Multi‐Carrier Spread Spectrum

Especially these days, we observe that frequency spectrum is a limited and valuable resource for wireless communications. A good example can be observed among network operators in Europe for the prices to pay for frequency bands for UMTS. Keeping this in mind, the first goal when designing future wireless communication systems has to be the increase in spectral efficiency. The development in digital communications in the past years has enabled efficient modulation and coding techniques for robust and spectral efficient data, speech, audio and video transmission. Here, we should mention two interesting and successfid techniques. These are the multi-carrier modulation (e.g. OFDM) and the spread spectrum technique (e.g. DS-CDMA). The breakthrough of multi-camer communications came in the 1990s as orthogonal fiequency division multiplexing (OFDM) was chosen for the European terrestrial digital audio broadcasting (DAB) and digital video broadcasting (DVBT) standards. Furthermore it has been chosen for the broadband wireless indoor standards ETSI HIPERLAN-I1 in Europe, IEEE-802.11 in the USA and MMAC in Japan. On the other side, we have observed the success of spread spectrum communications in mobile communications, whose first commercial widespread employment came with the CDMA based mobile radio standard IS-95 in the USA and nowadays with the use of CDMA for third generation mobile radio systems world wide, UMTSAMT 2000. Since 1993 various combinations of multi-carrier (MC) modulation and the spread spectrum (SS) technique have been introduced and the field of MC-SS communications has become an independent and important research topic with increasing activities. It has been shown that MC-SS offers high spectral efficiency, robustness and flexibility. Two different combination philosophies since 1993 have been followed: Namely MC-CDMA (or OFDM-CDMA) and MC-DSCDMA. MC-CDMA is based on a serial concatenation of DS spreading with MC modulation. The high rate DS spread data stream is MC modulated in that way that the chips of a spread data symbol are transmitted in parallel on several subcarriers. This means that the MC-CDMA system performs the spreading in the frequency domain, which allows kr simple signal detection strategies. As for DS-CDMA, a user may occupy the total bandwidth for the transmission of a single data symbol. The separation of the users' signals is performed in the code domain. This concept was first proposed with OFDM for optimum use of the available bandwidth for the downlink of a cellular system using orthogonal Walsh-Hadamard codes. MC-DS-CDMA converts the data stream onto parallel low rate sub-streams before applying DS spreading on each sub-stream in the time domain and modulating each sub-stream onto a sub-carrier. The sub-carrier spacing is proportional to the inverse of the chip rate. This will guarantee the orthogonality between the spectra of the sub-streams. If the spreading code length is smaller or equal to the number of sub-camers, a single data symbol is not spread in the fie quency, instead in time domain. By using high numbers of sub-carriers, this concept is advantageous in the sense af exploiting time diversity. However, due to the frequency non-selective fading per sub-channel, fi-equency diversity can only be exploited if channel coding with interleavjng or sub-camer hopping is employed. Furthermore, the sub-carrier spacing might be chosen larger than the reciprocal'of the chip duration in order to provide a higher frequency diversity. The main characteristics of MC-CDMA and MC-DS-CDMA are illustrated in Table 1.