This paper introduces a new parallel carrier recovery architecture suitable for ultra- high speed intradyne coherent optical receivers (e.g., ≥40Gb/s). The proposed scheme combines a novel low-latency parallel digital phase locked loop (DPLL) with a feedforward carrier phase recovery (CPR) algorithm. The new low-latency parallel DPLL is designed to compensate not only frequency offset, but also frequency fluctuations such as those induced by mechanical vibrations. It is well-known that nonzero frequency offset leads to higher phase error variance in a feedforward CPR. Furthermore, it has been recently shown that laser frequency instability caused by mechanical vibrations significantly degrades the performance of CPR. Other effects such as power supply noise may also introduce laser frequency fluctuations. To avoid receiver performance degradations, both frequency offset and frequency fluctuations should be compensated before the feedforward CPR block. We show that this task can be achieved by using a typical decision-directed serial DPLL. Then, a new approximation to the DPLL computation is introduced to enable a parallel-processing implementation in multigigabit per second receivers. The proposed technique reduces the latency within the feedback loop of the DPLL introduced by parallel processing, while providing a bandwidth and capture range close to those achieved by a serial DPLL. Simulation results demonstrate that the effects caused by frequency deviations can be eliminated with the proposed parallel carrier recovery architecture.

Reduced-complexity multiuser interference equalization and group decoding (ME-GD) is an attractive way to improve performance of multiuser transmissions for narrowband additive white Gaussian noise channels. ME-GD receivers have been investigated in the past for convolutionally coded synchronous multiple-access channels (MACs). In this paper, we propose a new ME-GD for trellis-coded asynchronous MACs. The coding information of a subset of users and the multiuser interference are jointly taken into account in the branch metric computation using per-survivor processing. The novel approach leads to a family of Viterbi decoders which offers a good tradeoff between performance and complexity. As a second contribution of this paper, we derive theoretical bounds for the performance of ME-GD considering the effect of error propagation, which is critical for asynchronous MACs. Numerical results show the excellent accuracy of the proposed bounds. Finally, based on the novel theoretical analysis, we derive a new criterion to achieve good tradeoffs between performance and complexity.

Joint reduced-state detection and group decoding (RSD-GD) is a promising approach to increase the spectral efficiency of multiple-access systems. Performance of RSD-GD receivers improves as coding information of more users is taken into account in the detection algorithm. For asynchronous channels, however, there is another approach to improve performance: use of more states of the multiple-access channel (MAC). Performance of RSD-GD receivers for asynchronous channels has been investigated using computer simulations; however, analytical studies of the performance of RSD-GD have not been reported so far. In this paper we derive upper bounds for the performance of RSD-GD over asynchronous MAC. The new results are based on an improved upper bound on the error event probability, which is significantly tighter than the traditional bound based on pairwise error events. Comparisons with values obtained from computer simulations show the excellent accuracy of the proposed bound.

Block-error processes in transmissions over slow-fading channels can be accurately modeled by a two-state Markov chain [the block Markov model (BMM)]. Another line of research has focused on the use of a channel-state Markov model (CSMM) to analyze block transmissions. Although both techniques provide results that agree well with observations, the relationship between both Markov models has not been recognized in the previous literature. In this letter, we show that the BMM for slow-fading channels can be directly derived from the CSMM. In addition, we introduce a greatly simplified channel-modeling methodology. In the new methodology, the BMM is the primary channel characterization tool, and the CSMM becomes essentially an estimation technique that provides parameters for the BMM. Results of packet transmissions in slow-fading channels show that our approach provides significant improvements in both accuracy and simplicity over previously proposed techniques.

The combination of space-time coding and multiuser detection is an effective way to increase the spectral efficiency of multiple access systems. In this paper we propose a new multiuser decoder for space-time coded asynchronous DS-CDMA systems over Rayleigh fading channels. The proposed scheme is an extension of the one introduced in Corral-Briones et al. (2002). Complexity is reduced exploiting the fact that the trellis for the entire decoder can be decomposed into a set of trellises for subsets of users with a lower total state count. Different tradeoffs between performance and complexity can be achieved by adjusting the number of users decoded by each trellis. We investigate the behavior of the proposed receiver for a few users with moderate and high correlation. Simulation results show that the new scheme achieves an order of magnitude lower error rate than the conventional receiver (i.e., a single-user decoder that ignores multiuser interference) at almost the same complexity. Moreover, doubling the complexity, the proposed receiver performs near the single user bound. We also address the performance and complexity of the new scheme in high bandwidth efficiency situations, which are of high practical interest. It is found that performance close to the single user bound is achieved with our scheme at practical frame error rates with significantly lower complexity than the receiver proposed in Corral-Briones.

Space-time coding (STC) is an efficient technique to combat severe fading effects and improve system capacity in wireless communications. However, for CDMA communications, multi-antenna transmission increases the multiuser interference (MUI). In addition, if the data rate is high, the processing gain can be quite small, and thus the MUI is further enhanced. In this paper we propose a new non-iterative multiuser decoder for space-time coded asynchronous DS-CDMA. Non-iterative decoding schemes are important in situations where decoding delay must be minimized, for example in interactive applications such as videoconferencing. The proposed scheme introduces a novel reduced-state multiuser sequence detection algorithm suitable to decode few co-located users with high data rates (i.e. users within a beam with short spread sequences). We evaluate the performance of the proposed scheme in high and moderate multiuser interference channels. Simulation results show that a major performance improvement over a conventional receiver is achieved with the proposed scheme even in moderate interference channels. In high interference channels, where the conventional receiver is useless, the proposed scheme performs near the single user bound at practical frame error rates and with reasonable complexity.

