Frequency diversity performance of coded multiband-OFDM systems on IEEE UWB channels

This paper investigates how convolutional and Reed-Solomon codes can be used to improve the performance of multiband-OFDM by utilizing the inherent frequency diversity of the new IEEE 802.15 UWB channel models. A normalized amplitude autocovariance function of the Fourier transform of the channel impulse response is defined. Then the average coherence bandwidths of CM1, CM2, CM3, and CM4 are estimated to be 31.6, 16.3, 11.0, 5.8 MHz, respectively. Using the central limit theorem, we can expect that the performance of an uncoded OFDM system on CM1-CM4 without shadowing is the same as in a Rayleigh fading channel with uniformly distributed phase. The performance of a convolutional code with rate 1/2 and constraint length 7 on CM2-CM4 without shadowing are up to 0.4 dB worse than that of on an uncorrelated Rayleigh fading channel. The loss for CM1 is around 1 dB. A block interleaver with 32 rows and 24 columns was used. This result is also valid for a convolutional code with rate 1/4 and constraint length 7. For code rates around 2/3, the performance of a punctured convolutional code with soft-decision decoding is much better than that of the Reed-Solomon codes with with 6, 7, and 8 bits per symbol and hard-decision decoding.