Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with intersymbol interference", IEEE Trans. Inf. Theory, Vol. IT-41, May 1995, pp. 704-713.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....in the VE trellis. The selective use of the channel taps in the VE was also applied in [7] A novel efficient equalizer for sparse channels (called SC MAP) is proposed in this paper, belonging to the symbolby symbol maximum a posteriori (MAP) family of equalizers recently revisited, e.g. in [5], 6] It simultaneously manages three different channel states ( visible , near hidden and far hidden ) to take into account the nonzero CIR coefficients, at the price of a small extra load with respect to the MAP equalizer for nonsparse channels with the same number Paper approved by S. ....

....(sampled at symbol interval ) received sequence is: 1) where is the transmitted data sequence, with (complex) constellation of size , is the CIR, and is a zero mean white noise sequence with spectral power density . The MAP equalizer for time invariant and (generally) nonsparse channels [5], 6] starts from the definition of the channel state vector , which includes the current symbol and can assume different configurations , and the associate a posteriori probability (APP) vector (2) collecting the probabilities of the possible channel states at step , conditional to the ....

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with inter-symbol interference," IEEE Trans. Inform. Theory, vol. 41, pp. 704--713, May 1995.

....therefore suitable for block oriented processing. The algorithm of [74] only requires a forward recursion and is suitable for continuous processing, although it is more complex and generally requires more total storage than the forward backward version of the algorithm. Following the convention of [75], we will denote the block oriented MAP algorithm as the type I MAP algorithm and the continuous MAP algorithm as the type H MAP algorithm. Because the type II MAP algorithm is more complex and requires more storage than the type I MAP algorithm, and because turbo encoding is a block oriented ....

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with intersymbol interference," IEEE Trans. Inform. Theory, vol. 41, pp. 704-713, May 1995.

....research has been done on these schemes. One of the major challenges is to decrease the computational complexity of both the ML MAP equalizer and the convolutional decoder, which is typically also applying a ML MAP detector, using suboptimal algorithms, and ecient use of the equalizer soft output [13]. Stateof the art systems include an interleaver after the encoder and a deinterleaver before the decoder to further improve the BER [13] Interleaving shu es symbols within a given time frame or block of data and thus decorrelates error events introduced by the equalizer between neighboring ....

.... and the convolutional decoder, which is typically also applying a ML MAP detector, using suboptimal algorithms, and ecient use of the equalizer soft output [13] Stateof the art systems include an interleaver after the encoder and a deinterleaver before the decoder to further improve the BER [13]. Interleaving shu es symbols within a given time frame or block of data and thus decorrelates error events introduced by the equalizer between neighboring symbols. The interleaver will play a central role later for iterative solutions. Finally, Ariyavisitakul and Li [14] proposed a joint ....

Y. Li, B. Vucetic, and Y. Sato, \Optimum soft-output detection for channels with intersymbol interference," IEEE Transactions on Information Theory, vol. 41, no. 3, pp. 704-713, May 1995.

.... ) is the vector containing the last L transmitted symbols; and v(i) C is a complex zero mean colored random sequence which accounts for the noise. The L variate random sequence N . i x C 1 A ) is a first order Markov chain known as state sequence of the channel [7]. The random variable X(i) may assume N M distinct and equally likely values , 1 i N , corresponding to the L long ordered sub sequences of constellation symbols, that is: N j , T ) j ( L ) j ( j 1 , 2) where A (1 i L, j N) is the i th component ....

....and estimation problems, such as in multiuser detection, iterative decoding and in all those situations where hard decisions are heavily employed despite the low reliability of the detection process. Moreover, this method makes the use of probabilistic symbol detectors, as the ones described in [7] and references therein, very appealing because they are able to generate three kinds of information: hard information (the hard decisions) soft information (the APPs) and a quasi soft intermediate case represented by the NL MMSE filtered and fixed lag smoothed estimates of the transmitted ....

Y. Li, B. Vucetic, Y. Sato, "Optimum Soft-Output Detection for Channels with Intersymbol Interference", IEEE Trans. Inform. Theory, vol.41, no.3, pp.704-713, May 1995.

