The current evolution of air interfaces often appears as a battle between CDMA/single carrier and OFDM/multi-carrier, a battle that OFDM/multicarrier would probably win. The questions that arise are: What would be the impact on globalization visa- vis adoption of the CDMA or OFDM? It the multi-carrier approach really needed? Together with Frequency Domain Equalization (FDE), what are the most efficient choices for air interfaces in future wireless systems?
CDMA is used in a number of 2G/3G standards for mobile communications such as IS-95, cdma2000, wideband CDMA (W-CDMA), time-division synchronous CDMA (TDSCDMA). Moreover, the principle of spread spectrum multiplexing has found applications in early IEEE 802.11 and 802.11b/g WLAN, Bluetooth, and cordless telephony.
OFDM and the related multiple access scheme, OFDMA, are used in IEEE 802.11a/g/n WLAN, HIPERLAN/2, WiMAX, DVBT, asymmetric digital subscriber line (ADSL), very high rate DSL (VDSL). Moreover, OFDM has also been chosen as the physical layer architecture for 3GPP long-term evolution (LTE) and it is easy to predict that even if with some CDMA/spread spectrum component in the near future OFDM will be used.
A multicarrier solution such as OFDM appears as the necessary replacement in 4G systems. The main reason is that OFDM offers a lower complexity solution than current single carrier systems to the problem of performance degradation over severely frequency selective channels, such as wireless channels for very high speed communications (several tens of Mbps): with one-tap equalizer and the introduction of the cyclic prefix is able to keep the orthogonality among different users’ signals and reduce the intersymbol interference (ISI).
Conventional DS-CDMA receivers consist of a RAKE combiner whose complexity increases with the increase of the number of propagation paths in the channel. This limits its performance over severely frequency selective channels. Moreover, in a frequency-selective channel, multicode operation severely suffers from the loss of orthogonality among the orthogonal spreading codes and the performance with coherent Rake combining severely degrades.
However, it is getting clearer that the advantages of OFDM over traditional CDMA systems are not related to the multicarrier nature of OFDM, but to the use of the equalization in the frequency domain (FD).
Cyclic-prefixed single-carrier systems (i.e. DS-CDMA, multicode DS-CDMA and also High Speed Downlink Packet Access) with FDE can achieve performance close to the one of a multicarrier-based system with the same receiver complexity. It is also worth noting that the best candidate for the uplink of 4G systems is the technique Interleaved FDMA (IFDMA), which is a single carrier scheme with FDE.
We can conclude that over frequency selective channel, the winning choice towards 4G is not multicarrier or single carrier but FDE-based: FDE brings the combination Cyclic Prefix + one tap equalizer + FFT implementation, which represents a low complexity and efficient way to remove ISI in frequency selective channels. Whilst this is a pretty straightforward conclusion, in this chapter we wonder if, in the evolution path towards 4G, it is possible to identify other key features that have not been clearly identified so far. One wonders whether once FD equalization is chosen, there are “optimal” choices that must be made to take full advantage of the FDE.
