In the 1970s, the consensus emerged that it was going to be easier and cost less to filter in the time domain rather than the frequency domain. This is the reason the wireline world abandoned frequency multiplexing and adopted time-division multiplexing for wireline backhaul transport. This thinking was taken into the wireless standardization
committees. GSM effectively was based, and is still based today, on an ISDN structure
and a time-division multiplex on the air interface. The disadvantage with the time-division multiplex is that RF bursts need to be shaped and modulated onto the RF channel. As we explain in the next chapter, it is proving quite difficult to deliver flexible bandwidth on demand from any of the TDMA
options, partly because of the challenge of pulse shaping in a multiple-slot handset. CDMAmoves the process of time domain filtering to baseband and delivers greater flexibility in terms of bandwidth on demand and multiple per-user traffic streams (we study how this is achieved in Chapter 3). Additionally, as described earlier, the CDMA multiplex allows a relaxation of RF channel spacing. CDMA only became possible in the early to mid-1990s, when it became feasible in cost and power budget terms to implement root raised cosine filters and low-cost, low-power budget PN code generators and numerically controlled oscillators (NCOs, studied in detail in Chapter 3).
In fourth-generation cellular, it is likely that CDMA will be combined with OFDM techniques to provide additional channel resilience (using 10, 15, or 20 MHz bandwidths). These hybrid time domain/frequency domain multiplexing schemes are generically described as coded orthogonal frequency-division multiplexing (COFDM). This is only possible when sufficient processing power is available to undertake handset transmit and receive Fast Fourier Transforms, but the benefit will be further improvements in the consistency of bandwidth quality (effectively an increase in coherence
bandwidth). For the present, attention is focused on making CDMA work well.

Leave a Reply

Your email address will not be published. Required fields are marked *

scroll to top