We propose a new bounded delay retransmission scheme that enhances the reliability of compressed video transmission in multicode DS-CDMA (MC-CDMA) systems. We consider H.263/H.263+ video at low data rates (i.e., <64 kbps) in an IS-95B system over slow fading Rayleigh channels. Our approach is designed to reduce not only packet loss, time delays, and complexity, but also interference effects. The latter is critical in CDMA systems. Based both on theoretical analysis and entire system simulation, we investigate the improvements in packet transmission and system capacity achieved by the new scheme, and provide results for H.263 video transmission over IS-95B. Our results show that the scheme proposed significantly improves the error resilience of video transmission over MC-CDMA systems with small additional complexity. We also introduce new techniques for the analysis of video transmission in MC-CDMA with error control based on ARQ protocols, and present the application of these techniques to the proposed retransmission scheme.

We propose a new reduced-complexity multiuser receiver for asynchronous trellis-coded CDMA systems in AWGN channels. Based on the coding information of a subset of users and the set partitioning principles inherent in TCM, reduced-state trellises are constructed. The Viterbi algorithm is thus used to jointly decode the trellis codes of this subset of users and equalize the multiuser interference of the channel. As a special case of the proposed receiver, we study a structure that decodes only one user and equalizes the multiuser interference. Simulation results show that the performance of the new reduced complexity receiver approaches the trellis-coded single-user bound for both low and high-interference channels. Moreover, we compare the performance of the proposed receiver with that of previously proposed suboptimal structures. Our scheme is found to significantly outperform previously known suboptimal structures.

In channels where the delay spread is smaller than the chip interval (e.g., an IS-95 system operating in indoor environments), spread-spectrum signals do not give rise to path diversity. In this situation, maximal-ratio combiner (MRC) RAKE receivers with resolution reduction (RR) of the diversity receiver branches may be used by the mobile stations to provide diversity gain, significantly improving system performance. A new resolution reduction technique based on the use of a minimum mean-square-error diversity combiner (MMSEC) is proposed in this work. We show that, under very general assumptions, this new method of RR is optimal. A detailed study of the performance of a dual-branch MMSEC-RAKE receiver with RR in a typical indoor office environment is presented. In order to allow a simple practical implementation, a suboptimal structure of the MMSEC is also proposed. Numerical results show that this new receiver scheme provides a 1.2-dB improvement over the previously proposed RR technique based on MRC, and a 4.9-dB improvement over conventional MRC-RAKE receiver without RR, at a frame-error rate of 0.01 for the downlink of the IS-95 system in a typical indoor office environment.

In channels where the delay spread is smaller than the chip interval, spread spectrum signals do not give rise to path diversity. In this situation, distributed antennas systems may be used to provide path diversity, significantly improving the system performance. The interference component caused by the same cell-site users, or interpath interference (IPI), is a dominant factor in the performance of these systems. In their analysis it is typically assumed that the probability density function (PDF) of the IPI can be approximated by the Gaussian distribution. This is called the Gaussian approximation (GA). We show that the GA for the IPI does not provide an accurate characterization of the performance of the downlink of DS-CDMA indoor systems with distributed antennas. The bit error rate (BER) at the output of a practical RAKE receiver is computed using an accurate approach for the IPI. We show that when the IPI density is thus modeled, the signal to noise ratio (SNR) at the output of the combiner is well approximated by a random variable with a Weibull distribution. Comparison of the BER obtained by the new approach for the IPI with those obtained by the GA shows that the latter method incurs in considerable error when used to evaluate performance of DS-CDMA indoor systems with distributed antennas.

We study the performance of the downlink of a DS-CDMA indoor system that incorporates distributed antennas and the convolutional code (CC) as error correction method. Based on the improved approach for the interpath interference (IA-IPI) introduced by Hueda (see Proc. 11/sup th/ PIRMC'00, London, 2000), we propose analytical expressions for the bit and frame error rate (BER/FER) at the output of the Viterbi decoder with soft decision decoding. The BER and FER of the forward traffic channel of an IS-95 system are then derived by computer simulation of the entire system and by numerical evaluation of analytic expressions. Results for several cases involving different numbers of antenna elements and different levels of interference from other cells are presented. The analytical models are shown to be most efficient in terms of computational cost while preserving excellent accuracy.

In channels where the delay spread is smaller than the chip interval, spread spectrum signals do not give rise to path diversity, thus a classical maximal ratio combiner (MRC)-RAKE receiver sets to unity the number of branches. In this situation, MRC-RAKE receivers with resolution reduction (RR) of the diversity branches may be used to significantly improve performance in the forward link of DS-CDMA systems (Yang 1995). As a result of the use of RR techniques, the noise components at the finger outputs are correlated. It has been shown that the MRC is suboptimal in the presence of correlated noise components. In this paper, a minimum mean squared error combiner (MMSEC) RAKE receiver with RR is proposed. The performance of the RAKE receiver with two branches is studied by means of theoretical analysis and computer simulations. To allow a simple practical implementation, a suboptimal structure of the MMSEC is also proposed. Frame error rate (FER) is obtained by using a simulation system based on the forward link of the digital cellular standard IS-95 in a typical indoor radio channel. Our results show that this new receiver scheme can achieve a gain of 3.6 dB over a conventional MRC-RAKE receiver at an FER of 0.01, with small increase in complexity.