....SNR. a) At 50 medium noise that consists of 50 position jitter and 50 width variation. b) At 95 medium noise that consists of 100 position jitter. VI. GENERATING SOFT OUTPUTS The present technique can be incorporated into soft output detectors, such as those studied in [7] 23] and [24]. Softoutput detectors find applications where the outer code is designed to take advantage of the reliability information provided by the channel detector [25] 26] or when iterative decoding is employed [27] 28] Incorporating signal dependent noise prediction into a soft output detector is ....

....modified; incorporating the signal dependent noise prediction, we obtain (41) where and are the noise prediction and its error variance, respectively, for branch . We may write (42) where s can be found from (6) and the error variance can be obtained from (7) Other soft output detectors [23] [24] can be modified using the same principle. VII. REDUCED STATE SOFT OUTPUT DETECTORS WITH SOFT DECISION FEEDBACK In this section, we consider reduced state soft output detectors that make use of soft decision feedback (SDF) The idea of using SDF has been considered in [30] for the AWGN channel. ....

[Article contains additional citation context not shown here]

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with intersymbol interference," IEEE Trans. Inform. Theory, vol. 41, pp. 704--713, May 1995.

....Galli, Student Member IEEE INFO COM Dpt. University of Rome La Sapienza , Via Eudossiana 18, 00184 Rome, Italy This work has been partially supported by the European R D, ACTS MOSTRAIN program. Abstract The performance Symbol by Symbol (SbS) AbendFritchman like combined equalizers decoders [5] performing diversity reception and coherent detection of Trellis CodedModulated (TCM) is analyzed for the case of fast time variant, ISI corrupted, Rayleigh faded data channels. New upperbounds for the resulting decoding error probability are derived and a related novel criterion for the design ....

....Sequence Estimators (MLSEs) at the receiver side. Simulation results [4] recently have demonstrated that SbS detectors can offer better performance on ISI degraded faded environments and today several low complexity algorithms make the implementation of such equalizers decoders very attractive [4,5]. Nevertheless, analytical performance bounds for SbS detectors are yet not available and only simulation results have been reported about SbS detectors performance [4,5] The purpose of this contribution is twofold. First of all, an analytical upper bound for the performance of the above ....

[Article contains additional citation context not shown here]

Y. Li, B. Vucetic, Y. Sato, "Optimum Soft-Output Detection for Channels with Intersymbol Interference", IEEE Trans. on Inform. Theory, vol.41, no.3, pp.704-713, May 1995.

....suggests that an optimum set of PDNP parameters as well as the actual PDNP gain relative to existing detection techniques depend highly on medium noise composition. VI. Generating Soft Outputs The present technique can be incorporated into softoutput detectors such as those studied in [7] 23][24]. Softoutput detectors find applications where the outer code is designed to take advantage of the reliablity information provided by the channel detector [25] 26] or when iterative decoding is employed [27] 28] Incorporating signal dependent noise prediction into a softoutput detector is ....

....where ) k n and ) p are the noise prediction and its error variance, respectively, for branch . We may write ) 42) where ) j q s can be found from (6) and the error variance ) p can be obtained from (7) Other softoutput detectors [23][24] can be modified using the same principle. VII. Reduced State Softoutput Detectors with Soft Decision Feedback In this section, we consider reduced state softoutput detectors that make use of soft decision feedback (SDF) The idea of using SDF has been considered in [30] for the AWGN channel. ....

[Article contains additional citation context not shown here]

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with intersymbol interference," IEEE Trans. Inform. Theory, vol. 41, no. 3, pp. 704-713, May 1995.

.... for the evaluation of (10) and (19) As far as the updating of from is concerned, a straightforward application of the Bayes rule and the total probability theorem followed by the exploitation of the Markovian property of the channel state sequence leads to the usual relationship [10] 11] [19] (20) It is also easy to prove that a reiterated application of the Bayes rule gives rise to the following expression for the computation of from [11] 19] 21) where, as a consequence of the Gaussianity property of the channel noise sequence in (1) the observation dependent diagonal matrix ....

.... theorem followed by the exploitation of the Markovian property of the channel state sequence leads to the usual relationship [10] 11] 19] 20) It is also easy to prove that a reiterated application of the Bayes rule gives rise to the following expression for the computation of from [11] [19]: 21) where, as a consequence of the Gaussianity property of the channel noise sequence in (1) the observation dependent diagonal matrix is composed by exponential terms as follows: 22) In summary, from (5) 12) it can be seen that during the SOM the channel estimator utilizes the ....

