive Radio Applications

The frequency spectrum is an essential resource of wireless communication. Special sections of the spectrum are used for military purposes, some frequency bands are sold by governments to broadcasting and mobile communications companies for commercial use, others such as ISM (Industrial, Science and Medical) bands are available for the public free of charge. As the spectrum becomes overcrowded, there seem to be two possible solutions: pushing the frequency limits higher to frequencies of 60 GHz and above, or the reaggregating the densely used licensed frequency bands. Both ideas lead to di culties: the use of higher frequencies requires the design and manufacturing of special analog devices, while the reaggreagtion of the spectra requires complex, intelligent and adaptive systems. As a result of the digital switchover of TV broadcasting, certain frequency bands become unused. According to recent plans, intelligent and opportunistically communicating radio systems aiming data transmission applications might be implemented on these frequencies. As the primary (incumbent) user of these frequencies will still be the broadcast sector, data communication systems operating in these frequencies will have to incorporate sophisticated intelligence and fast spectrum sensing capabilities to prevent interference. The physical layer of these communication systems must meet some special requirements implied by these special constraints. Cognitive radio based opportunistic exploitation for spatially and temporally unused frequencies is considered to be a feasible approach to improve the spectral e ciency and to introduce new services in the legacy bands. The selection of the appropriate modulation technique is a major issue. The frequency-selective nature of the radio transmission channel and the need for multiuser applications call for multicarrier modulation schemes, which are the subject of this thesis. Today the Orthogonal Frequency Division Multiplexing (OFDM) modulation scheme is the most widespread technique for high-speed wireless data transmission. It is used in many broadcasting and communication systems such as Digital Video Broadcasting (DVB), Digital Audio Broadcasting (DAB) and certain types of IEEE 802.11 Wireless Local Area Networks (WLAN). Using OFDM, low-complexity demodulation and modulation can be performed by Fast Fourier Transform (FFT) and Inverse FFT (IFFT) operations, respectively. Using Cyclic Pre x (CP) Inter Symbol Interference (ISI) can be eliminated which is the precondition of e cient channel equalization in the frequency domain. Nevertheless, this scheme has some drawbacks. Due to the large dynamic range of the transmitted signal, OFDM is highly sensitive to the nonlinear characteristics of the Power Ampli ers (PA). This nonlinearity induces in-band and out-of-band spurious products, which might degrade the system's performance. OFDM system performance can also be severely degraded by the frequency mismatch of the transmitter and receiver local oscillators, therefore, a very robust synchronization technique is required. Moreover, without ltering the transmitted signal, the out-of-band radiation of OFDM is considered as only moderate, which is a major drawback in the opportunistic context, where signi cant adjacent-channel leakage leads to interference. There are many issues regarding cognitive radio scenarios which cannot be met by OFDM. This is why other multicarrier schemes have become a point of interest. In this thesis, besides OFDM, the focus is on three possible alternatives: DFT-Spread OFDM (DFTS-OFDM), Constant Envelope OFDM (CEOFDM) and Staggered MultiTone (SMT). Each of these schemes have some advantages over OFDM which makes them bene cial for use in cognitive radio applications. In this thesis the above mentioned multicarrier schemes are closely investigated and compared. First, their signal model and the required signal processing blocks are introduced. Then the advantages and drawbacks derived from the transmitted signal properties are presented. Later, the e ects of nonlinearities, synchronization impairments and multipath propagation will be investigated on the baseband transceiver model. Novel channel equalization schemes are also presented for SMT exhibiting better system performance than conventional techniques. Peak-to-Average Power Ratio (PAPR) reduction by clipping is discussed for OFDM signals. Iterative compensation techniques employing the Turbo principle are introduced to improve the performance of amplitude limited OFDM and SMT signals. OFDM based techniques are also presented for the SMT scheme in order to reduce the large dynamic range of the transmitted signal. Finally, applications of recursive discrete fourier transform in cognitive radio receivers for lter bank multicarrier modulation are investigated.