Turbo codes are parallel concatenated codes whose performance in the additive white Gaussian noise (AWGN) channel has been shown to be near the theoretical limit. In this paper, we describe a low-rate superorthogonal turbo code that combines the principles of low-rate convolutional coding and that of parallel concatenation. Due to the bandwidth expansion, this code outperforms the ordinary turbo code both in AWGN and especially in fading channels. Thus, superorthogonal turbo codes are suited mainly for spread-spectrum applications. For the purposes of iterative decoding, we concisely describe the connection between the optimal maximum a posteriori symbol estimation and suboptimal soft-output decoding based on sequence estimation. The suboptimal decoder produces outputs that can directly be used as additive metrics at successive decoding iterations, without the need for estimating channel noise variance. Simulation results in AWGN and flat Rayleigh fading channels are also presented, along with analytical upper bounds of bit- and frame-error probabilities.

Universal Mobile Telecommunications System (UMTS) has a target maximum user bit rate of 2 Mbit/s. We present a performance comparison between four TDMA schemes and one CDMA scheme that all can provide the required user data rate. Evaluation considers link level performance, noise limited uplink range and cellular capacity. Three different environments, indoor large room, street micro cell and urban macro cell, have been used. Results show that when sophisticated methods are utilized schemes based on both TDMA and CDMA access methods can provide very similar performance for implementing UMTS high bit rate services when many different environments are considered.

Turbo-codes are parallel concatenated convolutional codes (PCCC) whose performance in an AWGN channel has been shown to be near the Shannon limit. We describe a low rate superorthogonal turbo-code that outperforms the ordinary turbo-code at the expense of bandwidth expansion both in AWGN and especially in fading channels. Thus superorthogonal turbo-codes are suited mainly for e.g. CDMA applications. We concisely describe the soft output Viterbi (1995) algorithm using a priori inputs (APRI-SOVA) with connection to optimal maximum a posteriori (MAP) symbol estimation. Simulation results in AWGN and fading channels are also presented.