[Article contains additional citation context not shown here]

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with intersymbol interference," IEEE Trans. Inform. Theory, vol. 41, pp. 704--713, May 1995.

....experienced for enhanced modems which outperform those currently used in HDSL1 systems [1, Chap. 7] 11] 13] Although in these environments the optimal receiver minimizing the symbol error probability is the so called symbol by symbol maximum a posteriori probability (SbS MAP) equalizer [6] [17], unfortunately its implementation complexity grows (at least) exponentially with the length of the channel impulse response [17] so that its utilization in DSL applications is out of discussion ( 1] Chap. 7) 6] 7] 21] However, the excellent performance achieved by iterative decoding of ....

.... environments the optimal receiver minimizing the symbol error probability is the so called symbol by symbol maximum a posteriori probability (SbS MAP) equalizer [6] 17] unfortunately its implementation complexity grows (at least) exponentially with the length of the channel impulse response [17] so that its utilization in DSL applications is out of discussion ( 1] Chap. 7) 6] 7] 21] However, the excellent performance achieved by iterative decoding of turbocodes [9] 10] has encouraged the application of iterative architectures to other receiving schemes (see, for example, 14] ....

[Article contains additional citation context not shown here]

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with intersymbol interference," IEEE Trans. on Inf. Theory, vol. 43, pp. 704--713, May 1995.

....Schemes Over Fast Time Variant Multipath Degraded Faded Environments 1 E. Baccarelli, R. Cusani, S. Galli INFO COM Dpt. University of Rome La Sapienza , Via Eudossiana 18, 00184 Rome, Italy Abstract The performance Symbol by Symbol (SbS) AbendFritchman like combined equalizers decoders [5] performing diversity reception and coherent detection of Trellis CodedModulated (TCM) is analyzed for the case of fast time variant, ISI corrupted, Rayleigh faded data channels. New upperbounds for the resulting decoding error probability are derived and a related novel criterion for the design ....

....Sequence Estimators (MLSEs) at the receiver side. Simulation results [4] recently have demonstrated that SbS detectors can offer better performance on ISI degraded faded environments and today several low complexity algorithms make the implementation of such equalizers decoders very attractive [4,5]. Nevertheless, analytical performance bounds for SbS detectors are yet not available and only simulation results have been reported about SbS detectors performance [4,5] The purpose of this contribution is twofold. First of all, an analytical upper bound for the performance of the above ....

[Article contains additional citation context not shown here]

Y. Li, B. Vucetic, Y. Sato, "Optimum Soft-Output Detection for Channels with Intersymbol Interference", IEEE Trans. on Inform. Theory, vol.41, no.3, pp.704-713, May 1995.

....S 0090 6778(00)07532 2. SER) and perform soft input soft output detection suitable for iterative (Turbo) decoding [6] 7] However, although there is rich literature on MLSDs for time varying frequency selective channels [4] 5] 8] 11] the MAP research is less complete [6] 12] [15]. In this unified paper, we provide a novel performance analysis of MAPDs in time varying channels, and we coherently and comprehensively address the following practically important topics found also in the MLSD literature: linear modulations, time varying frequency selective channels, ....

....and [11] yet its complexity is notably higher. In addition, the following MAP specific areas are covered: detection when all observations are available (the forward backward MAPDs [13, Appendix] detection when only a fixed number of future observations are available (the fixed lag MAPDs [14] [15]) channel parameter estimation, bit detection, and symbol detection. Bit and symbol detection arise because the MAP criterion may be applied either to the bits into a constellation mapper or its output symbols, as both are discrete. When applied to the output symbols, the result is a conventional ....

[Article contains additional citation context not shown here]

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with intersymbol interference," IEEE Trans. Inform. Theory, vol. 41, pp. 704--713, May 1995.

....[16, Sec. 5.3. 5] 11] 15] An attractive alternative to MLSE s is constituted by the so called symbol by symbol (SBS) maximum a posteriori probability (MAP) detectors, originally developed by Abend and Fritchman [3] Although they offer a well stated performance improvement over MLSE s [17], 18] 21] 22] and could be effectively employed for decoding trellis encoded data [2] 7] 17] 18] for a long time, practical interest in SBS detectors has been limited by their large computational complexity. Paper approved by S. G. Wilson, the Editor for Coding Theory and Applications ....

