29-11-2012, 06:19 PM
ADAPTIVE EQUALIZATION FOR TDMA DIGITAL MOBILE RADIO
ADAPTIVE EQUALIZATION FOR TDMA DIGITAL MOBILE RADI1.doc (Size: 43 KB / Downloads: 21)
Abstract:-
Adaptive equalization for a TDMA (time-division multiple-access) digital cellular system is discussed.
A survey of adaptive equalization techniques that includes their performance characteristics and their implementation complexity is presented.
The design of adaptive equalization algorithms for a narrowband TDMA system is considered. It is concluded that, on the basis of implementation complexity and performance in the presence of multipath distortion and signal fading, MLSE (maximum-likelihood sequence estimation) and DFE (decision feedback equalization) are viable equalization methods for mobile radio.
INTRODUCTION
The demand for mobile radio telephone service through out US,Canada,Europe and japan has accelerated the development spectrally efficient digital modulation/demodulation techniques for replacing the spectrally inefficient systems based on analog modulation.
The proposed narrow band and wide band TDMA digital cellular system require adaptive equalization at the demodulator to combat the ISI resulting from the time Variant multipath propagation of the signal through the channel.
EQUALIZATION OF DIGITAL MOBILE
RADIO CHANNEL
Mobile radio channels are generally characterized as fading multipath channels with time dispersion ranging from few microseconds upto 100microseconds. Such large multipath spreads results in ISI which necessitates the use of an equalizer.The frequency selective channel characteristics in mobile radio generally results in channel spectral nulls. As a consequence,linear equalizers must be ruled out since it is generally recognized that their performance is poor on such channel characteristics. Hence,our choice is limited to non linear equalization techniques namely DFE,MLSE and MAP.
BLOCK PROCESSING ALGORITHMS
This approach has been investigated by HSU,DAVIDSON and CROZIER. In these schemes a typical block consists of training symbols followed by data symbols.
HSU investigated a non-linear decision feed back symbol detection technique in which the symbol blocks on each side of a data block are used for channel estimation.
CROZIER investigated a non-linear decision feed back symbol detection technique in which the symbol blocks on both side of a data block are used for channel estimation.
The other schemes based on block adoption with channel tracking via interpolation between successive blocks of training symbols have been proposed by DAVIDSON.