Decision feedback differential detection using weighted phase references

An architecture of decision feedback differential phase detection (DF-DPD) of M-ary DPSK signals is proposed. Conventional differential detection of M-ary DPSK signals uses the phase difference between a current symbol and the previous symbol. This method can make the receiver architecture simple. But it has poorer BER performance than coherent detection because it uses the previous noisy signal as the phase reference. To improve the BER performance of conventional differential detection, multiple symbol differential detection methods, such as maximum likelihood differential detection (ML-DD), Viterbi-DD, decision feedback differential detection (DF-DD), and DF-DPD, using L-1 past detected symbols, have been proposed. As L increases, the BER performance of DF-DPD improves but the complexity of the architecture increases dramatically. Thus, this paper proposes the simplified DF-DPD architecture, replacing the conventional delay and addition architecture with one accumulator. The proposed architecture also has a good BER performance with minimizing the phase error of the current received symbol using the accumulated phase differences of all DPD. The simulation results show that the BER performance of the proposed architecture approaches that of coherent detection.

A differential recursive detection (DRD) technique for multiple phase shift keying (MPSK) is proposed, and its performance is analyzed and computer simulated. By combining the multiple-symbol differential detection algorithm with conventional (one-symbol) differential detection, based on maximum-likelihood estimation for a multisymbol observation interval, it is possible to construct a recursive algorithm with a much simpler structure and comparable performance in error rate. Results from computer simulation show 1.25 dB, 2 dB, and 2.55 dB gain over conventional differential detection for the input symbol sequence length of N=3, 5, and 10, respectively, at a bit error rate of 10/sup -5/ in additive white Gaussian noise. For N of more than 10, the performance of DRD is close to that of ideal coherent detection for the differentially encoded signal.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Hamamoto (see ibid., vol.44, no. 8, p.959-66, 1996) presented a detection technique for differential phase-shift keying (DPSK) with improved performance over conventional differential coherent detection. Leib (see ibid., vol.43, no.2/3/4, p.722-25, 1995) comments that this technique had been introduced much earlier. As correctly stated by Hamamoto the essence of the new DPSK detection is based on a technique introduced by Leib and Pasupathy (1988). Differential coherent detection can be viewed as a scheme that uses the previous data symbol as a phase reference. The problem is that this phase reference suffers from the channel noise in very much the same way as the data symbol does and, therefore, it results in a performance degradation.

Approximately 3 dB signal-to-noise ratio (SNR) loss is always paid with conventional differential space-time block codes (STBCs), compared with coherent STBCs. In this paper, a multiple-symbol differential detection (MSDD) technique is proposed for M-ary phase-shift keying (PSK) STBCs. The new scheme can greatly narrow the 3-dB performance gap by extending the observation interval for differential decoding. The technique uses maximum-likelihood sequence detection instead of traditional symbol-by-symbol detection, and is carried out on the slow, flat Rayleigh fading channel. A generalized decision metric is derived for an observation interval of arbitrary length. It is shown that for a moderate number of symbols, MSDD provides approximately 1.5 dB performance improvement over conventional differential detection. In addition, a closed-form pairwise error probability and approximate bit-error probability (BEP) are derived for multiple-symbol differential binary PSK STBC. Results show that the theoretical BEP matches simulation results well. The BEP is shown to converge asymptotically with the number of symbols in the observation interval to that of the differential scheme with coherent detection.

In this paper, noncoherent receivers for differential space-time modulation (DSTM) are investigated. It is shown that the performance of the previously proposed conventional differential detection (DD) receiver is satisfactory only for very slow flat fading channels. However, conventional DD suffers from a considerable loss in performance even for moderately fast fading, especially if more than one transmit antenna is used. In order to overcome this problem, two improved noncoherent receivers are considered. The first one is the multiple-symbol detection (MSD) receiver. Because of the high computational complexity of MSD, also a low-complexity decision-feedback differential detection (DF-DD) receiver is derived. Analytical and simulation results confirm that both receivers perform equally well and can take full advantage of the enhanced diversity provided by multiple transmit antennas even for fast fading.

A novel differential detection technique for multiple phase shift keying (MPSK), which uses multiple-symbol observation interval, is proposed and its performance is analyzed. The proposed technique detects the transmitted symbol with accumulation of the differential signal vector projected on the previously received symbol. As with the multiple-symbol differential detection (MSDD) technique, the performance of the proposed scheme fills the gap between the conventional differential and the ideal coherent detection. The amount of improvement depends on the number of phases M and the number of additional symbol intervals N of observation. Performance analysis and simulation show considerable improvement with only 1 or 2 additional symbol intervals of observation. In addition, the complexity increases linearly with observation interval and is smaller than that of Divsalar and Simon (1990) and of Edbauer (1992).

In distributed space-time coding cooperative systems, differential detection gains much attention because it needs no channel state information at the receivers. But conventional differential detection performs about 3dB loss than coherent detection. In this paper, we propose a multiple-symbol differential detection algorithm for relay networks based on maximum-likelihood principle. The simulation results show that the multiple-symbol differential detection can narrow the performance gap between the conventional differential detection and coherent detection for Alamouti and square real orthogonal coding systems with corresponding numbers of relays, especially in high signal-tonoise ratio regime. And with the increase of observation window size, the Doppler effect on performance is decreased and the error floor encountered in conventional differential detection can be eliminated provided that the value of observation window size is sufficiently high.

In conventional differential detection (CDD), the previous received symbol is used as the phase reference for the current received symbol, therefore the channel variation due to fast fading causes an error floor. We propose a novel iterative decision-directed differential detection (iD4) algorithm for orthogonal frequency division multiplexing (OFDM) systems employing differential PSK modulation in each subcarrier. The core of iD4 is a low complexity time and frequency 2-D estimator which can effectively estimate the channel variation experienced by adjacent time domain symbols. This estimated channel variation is then compensated in the differential detection process. Both theoretical analysis and computer simulations show that significant performance improvement can be achieved using iD4 over the CDD for OFDM receivers in fast fading channels.