.... symbol by symbol (SBS) maximum a posteriori probability (MAP) detectors, originally developed by Abend and Fritchman [3] Although they offer a well stated performance improvement over MLSE s [17] 18] 21] 22] and could be effectively employed for decoding trellis encoded data [2] 7] [17], 18] for a long time, practical interest in SBS detectors has been limited by their large computational complexity. Paper approved by S. G. Wilson, the Editor for Coding Theory and Applications of the IEEE Communications Society. Manuscript received March 13, 1995; revised September 2, 1996. ....

[Article contains additional citation context not shown here]

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with inter-symbol interference," IEEE Trans. Inform. Theory, vol. 41, pp. 704--713, May 1995.

....fast timevarying multipath channels that combines the channel estimation and data detection tasks is presented. The a posteriori probabilities (APP s) of the states of the intersymbol interference (ISI) channel are recursively computed from the received data by a symbol by symbol (SbS) detector [7] and are then employed by a Kalman type nonlinear channel estimator. Robust channel tracking and good data detection performance are obtained, with a reasonable receiver complexity. Index Terms Digital communications, fading channels, equalizers, time varying channels. I. THE SYSTEM AND THE ....

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with inter-symbol interference," IEEE Trans. Inform. Theory, vol. 41, pp. 704--713, May 1995.

....between the vectors x and s, the term codeword will be used interchangeably to describe either of the two vectors. C. Sum product decoding by message passing In the literature, several methods exist for soft detection of symbols transmitted over intersymbol interference (ISI) channels [25] 26] [27], 28] 29] 30] There also exist 3 The true code rate of a code de ned by a graph will always be greater or equal to the design code rate. In practice, they are often extremely close, so we do not distinguish between them throughout the paper. February 22, 2001 DRAFT SUBMITTED TO IEEE ....

Y. Li, B. Vucetic, and Y. Sato, \Optimum soft-output detection for channels with intersymbol interference," IEEE Trans. Inform. Theory, vol. 41, pp. 704-713, May 1995.

.... Street, New Jersey 07960, USA Abstract Several contributions have recently shown that adaptive receivers based on Symbol by Symbol Maximum A Posteriori (SbS MAP) equalizers are an attractive alternative to the more common ones based on Maximum Likelihood Sequence (MLS) detectors (see, e.g. [1,6]) However, these two equalizers present advantages that, if suitably exploited in a combined form, can be used to develop high performance adaptive receivers. In this paper, a novel adaptive receiver that merges the advantages of both SbS MAP and MLS receivers is presented. I. AN OVERVIEW ON MLS ....

.... present two opposite features: MLS equalizers deliver reliable data estimates when D is very large (at least five six times the memory d introduced by the channel) 4] whereas SbS MAP equalizers deliver reliable decisions for low values of D (of the order of the memory introduced by the channel) [1]. The optimality criterion on which MLS receivers are based is the minimization of the error probability on a persequence basis. The complexity of this algorithm grows linearly with the sequence length and is optimum, in an MLS sense, when the decision delay D is infinite. However, nearly ....

[Article contains additional citation context not shown here]

Y. Li, B. Vucetic, Y. Sato, "Optimum Soft-Output Detection for Channels with Intersymbol Interference", IEEE Trans. on Inform. Theory, Vol.41, No.3, pp.704-713, May 1995.

....receiver for TDMA systems is presented, which employs a single channel estimator. Its basic architecture is drawn in Fig. 2. Roughly speaking, the proposed adaptive receiver is constituted by a symbol by symbol maximum likelihood (SbSML) detector of the Abend Fritchman family [8, Sec. 6. 6] 12] [13] which recursively delivers reliable hard decisions and feeds a minimum mean square error (MMSE) nonlinear Kalman like recursive channel estimator with the a posteriori probabilities (APP s) of the states of the ISI channel. There are several reasons which support the choice of an SbS ML detector ....

