Zihua Guo

An effective multiuser receiver for DS/CDMA systems

Multiuser detection (MUD) is a key technique for combating multiple access interference (MAI) in CDMA systems. We propose a simple and effective multiuser receiver, which we refer to as the objective function based multiuser detector (OFMUD). This is a successive multistage joint detector based on the statistical optimization of an objective function. Two objective functions are considered in this paper. The first one is based on a maximum likelihood solution, and the second one is based on the least squared error solution. A key feature of the proposed detector is that among all the K user bits, we update the estimate of one and only one bit (the one that is most likely to be wrong) in every stage. Analytical and numerical results show that significant performance improvement can be achieved compared with the conventional DS/CDMA receiver. At the same time, the near-far problem is alleviated. In addition, it is shown that the proposed receiver can outperform the popular multiuser parallel interference cancellation receiver in most cases with a smaller computational cost.

Hybrid inverse halftoning using adaptive filtering

We propose a novel fast inverse halftoning technique using a combination of spatial varying filtering and spatial invariant filtering. The proposed algorithm is significantly simpler than most existing algorithms. Without explicit edge region classification, the proposed spatial varying filter is shown to be capable of preserving edges effectively.

A low complexity reduced-rank MMSE receiver for DS/CDMA communications

The minimum mean squared error (MMSE) receiver is a linear filter which can achieve optimal near-far resistance in direct-sequence code-division multiple-access communications. However, one of the main problems of this receiver is the required number of filter taps, which is typically large. This is especially true in systems with a large processing gain in which case the receivers computation burden becomes very high. As a result, methods for reducing the complexity of the MMSE receiver have been of great interest in recent years. We propose an efficient partitioned MMSE receiver based on a classification algorithm. It is shown that the computational complexity (in terms of the filter taps) of the proposed receiver can be reduced significantly while good performance is maintained. Based on the special structure of our proposed receiver, we also propose a release-merge adaptive partition algorithm which can update the partition and the receivers coefficients simultaneously. In particular, it is demonstrated that the proposed receiver can perform much better than previously proposed reduced-rank MMSE receivers, such as the partial despreading MMSE receiver and the cyclically shifted filter bank receiver, with even a smaller number of taps.

Performance of VSG-CDMA and MC-CDMA in multirate systems

The performance of two multirate schemes in DS/CDMA systems, variable spreading gain (VSG-CDMA) and multicode (MC-CDMA) are compared. We first present the signal model for a multirate DS/CDMA system. Then, the performance in terms of asymptotic multiuser efficiency (AME), near-far resistance (NFR), and BER are evaluated. Some typical detectors are considered in multirate systems. These include the optimal maximum-likelihood detector, the sub-optimal linear detector, the interference canceller and the conventional MF detector. We show that in multirate systems, MC-CDMA has a very similar performance to VSG-CDMA employing low-rate detection (LRD). But it performs much better than VSG-CDMA employing high-rate detection (HRD). Furthermore, we show some important distinctions between multirate systems and single-rate systems.

Parameter estimation for image/video transcoding

In most of the current image/video compression standard, such as JPEG, MPEG, H.263, etc, the DCT is used. The DCT coefficients for every AC frequency component usually have Laplacian distribution. Such a distribution is critical in the image or video transcoding system. Therefore, the parameter in the distribution should be well estimated. Different from the normal parameter estimation, in the transcoding system, the original DCT value is unavailable, and only the dequantized value is available. In this paper, we propose 3 estimation methods based on the dequantized value for the transcoding system. The simulation demonstrates that they can yield very good results. In addition, under a fixed bit rate, they can work even better than estimation with the original DCT value in the sense of image/frames PSNR.

Robust joint channel estimation and multiuser detection with differential modulation in DS/CDMA systems

Multiuser detection (MUD) is a powerful technique for mitigating the effects of multiple access interference (MAI) and near-far problems in DS/CDMA systems. However, most of the previous work on MUD assumes perfect channel information at the receiver. In this paper, we propose an effective joint channel estimation and data detection method with differential modulation and interference cancellation over slow varying fading channels. It is shown that we can achieve near one-user performance with few pilots. Furthermore, we show that the proposed receiver is near-far resistant. In particular, it is shown that the complexity of the proposed scheme is quite small and that it can be reliably used in relatively fast fading channels.

Erlang capacity for satellite CDMA systems with adaptive power control and interference cancellation

In satellite DS/CDMA communication systems, power control (PC) is critical, and only open loop power control (OLPC) is feasible due to the large propagation delay. We propose an adaptive OLPC method for DS/CDMA satellite communication systems based on a Markov chain channel model. Then, under the assumption that multiuser interference cancellation (IC) is employed, we derive the Erlang capacity of the satellite DS/CDMA system with the OLPC scheme. It is shown that IC and power control error (PCE) affects the Erlang capacity significantly. In addition, dual-satellite diversity is considered and we show that diversity can improve the Erlang capacity of such a system greatly.

Optimal power allocation for multirate CDMA systems with interference cancellation

We derive the optimal (minimal) power allocation and capacity for a multirate DS/CDMA system with various QoS requirements. Most of the previous work on power allocation only dealt with the conventional matched filter (MF) receiver and no multiuser detection (MUD) was considered. It is well known that MUD and in particular interference cancellation (IC) can be used to improve the system performance significantly. Hence, after proposing a novel graph-aided method to do optimal power allocation with the conventional MF receiver, parallel IC (P/C) is considered in multirate systems. We propose an iterative approach for optimal power allocation with PIC in multirate systems. It is shown that the proposed power allocation algorithm can converge to the optimal solution quickly. Furthermore, it is demonstrated that when PIC is combined with optimal power allocation, the power dissipation decrease and the system capacity increase significantly.

An adaptive receiver for dynamic traffic DS/CDMA packet transmission

The adaptive minimum mean squared error (MMSE) based detector is one of the most attractive CDMA receivers because of its relatively simple structure and good performance. However, most of the previous research on such receivers was based on the assumption that the system traffic is fixed during the receiver processing. We propose the use of the MMSE filter in a dynamic traffic system, such as in packet transmission systems. With decision directed (DD) update of the filter coefficients, we can adapt to the dynamic traffic. Under some extreme cases when the user traffic increases significantly, the pure DD mode may fail or diverge and the condition under which it fails is also derived. To assure convergence, the blind PASTd (Projection Approximation Subspace Tracking) method is first used before switching to the DD mode. It is shown that the combined PASTd and DD based MMSE filter is a robust receiver and can achieve almost the same performance as a receiver with training sequences for every packet, both in the uplink and the downlink. We also show that the receiver behavior is quite different when the traffic increases or decreases.

A new objective function based multi-user detector for DS/CDMA over multipath fading channels

There has been a lot of interest in using multiuser detection (MUD) for combating the detrimental effects of multiple access interference (MAI) in DS/CDMA communication systems. In this paper, we propose a simple and effective MUD receiver over multipath fading channels. It is shown that the proposed receiver can easily take advantage of path diversity. By doing so, significant capacity and performance improvement can be achieved compared with the conventional DS/CDMA receiver. It is also shown that the proposed MUD receiver can outperform the popular parallel interference cancellation (PIC) detector which does not employ a RAKE receiver. Furthermore, it can achieve similar performance as that of the PIC receiver which employs a RAKE with less computation complexity.