In this paper we present decision-feedback differential detection (DF-DD) schemes for autocorrelation-based detection in impulse-radio ultra-wideband (IR-UWB) systems, a signaling scheme regarded as a promising candidate in particular for low-complexity wireless sensor networks. To this end, we first discuss ideal noncoherent sequence estimation and approximations thereof based on block-wise multiple-symbol differential detection (MSDD) and the Viterbi algorithm (VA) from the perspective of tree-search/trellis decoding. Exploiting relations well-known from tree-search decoding, we are able to derive the novel decision-feedback differential detection (DF-DD) schemes. A comprehensive comparison with respect to performance and complexity of the presented schemes in a typical IR-UWB scenario reveals-along with novel insights in techniques for complexity reduction of the sphere decoder applied for MSDD- that sorted DF-DD achieves close-to-optimum performance at very low, and in particular constant receiver complexity.

The performance of conventional and decision feedback differential detection receivers for Gaussian minimum-shift keying (GMSK) signals transmitted in the presence of cochannel interference (CCI) and additive white Gaussian noise (AWGN) is evaluated. For the interference, the authors adopt a model which includes N statistically independent static as well as faded CCI. Various bit-error-rate (BER) performance evaluation results have indicated that the receiver under investigation performs better as compared to other more conventional receiver structures. Especially significant BER improvements are obtained for the static CCI channel. For example, it is shown that with a carrier-to-interference ratio of 14 dB, the performance of a 2-b decision feedback differential receiver outperforms a conventional 2-b differential detector by more than 14 dB (at a BER 10/sup -3/). For the faded CCI, the improvement is less: mainly they result in error-floor reductions of about half an order of magnitude. By comparing the performance of static and faded CCI, it was also found that for a given C/I, the performance of the former would depend on the number of interferers, whereas this is not the case for the latter. >

Multiple-symbol differential detection (MSDD) is a robust maximum-likelihood (ML) receiver technique for frequency-nonselective fast Rayleigh fading channels. However, its complexity grows exponentially with the block size N and that makes it impractical to implement. Recently, multiple-symbol differential sphere decoder (MSDSD) is developed to alleviate this problem but its complexity at low signal-to-noise ratio (SNR) grows exponentially with decreasing SNR. In this paper, we explore the potential of the Fano algorithm as an efficient MSDD receiver. The bit-error performance and the complexity of the Fano-MSDD are evaluated and compared with other reduced complexity techniques. Our preliminary results indicate that Fano-MSDD is more attractive than decision-feedback differential detection (P.Y. Kam and CH Teh, 1983) and (R Schober and WH Gerstacker, 2000) from both the error performance and the complexity points of view. When compared to the MSDSD, the Fano-MSDD suffers a moderate degradation in power efficiency. However, its computational complexity is very steady even at low SNR. This translates into a dramatic saving in complexity over the MSDSD when SNR is low. Even at large SNR, the Fano-MSDD still provides a small edge over the MSDSD in terms of complexity. We believe that with some fine tuning of the decoder parameters, such as the bias and the threshold's step size, it is possible to extract better performance from the Fano decoder than it is now.

The types of manipulated MSK (Minimum Shift Keying) signals obtained by narrowing MSK band which are currently attracting attention include TFM (Tamed Frequency Modulation) and GMSK (Gaussian filtered MSK) signals. A new 2‐bit delay differential detection technique for these manipulated signals is proposed and its performance is compared with the conventional technology. In the proposed technique, a decision feedback circuit which controls the phase references of the delayed signals is used to reduce the intersymbol interference caused by the transmit and receive filters. the proposed technique is especially effective for manipulated MSK signals. Computer simulation results show that the proposed technique requires a smaller receive filter bandwidth and the bit error performance results in an improvement of 3 to 8 dB in the SN ratio compared with the conventional technique.

The performance of multiple symbol differential detection applied to differentially encoded QPSK signals, is examined by simulation in the mobile radio channel with slow Rayleigh and Rician fading, and cochannel interference (CCI). Performance gains observed in the slow Rician fading channel when compared to conventional differential detection, are maintained in the CCI environment. Using a signal quality measure from the multiple symbol differential detection algorithm for selection diversity, performance gains are observed over conventional differential detection in Rayleigh fading as well. >

Motivated by the recent development of Spatial Modulation (SM) and Differential Space-Time Shift Keying (DSTSK), we propose a reduced-complexity Conventional Differential Detector (CDD) as well as a reduced-complexity Multiple-Symbol Differential Sphere Detector (MSDSD) for DSTSK. Both schemes operate on a symbol-by-symbol basis in order to reduce the complexity of the classic block-by-block-based CDD and MSDSD, whilst still approaching the optimum performance of the full-search-based Maximum Likelihood (ML) detector.

Differential distributed space-time coding (DDSTC) technique has been considered for relay networks to provide both diversity gain and high throughput in the absence of channel state information. Conventional differential detection (CDD) or two-symbol non-coherent detection over slow-fading channels has been examined and shown to suffer 3–4 dB loss when compared to coherent detections. Moreover, it has also been shown that the performance of CDD severely degrades in fast-fading channels and an irreducible error floor exists at high signal-to-noise ratio region. To overcome the error floor experienced with fast-fading, a nearly optimal “multiple-symbol” differential detection (MSDD) is developed in this paper. The MSDD algorithm jointly processes a larger window of received signals for detection and significantly improves the performance of D-DSTC in fast-fading channels. The error performance of the MSDD algorithm is illustrated with simulation results under different fading scenarios.