Techniques for UWB-OFDM
Modern wireless transmission systems usually transmit in one part of the total radio spectrum, which is reserved by local regulators (the RegTP in Germany or the FCC in the US) for exclusive usage. These frequencies are often sold to a service provider; however some other ranges of the spectrum are available for free use.
A new approach is triggered by the lacking of free frequencies. Recently, the American FCC has decided to allow transmitting over a wide spectral range (>500 MHz), even where other systems have been granted exclusive excess. This remains valid under the premise that a given threshold of the power spectral density (i.e. the transmit power related to the bandwidth) is not exceeded, in order to prevent this "parasitic system" from distorting the traditional services like e.g. WLAN, Bluetooth or UMTS. The approach is visualized in Fig. 1 and in the research community, it is known under the abbreviation UWB for ultra-wideband.
A solution for such a system has been proposed in terms of the multiband OFDM (MB-OFDM) standard, where orthogonal frequency division multiplex (OFDM) separates like depicted in Fig. 2 the employed spectrum into several subcarriers. On these subcarriers independent data streams are transmitted.
Because of its high flexibility and adaptability, OFDM offers smart solutions for two main challenges of UWB systems: the frequency selective channel due to multipath propagation of the signal and interference resilience against signals of simultaneously operating "narrowband" systems.
The frequency selective channel usually introduces distortions from temporally adjacent symbols into the received signals (inter-symbol interference). This is circumvented in OFDM by using several small portions (the subcarriers) of the frequency spectrum for transmitting signals. Then, they experience only a frequency flat channel. Hence, they are individually not harmed by frequency selectivity and are free of inter-symbol interference. However, for OFDM in UWB scenarios, a more and more important receiver component with respect to the achievable data rate becomes the channel estimation, that suffers from the increased amount of channel parameters as already reported in several of our research publications, lately also in the context of UWB.
Narrowband interference can easily be handled by switching off the subcarriers, where the respective narrowband interferer is persistent (see Fig. 2), if the data is coded across all subcarriers. This is done in the receiver, but can also be additionally performed in the transmitter in order to save battery power. Other techniques comprise spreading a single signal over various subcarriers and time instants in order to always have sufficient undistorted information on each transmitted signal. The basis for active interference handling is an efficient detection method that is capable of detecting even low power interferers with a sufficient precision. Besides suppressing a distorted subcarrier, an estimation of the interference and a following subtraction from the received signal (cancelling) becomes an alternative option to deal with this information. E.g. optimal points for switching between suppressing and cancelling in the receiver have to be identified.