We show that because of drastic differences in the fading statistics between ultra wide bandwidth (UWB) multi-carrier and direct sequence code division multiple access (DS-CDMA) approaches, DS easily scales to Gbps rates while multi-carrier architectures have severe difficulty. Both DS-CDMA and orthogonal frequency division multiple-access (OFDM) are well understood and proven modulation techniques in conventional (narrowband) commercial technologies (e.g. DS-CDMA in cell phones; OFDM in IEEE 802.11a/g). The maturity of these approaches, however, is vastly different when applied to ultra-wideband (UWB) systems. Already implemented and operating in silicon, DS-CDMA architectures have proven to be the most mature and scaleable for UWB on both a theoretical as well as implementation basis. Among the proposed approaches before the IEEE 802.15.3a standards committee, the DS-CDMA transmitted waveform (which is the "thing" being standardized) is uniquely capable of serving the broadest diversity of applications. It can, for example, allow very low-cost low-power transmit-only devices (even at Gbps rates) because it requires no FFT or DAC or DSP. At the same time, receivers can incorporate varying degrees of DSP to provide scaleable power/cost versus performance. We present performance comparisons of DS-CDMA [8] vs. the proposed multi-band MB-OFDM architecture [11] for outage range in a variety of multipath environments. Moreover, we describe how DS-CDMA UWB architectures can support robust and flexible multi-user capabilities, protect against in-band interference, and provide high-resolution ranging capabilities for safety-of-life applications.
