In this paper, we consider a low-complexity iterative receiver for decoding multi-user information data in a bandwidth efficient convolutionally coded TH-UWB-CDMA system introduced in [1]. The proposed receiver structure consists of a bank of soft interference canceller-likelihood calculators (SICLC), each followed by :a soft-input soft-output (SISO) convolutional decoder. Information is fed back from SISO decoders to SICLCs for interference cancellation. Each SICLC then provides soft information about coded symbols in the form of log-likelihood ratio (LLR), which is used by the corresponding SISO decoder as a priori information about coded symbols, We evaluate the performance of the proposed receiver in AWGN channels. Our numerical results show that the new receiver significantly improves the system performance. Moreover, at relatively high signal-to-noise ratios, after few iterations, multiple access interference (MAI) is efficiently suppressed, and single-user performance is achieved.