....and feeds a minimum mean square error (MMSE) nonlinear Kalman like recursive channel estimator with the a posteriori probabilities (APP s) of the states of the ISI channel. There are several reasons which support the choice of an SbS ML detector in the scheme of Fig. 2. First, simulation results [13] [15] and analytical bounds [16] confirm that SbS ML detectors are able to deliver reliable final decisions with a limited delay equal to the delay spread (in multiples of the signaling period) of the transmission channel. So, the abovecited tracking capability loss suffered by the conventional ....

[Article contains additional citation context not shown here]

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with intersymbol interference," IEEE Trans. Inform. Theory, vol. 41, pp. 704--713, May 1995.

....core of the receivers consists of an APPcomputer that computes the APP sequence ) i i of the states of the ISI channel on a step by step basis. As far as the recursive computation of the APP vectors is concerned, it can be carried out following, for example, the algorithms proposed in [4] [5]. The actual implementation of the APP computation is not binding for the structure of the proposed receiver. However it is worth pointing out that all implementations share a very important feature: recursion. In fact, vector ) i i can be recursively computed on the basis of ) 1 1 ( i ....

Y. Li, B. Vucetic, Y. Sato, "Optimum Soft-Output Detection for Channels with Intersymbol Interference", IEEE Trans. on Inform. Theory, Vol.41, No.3, pp.704-713, May 1995.

....probabilities (A Posteriori Probabilities, APPs) are computed, the receiver selects as the decided symbol the one that has the highest a posteriori probability. The main drawback of an SbS MAP receiver is due to its high computational complexity that is at least linear in the decision delay D [2] and this has limited their application in practical environments [5,Sect.6.6] However, these receivers have been recently rediscovered because they are able to generate softinformation in the form of APPs and this new interest has led to the derivation of new algorithms for the computation of ....

.... rediscovered because they are able to generate softinformation in the form of APPs and this new interest has led to the derivation of new algorithms for the computation of the APPs that present a computational complexity of the same order of the VA for limited values of the decision delay [2]. In particular, in [6] a new channel estimator that is fed by the APPs of the states of the ISI channel (instead of the less informative hard decisions) and that is able of generating reliable zero delayed channel estimates has been presented. The adaptive receiver presented in [6] is based on ....

Y. Li, B. Vucetic, Y. Sato, "Optimum SoftOutput Detection for Channels with Intersymbol Interference", IEEE Trans. on Inform. Theory, Vol.41, No.3, pp.704-713, May 1995.

....by the receiver [11, Sect. III] Many different SBS MAP detectors have been proposed in the past years, with the purpose of complexity reduction with respect to the original algorithm of [10] which exhibit the same performance and, in principle, are substantially equivalent (see, for example, [11] besides the mentioned [9] We observe here that also the MLSE makes soft statistics available (as in the well known Soft Output VA solutions exploited, for example, in turbo decoding schemes) but again with large delay. Referring now to the channel estimation problem, classic decision driven ....

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with inter-symbol interference," IEEE Trans. Inform. Theory, vol. 41, pp. 704--713, May 1995.

No context found.

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with intersymbol interference", IEEE Trans. Inf. Theory, Vol. IT-41, May 1995, pp. 704-713.

No context found.

Y. Li, B. Vucetic, Y. Sato, "Optimum soft-output detection for channels with intersymbol interference," IEEE Trans. on Information Theory, vol. 41, pp. 704--713, May 1995.

No context found.

Y. Li, B. Vucetic, Y. Sato, "Optimum Soft-Output Detection for Channels with Intersymbol Interference", IEEE Trans. Inf. Theory, vol.41, no.3, pp.704-713, May 1995.

No context found.

Y. Li, B. Vucetic, Y. Sato, "Optimum SoftOutput Detection for Channels with Inter-symbol Interference", IEEE Trans. Inf. Theory, vol.41, no.3, pp.704-713, May 1995.

No context found.

Y. Li, B. Vucetic, and Y. Sato, \Optimum soft-output detection for channels with intersymbol interference," IEEE Transaction on Information Theory, vol. 41, no. 3, pp. 704-13, May 1995.

No context found.

Y. Li, B. Vucetic, and Y. Sato, "Optimum soft-output detection for channels with intersymbol interference," IEEE Trans. Inform. Theory, vol. 41, pp. 704--713, May 1995.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC