Publication number | US20050264906 A1 |

Publication type | Application |

Application number | US 10/853,090 |

Publication date | Dec 1, 2005 |

Filing date | May 25, 2004 |

Priority date | May 25, 2004 |

Also published as | US7702991, US20070189424 |

Publication number | 10853090, 853090, US 2005/0264906 A1, US 2005/264906 A1, US 20050264906 A1, US 20050264906A1, US 2005264906 A1, US 2005264906A1, US-A1-20050264906, US-A1-2005264906, US2005/0264906A1, US2005/264906A1, US20050264906 A1, US20050264906A1, US2005264906 A1, US2005264906A1 |

Inventors | Erich Haratsch |

Original Assignee | Haratsch Erich F |

Export Citation | BiBTeX, EndNote, RefMan |

Patent Citations (28), Referenced by (9), Classifications (13), Legal Events (2) | |

External Links: USPTO, USPTO Assignment, Espacenet | |

US 20050264906 A1

Abstract

A method and apparatus are disclosed for improving the maximum data rate of reduced-state Viterbi detectors with local feedback in magnetic recording systems. A read channel signal is processed in a magnetic recording device by precomputing branch metrics, intersymbol interference estimates or intersymbol interference-free signal estimates for speculative sequences of one or more channel symbols; selecting one of the precomputed values based on at least one decision from at least one corresponding state; and selecting a path having a best path metric for a given state.

Claims(30)

precomputing branch metrics for speculative sequences of one or more channel symbols;

selecting one of said precomputed branch metrics based on at least one decision from at least one corresponding state; and

selecting a path having a best path metric for a given state.

processing a received signal using an FIR equalizer to generate an equalized received signal; and

processing said equalized received signal using a noise-predictive FIR filter to generate said signal.

processing a received signal using an FIR equalizer to generate an equalized received signal; and

processing said equalized received signal using a plurality of noise-predictive FIR filters to generate a plurality of signals.

a branch metric unit for precomputing branch metrics for speculative sequences of one or more channel symbols;

at least one multiplexer for selecting one of said precomputed branch metrics based on at least one decision from at least one corresponding state; and

an add-compare-select unit for selecting a path having a best path metric for a given state.

precomputing intersymbol interference-free signal estimates for speculative sequences of one or more channel symbols;

selecting one of said precomputed intersymbol interference-free signal estimates based on at least one decision from at least one corresponding state; and

selecting a path having a best path metric for a given state.

processing a received signal using an FIR equalizer to generate an equalized received signal; and

processing said equalized received signal using a noise-predictive FIR filter to generate said signal.

processing a received signal using an FIR equalizer to generate an equalized received signal; and

processing said equalized received signal using a plurality of noise-predictive FIR filters to generate a plurality of signals.

an intersymbol interference unit for precomputing intersymbol interference-free signal estimates for speculative sequences of one or more channel symbols;

at least one multiplexer for selecting one of said precomputed intersymbol interference-free signal estimates based on at least one decision from at least one corresponding state; and

an add-compare-select unit for selecting a path having a best path metric for a given state.

Description

- [0001]The present application is related to United States patent application entitled “Method and Apparatus for Generating Filter Tap Weights and Biases for Signal Dependent Branch Metric Computation,” filed on May 6, 2004 and United States patent application entitled “Method and Apparatus for Multiple Step Viterbi Detection with Local Feedback,” United States patent application entitled “Method and Apparatus for Precomputation and Pipelined Selection of Branch Metrics in a Reduced-State Viterbi Detector,” and United States patent application entitled “Method and Apparatus for Precomputation and Pipelined Selection of Intersymbol Interference Estimates in a Reduced-State Viterbi Detector,” each filed contemporaneously herewith, and each incorporated by reference herein.
- [0002]The present invention relates generally to equalization, detection and decoding techniques and, more particularly, to sequence estimation techniques with reduced complexity.
- [0003]A magnetic recording read channel converts an analog read channel into an estimate of the user data recorded on a magnetic medium. Read heads and magnetic media introduce noise and other distortions into the read signal. As the information densities in magnetic recording increase, the intersymbol interference (ISI) becomes more severe as well, (i.e., the channel impulse response becomes longer). In read channel chips, a Viterbi detector is typically used to detect the read data bits in the presence of intersymbol interference and noise. When the channel impulse response is long, however, the hardware complexity associated with the Viterbi detector becomes prohibitively large, as the number of states considered by the Viterbi detector grows exponentially with the length of the channel impulse response. A number of techniques have been proposed or suggested for reducing the complexity of Viterbi detectors.
- [0004]For example, the hardware complexity of the Viterbi detector can be reduced by using a reduced-state trellis that considers only a shortened impulse response, and canceling intersymbol interference due to the tail of the impulse response for each state by using past survivor symbols as local feedback. See, e.g., J. W. M. Bergmans, “Digital Baseband Transmission and Recording,” Kluwer Academic Publishers, 326 (1996) or U.S. Pat. No. 6,690,754, issued to Haratsch et al., entitled “Method and Apparatus for Reducing the Computational Complexity and Relaxing the Critical Path of Reduced-State Sequence Estimation (RSSE) Techniques,” incorporated by reference herein
- [0005]The error rate performance of reduced-state Viterbi detectors with local feedback can approach the performance of full-state Viterbi detectors without local feedback that implement maximum likelihood sequence estimation (MLSE). The maximum achievable data rate of a Viterbi detector implementation with local feedback, however, is considerably lower compared to a Viterbi detector implementation without local feedback, as significantly more operations have to be performed within one clock period. A need therefore exists for a method and apparatus for performing reduced-state Viterbi detection with local feedback at the high data rates that are required by evolving high-end storage applications.
- [0006]Generally, a method and apparatus are disclosed for improving the maximum data rate of reduced-state Viterbi detectors with local feedback in magnetic recording systems. The maximum data rate that may be achieved by the disclosed reduced-state Viterbi detectors is improved by precomputing a number of candidate branch metrics and performing pipelined selection of one of the precomputed values. Precomputing the branch metrics for possible symbol combinations in the channel memory makes it possible to shorten the critical path. In an alternative embodiment, intersymbol interference estimates or intersymbol interference-free signal estimates are precomputed, and branch metric are computed based on the selected precomputed values.
- [0007]A read channel signal is processed in a magnetic recording device by precomputing branch metrics, intersymbol interference estimates or intersymbol interference-free signal estimates for speculative sequences of one or more channel symbols; selecting one of the precomputed values based on at least one decision from at least one corresponding state; and selecting a path having a best path metric for a given state.
- [0008]A more complete understanding of the present invention, as well as further features and advantages of the present invention, will be obtained by reference to the following detailed description and drawings.
- [0009]
FIG. 1 is a schematic block diagram of a conventional system model for a baseband communications channel with ISI and additive noise; - [0010]
FIG. 2 illustrates a trellis diagram for a channel with memory L=1; - [0011]
FIG. 3 illustrates a trellis diagram for a channel having a memory L=4; - [0012]
FIG. 4 illustrates a reduced-state trellis diagram corresponding to the full-state trellis ofFIG. 3 , for a channel having a memory L=4 and a shortened channel memory K=1; - [0013]
FIG. 5 is a schematic block diagram for an exemplary conventional reduced-state Viterbi detector with local feedback; - [0014]
FIG. 6 illustrates a detailed state-parallel implementation of a reduced-state Viterbi detector with local feedback corresponding to the trellis ofFIG. 4 ; - [0015]
FIG. 7 is a schematic block diagram of a reduced-state Viterbi detector that incorporates precomputation of the branch metrics; - [0016]
FIG. 8 is a schematic block diagram showing the selection of a precomputed branch metric by the reduced-state Viterbi detector ofFIG. 7 using survivor symbols; - [0017]
FIG. 9 is a schematic block diagram of a reduced-state Viterbi detector that incorporates precomputation of the ISI-free signal estimates; - [0018]
FIG. 10 is a schematic block diagram showing the selection of a precomputed ISI-free signal estimate by the reduced-state Viterbi detector ofFIG. 9 using survivor symbols; - [0019]
FIG. 11 is a schematic block diagram showing the selection of a precomputed intersymbol interference estimate using survivor symbols; - [0020]
FIG. 12 is a schematic block diagram of a reduced-state Viterbi detector incorporating pipelining of the branch metric selection; - [0021]
FIG. 13 is a schematic block diagram showing the pipelined selection of a branch metric by the reduced-state Viterbi detector ofFIG. 12 using ACS decisions; - [0022]
FIG. 14 is a schematic block diagram of a reduced-state Viterbi detector that incorporates pipelining of the ISI-free signal estimate selection; - [0023]
FIG. 15 is a schematic block diagram showing the pipelined selection of an ISI-free signal estimate by the reduced-state Viterbi detector ofFIG. 14 using ACS decisions; - [0024]
FIG. 16 is a schematic block diagram showing the pipelined selection of an intersymbol interference estimate using ACS decisions; - [0025]
FIG. 17 is a functional block diagram of the ACS operation performed inFIGS. 8, 10 ,**11**,**13**,**15**and**16**; - [0026]
FIG. 18 is a functional block diagram of a read channel detector that implements noise-predictive data detection and uses one of the reduced-state Viterbi detectors ofFIGS. 7-16 incorporating features of the invention; and - [0027]
FIG. 19 is a functional block diagram of a read channel detector that implements signal-dependent noise-predictive data detection and uses one of the reduced-state Viterbi detectors ofFIGS. 7-16 incorporating features of the invention. - [0028]The present invention increases the maximum data rate that may be achieved by reduced-state Viterbi detectors. According to one aspect of the invention, branch metrics, ISI-free signal estimates or ISI estimates are precomputed, and the correct values are selected based on survivor symbols or ACS decisions. In this manner, the computations of ISI estimates, ISI-free signal estimates or branch metrics are removed from the critical path. According to another aspect of the invention, branch metrics, ISI-free signal estimates or ISI estimates can be selected in a pipelined fashion using a multiplexer network structure that corresponds to the structure of the trellis considered by the detector.
- [0029]For a detailed discussion of reduced-state Viterbi detection with local feedback, which is also known as Reduced-State Sequence Estimation (RSSE), (Delayed) Decision-Feedback Sequence Estimation (DFSE), and Parallel Decision-Feedback Equalization (PDFE), see, for example, U.S. Pat. No. 6,690,754 to Haratsch et al., entitled “Method and Apparatus for Reducing the Computational Complexity and Relaxing the Critical Path of Reduced-State Sequence Estimation (RSSE) Techniques,” incorporated by reference herein, and the references cited therein. See also, Lee and Messerschmidt, “Digital Communication,” Kluwer Academic Publishers, 2
^{nd }ed. (1994). - [0030]
FIG. 1 is a schematic block diagram of a conventional system model for a baseband communications channel**100**with ISI and additive noise. While the exemplary embodiment is discussed in the context of baseband communications, the techniques discussed herein can also be applied to passband communications systems, as would be apparent to a person of ordinary skill in the art. Further, while it is assumed that trellis-coded modulation (TCM) is not employed for ease of illustration, the disclosed techniques generalize to communication systems using TCM or other modulation schemes. - [0031]The modulator
**110**maps an information symbol b_{n }into a channel symbol a_{n}. For ease of illustration, it is assumed that the number of information bits per information symbol is one. In other words, the information symbol b_{n }is equivalent to a single information bit b_{n}. The modulator**110**maps an information symbol b_{n }to a two-level channel symbol a_{n }according to following rule:$\begin{array}{cc}{a}_{n}=\{\begin{array}{c}-1,{b}_{n}=0\\ 1,{b}_{n}=1\end{array}.& \left(1\right)\end{array}$ - [0032]The techniques discussed herein can easily be applied to other modulation schemes and more than two signal levels. For a discussion of reduced-state Viterbi detection for an exemplary modulation scheme with five signal levels, see, U.S. patent application Ser. No. 09/471,920, entitled, “Method and Apparatus for Shortening the Critical Path of Reduced Complexity Sequence Estimation Techniques,” incorporated by reference herein.
- [0033]The ISI channel
**100**is modeled as an FIR filter, and the channel output at time n is given by$\begin{array}{cc}{r}_{n}={z}_{n}+{w}_{n}=\sum _{i=0}^{L}{f}_{i}\xb7{a}_{n-i}+{w}_{n},& \left(2\right)\end{array}$

where z_{n }is the ISI channel output, {f_{i}}, 0≦i≦L are the channel coefficients, L is the channel memory, and w_{n }is noise. The decision of a detector**120**that corresponds to b_{n }is denoted by b′_{n}. - [0034]The ISI channel output z
_{n }depends on the current channel symbol a_{n }and the past L transmitted channel symbols {a_{n-i}}, 1≦i≦L. This output can be described as a function of the L past transmitted channel symbols using a finite state machine (FSM) model, where the channel state at time n is defined by

α_{n}=(*a*_{n−1}*, a*_{n−2}*, . . . , a*_{n−L}). (3) - [0035]The channel state is equivalently defined in terms of the L past transmitted information bits:

β_{n}=(*b*_{n−1}*, b*_{n−2}*, . . . , b*_{n−L}). (4) - [0036]It is apparent from equations (3) or (4) that the number of channel states is given by

2^{L}. (5) - [0037]To simplify the notation, the integer value corresponding to the vector (b
_{n−1}, b_{n−L+1}, b_{n−L}) will be used to represent the channel state β_{n}. For example, 0_{n }will stand for β_{n}=(0, . . . , 0, 0), and 1_{n }will stand for β_{n}=(0, . . . , 0, 1). - [0038]The FSM process describing the ISI channel
**100**can be visualized using a trellis diagram**200**, shown inFIG. 2 , for a channel with memory L=1. For the considered exemplary uncoded channel model, a trellis state at time n is denoted by σ_{n}, and is equal to the channel state, i.e., σ_{n}=β_{n}. InFIG. 2 , solid lines correspond to survivor paths, dotted lines to discarded transitions, and dashed lines to path extensions. There are two channel states, and two branches corresponding to the information symbols b_{n}=0 and b_{n}=1 leave each state σ_{n }to reach respective successor states {σ_{n+1}}. It can be seen from equation (5) that the number of channel states grows exponentially with respect to the channel memory. - [0039]
FIG. 2 depicts the operation of the Viterbi algorithm at time step n. At this point, the Viterbi algorithm has already determined the survivor path into state 0_{n}, which corresponds to the surviving state sequence {0_{n}, 1_{n−1}, 0_{n−2}, 1_{n−3}, . . . }. The survivor path into state 1_{n }corresponds in this example to the state sequence {1_{n}, 0_{n−1}, 0_{n−2}, 1_{n−3}, . . . }. Based on these two survivor paths, the Viterbi algorithm decides on the survivor paths into states 0_{n+1 }and 1_{n+1}, in the manner described below. - [0040]First, the Viterbi algorithm calculates branch metrics for the state transitions from σ
_{n }to σ_{n+1}. For a channel with additive white Gaussian noise, the optimum branch metric is the Euclidean distance between the received symbol r_{n }and the ideal ISI channel output z_{n }that corresponds to the respective state transition. For a transition from state σ_{n}, the branch metric is given by$\begin{array}{cc}{\lambda}_{n}\left({\sigma}_{n},{a}_{n}\right)={\left({r}_{n}-{z}_{n}\right)}^{2}={\left({r}_{n}-\sum _{i=0}^{L}{f}_{i}\xb7{a}_{n-i}\right)}^{2},& \left(6\right)\end{array}$

where a_{n }is the channel symbol that is associated with a transition from state σ_{n }to a successor state σ_{n+1}. The techniques described herein are independent from the way branch metrics are computed, i.e., branch metrics can also by computed by using the absolute value of the difference between the received symbol r_{n }and the ideal ISI channel output z_{n}. - [0041]In the trellis
**200**ofFIG. 2 , there are two path extensions into any state σ_{n+1}, e.g., state 0_{n+1 }can be reached from states 0_{n }and**1**_{n}. Out of the two path extensions into a particular state σ_{n+1}, the Viterbi algorithm keeps only the one extension with the smallest path metric, as it corresponds to the most likely path into this state. The metric for the path that emanates from state σ_{n }and leads into σ_{n+1 }is calculated by adding the path metric for the preceding state σ_{n}, Γ_{n}(σ_{n}) and the branch metric λ_{n}(σ_{n}, a_{n}) for the transition. - [0042]The three operations to determine the best survivor path into a new state σ
_{n+1}, i.e., adding up corresponding path metrics of predecessor states σ_{n }and branch metrics for the extensions into the new state σ_{n+1}, comparing the path metrics of these extended sequences, and selecting the extension with the minimum path metric as the survivor sequence for the new state, are referred to as add-compare-select (ACS), which can be described by the following equation:$\begin{array}{cc}{\Gamma}_{n+1}\left({\sigma}_{n+1}\right)=\underset{\left\{{\sigma}_{n}\right\}->{\sigma}_{n+1}}{\mathrm{min}}\left({\Gamma}_{n}\left({\sigma}_{n}\right)+{\lambda}_{n}\left({\sigma}_{n},{a}_{n}\right)\right).& \left(7\right)\end{array}$ - [0043]As previously indicated, the invention can also be applied when branch metrics are computed differently. As known in the art, for certain branch metric definition, the best path into a state is given by the path with the maximum (instead of minimum) path metric. For such cases, the ACS operation described by equation (7) involves a maximum instead of a minimum operation.
- [0044]In the example of
FIG. 2 , the two survivor sequences into states 0_{n }and 1_{n }merge into a single path at time step n−2. In general, all survivor paths merge into a single path after some detection delay D with high probability. Thus, information symbols can be uniquely detected from this time step on. Therefore, it is possible to implement the Viterbi algorithm with a fixed detection delay. It is not required to process the whole transmitted sequence before the first information symbols can be detected. Generally, the detection delay D should be approximately five times the memory of the underlying FSM process. For ISI channels, the memory is equal to L. Typically, a good value for D is determined by running error rate simulations for different values of D. - [0045]
FIG. 3 illustrates a trellis**300**describing an ISI channel having a memory L=4. A trellis state at time n is denoted by σ_{n}, and for the considered exemplary uncoded channel model, it is equal to the channel state, i.e., σ_{n}=β_{n}. There are 16 channel states, and two branches corresponding to the information symbols b_{n}=0 and b_{n}=1 leave each state σ_{n }to reach respective successor states {σ_{n+1}}. - [0046]As indicated above, the disadvantage of MLSE is that its complexity grows exponentially with the channel memory. Considering fewer states for the detection of the most likely data sequence reduces the required hardware or computational effort. Reduced-state Viterbi Detection with local feedback accomplishes this by merging several separate states into one single reduced state and keeping only one survivor path per reduced state. The ISI that is not considered in the reduced state is cancelled for each reduced-state by using channel symbols from the corresponding survivor path in a local feedback fashion. Reduced-state Viterbi detection with local feedback is also known as “Reduced-State Sequence Estimation (RSSE)”, “(Delayed) Decision-Feedback Sequence Estimation”, “Parallel Decision-Feedback Equalization”, etc.
- [0047]In the simplest variant of RSSE, a reduced state β′
_{n }is obtained by not considering all L information symbols, but only the past K information symbols for the definition of a trellis state:

β′_{n}=(*b*_{n−1}*, b*_{n−2}*, . . . , b*_{n−K}), 0*≦K≦L,*(8)

where K is referred to as the truncated channel memory. The number of states in the reduced-state trellis is then given by

2^{K}. (9) - [0048]The reduced state β′
_{n }does not contain information about the ISI caused by the channel symbols (a_{n−K−1}, a_{n−K−2}, . . . , a_{n−L}). Conceptually, this reduced state is obtained by grouping all original states β_{n }as defined in Equation (4) with the same information symbol sequence (b_{n−1}, b_{n−2}, . . . , b_{n−K}), but different sequences (b_{n−K−1}, b_{n−K−2}, . . . , b_{n−L}) into one single reduced state β′_{n}. Therefore, this reduced state does not make any statement about the ISI associated with the channel coefficients (f_{K+1}, f_{K+2}, . . . , f_{L}). But an estimate for this ISI component can be computed by considering the respective channel symbols from the survivor sequence into this state. The ISI corresponding to a state is not known a-priori as in MLSE, but must be determined at each detection step by using channel symbols from the corresponding survivor path. Let σ_{n }denote a state in the reduced-state trellis, i.e., σ_{n}=β′_{n}. The ISI estimate u_{n}(σ_{n}) for a state σ_{n }is calculated at time step n as$\begin{array}{cc}{u}_{n}\left({\sigma}_{n}\right)=\sum _{i=1}^{K}{f}_{i}\xb7{a}_{n-i}+\sum _{i=K+1}^{L}{f}_{i}\xb7{\hat{a}}_{n-i}\left({\sigma}_{n}\right),& \left(10\right)\end{array}$

where â_{n−i}(σ_{n}) is the channel symbol that corresponds to the survivor sequence into state σ_{n }and that is associated with trellis step n−i. The first term on the right hand side of equation (10) computes the ISI component that is known a-priori due to the definition of the reduced state in equation (8). The second term on the right hand side of equation (10) is the ISI component caused by channel taps that were ignored in the reduced-state definition of equation (8). This ISI term is calculated at each detection step for a given state by using respective survivor symbols as local feedback. - [0049]With the ISI estimate u
_{n}(σ_{n}), the branch metric for the transition that emanates from state σ_{n }to reach a successor state σ_{n+1 }and corresponds to channel symbol a_{n }can be computed as:

λ_{n}(σ_{n}*, a*_{n})=(*r*_{n}*−f*_{0}*·a*_{n}*−u*_{n}(σ_{n}))^{2}. (11) - [0050]As in MLSE, the most likely survivor path into the state σ
_{n+1 }with the path metric Γ_{n+1}(σ_{n+1}) among the path extensions from all possible predecessor states {σ_{n}} is determined with an ACS operation:$\begin{array}{cc}{\Gamma}_{n+1}\left({\sigma}_{n+1}\right)=\underset{\left\{{\sigma}_{n}\right\}->{\sigma}_{n+1}}{\mathrm{min}}\left({\Gamma}_{n}\left({\sigma}_{n}\right)+{\lambda}_{n}\left({\sigma}_{n},{a}_{n}\right)\right).& \left(12\right)\end{array}$ - [0051]The version of RSSE where a reduced state is defined by considering just a truncated channel memory as in equation (8) is referred to as (Delayed) Decision-Feedback Sequence Estimation (DFSE), described, for example, in A. Duel-Hallen and C. Heegard, “Delayed Decision-Feedback Sequence Estimation,” IEEE Transaction on Communications, 428-436 (May 1989). A reduced-state trellis can also be constructed by applying set partitioning principles to the channel symbol alphabet, as suggested in M. V. Eyuboglu and S. U. Qureshi, “Reduced-State Sequence Estimation With Set Partitioning and Decision-Feedback,” IEEE Transactions on Communications, 13-20 (January 1988). Recently, even more general rules for the definition of reduced states were given in B. E. Spinnler and J. Huber, “Design of Hyper States for Reduced-State Sequence Estimation,”, AEÜ (Electronics and Communication), 17-26 (1996). The present invention can be applied to such general RSSE methods. In addition, the present invention can be applied to another subclass of RSSE, referred to as Parallel Decision-Feedback Equalization, described in Lee and Messerschmidt, “Digital Communication,” 2
^{nd }ed. (1994). These publications are each incorporated by reference herein. - [0052]Now, RSSE will be explained for the case that L=4 and K=1. Then, a state in the reduced-state trellis is defined according to equation (8) as:

β′_{n}=(*b*_{n−1}) (13)

and the number of states in the reduced-state trellis is equal to 2^{1}=2.FIG. 4 illustrates the reduced-state trellis**400**corresponding to the full state trellis**300**ofFIG. 3 that describes an ISI channel having a memory L=4. A state at time n in the reduced-state trellis is denoted by σ_{n}, i.e., σ_{n}=β′_{n}. There are two channel states, and two branches corresponding to the information symbols b_{n}=0 and b_{n}=1 leave each state σ_{n }to reach respective successor states {σ_{n+1}}. - [0053]
FIG. 5 is a schematic block diagram for an exemplary conventional reduced-state Viterbi detector**500**with local feedback. As shown inFIG. 5 , the reduced-state detector**500**includes a decision-feedback unit (DFU) that computes separate ISI estimates for each trellis state according to equation (10) using local feedback, a branch metric unit (BMU) that computes branch metrics for all transitions, an add-compare-select unit (ACSU) that determines the best survivor path into each state, and a survivor memory unit (SMU) that stores the survivor paths. - [0054]As shown in
FIG. 5 , due to the local feedback the critical path**510**is comprised of a recursive loop that includes each of the processing blocks (i.e., the BMU, ACSU, SMU and DFU). As all operations along this critical path**510**have to be performed within one clock period, this recursive loop limits the maximum achievable data rate. Therefore, the maximum data rate of a reduced-state Viterbi detector with local feedback is significantly lower than the maximum data rate of a Viterbi detector without local feedback, which is only limited by the ACS function. - [0055]
FIG. 6 illustrates a detailed state-parallel reduced-state Viterbi detector implementation**600**with local feedback corresponding to the trellis**400**ofFIG. 4 , having a memory L=4 and a shortened channel memory K=1. â_{n−4}(0_{n}) is the survivor symbol for time step n−4 from the survivor path into state 0_{n}·s_{n+1}(0_{n+1}) is the ACS decision for the two path extensions into state 0_{n+1}. The part of the SMU that stores the L−K survivor symbols â_{n−K−1}(σ_{n}), â_{n−K−2}(σ_{n}), . . . â_{n−L}(σ_{n}) for each reduced state is implemented with a register-exchange-architecture, as these decisions are required for the computation of ISI estimates in the DFU without delay. The implementation of the SMU using a register-exchange architecture is described, e.g., in R. Cypher and C. B. Shung, “Generalized Trace-Back Techniques for Survivor Memory Management in the Viterbi Algorithm,” Journal of VLSI Signal Processing, 85-94 (1993). Because the discussed exemplary channel uses two signal levels, the multipliers in the DFU can be implemented with a shift operation. The squaring operation for the Euclidean distance computation in the BMU can be approximated using random logic or a look-up table. - [0056]Reduced-state Viterbi detection with local feedback that implements, e.g., RSSE, is associated with less computational complexity than full-state Viterbi detection that implements MLSE for the same channel memory L, as it processes less states. However, this comes at the expense of a significantly longer critical path, which is drawn in
FIG. 6 using dotted lines. The critical path comprises one symbol multiplication and L−K additions in the DFU (the first term in the right hand side of equation (10) can be computed outside the loop), one addition, subtraction and squaring operation in the BMU, one add-compare in the ACSU, and a 2-to-1 MUX in the SMU. All the operations along this critical path must be completed within one symbol period and cannot be pipelined. In contrast to this, the critical path in a Viterbi detector just comprises the ACS operation. Therefore, the maximum data rate of a reduced-state Viterbi detector implementation with local feedback is potentially significantly lower compared to a Viterbi detector that performs MLSE. Furthermore, the maximum throughput of a reduced-state Viterbi detector implementation with local feedback depends on the channel memory such that it decreases for increasing L. - [0057]The present invention employs two techniques to increase the maximum data rate that may be achieved by the reduced-state sequence estimator
**500**. First, as discussed below in conjunction withFIGS. 7-11 , branch metrics, ISI-free signal estimates or ISI estimates are precomputed, and the correct values are selected based on survivor symbols or ACS decisions. In this manner, the computations of ISI estimates, ISI-free signal estimates or branch metrics are removed from the critical path. Second, branch metrics, ISI-free signal estimates or ISI estimates can be selected in a pipelined fashion using a multiplexer network structure that corresponds to the structure of the trellis considered by the detector, as shown inFIGS. 12-16 . - [0058]
FIG. 7 is a schematic block diagram of a reduced-state Viterbi detector**700**with local feedback incorporating features of the present invention. As shown inFIG. 7 , the branch metric computation is removed from the critical path by precomputing branch metrics at stage**710**. For a detailed discussion of the precomputation of branch metrics for a trellis, see, for example, U.S. patent application Ser. No. 09/471,920, entitled, “Method and Apparatus for Shortening the Critical Path of Reduced Complexity Sequence Estimation Techniques,” incorporated by reference herein. - [0059]The correct branch metrics are selected at stage
**720**, discussed below in conjunction withFIG. 8 , based on survivor symbols. As previously indicated, the data rate can be increased by precomputing branch metrics and selecting the appropriate ones based on past survivor symbols. As shown inFIG. 7 , a pipeline stage**715**can be inserted between the branch metric precomputation**710**and branch metric selection**720**. In the implementation shown inFIG. 7 , the critical path comprises the branch metric selection**720**, ACSU**730**and SMU**740**. However, the computation of ISI estimates and branch metrics is not part of the critical path in contrast to the conventional reduced-state Viterbi detection implementation shown inFIG. 5 . - [0060]In the exemplary uncoded channel model described above, the input into the reduced-state detector is given by equation (2), and a state in the reduced-state trellis is defined by equation (8). A state in the reduced-state trellis is denoted by σ
_{n}, i.e., σ_{n}=β′_{n}. A branch metric for the transition from state σ_{n }to σ_{n+1 }that corresponds to the information bit sequence b_{n}, b_{n−1}, . . . b_{n−L }is given by$\begin{array}{cc}{\lambda}_{n}\left({b}_{n}{b}_{n-1}\text{\hspace{1em}}\dots \text{\hspace{1em}}{b}_{n-L}\right)={\left({r}_{n}-\sum _{i=0}^{L}{f}_{i}\xb7{a}_{n-i}\right)}^{2},& \left(14\right)\end{array}$

where a_{n−i }is the channel symbol that corresponds to the information bit b_{n−i}. This is the same branch metric that was referred to as λ_{n}(σ_{n}, a_{n}) in the context of equation (11). To account for all possible bit sequences, 2^{L+1 }branch metrics have to be precomputed. For a transition from state σ_{n }to σ_{n+1}, there are 2^{L−K }branch metric candidates that correspond to the same bit sequence b_{n−1}, b_{n−2}, . . . b_{n−K}, which is determined by σ_{n}, but different speculative bit sequences b_{n−K−1}, b_{n−K−2}, . . . b_{n−L}. For each state and transition in the reduced-state trellis, the appropriate branch metric is selected based on the L−K survivor symbols â_{n−K−1}, â_{n−K−2}, . . . â_{n−L }that correspond to this state. The selection of the branch metric associated with the transition from state σ_{n}=0_{n }to σ_{n+1}=0_{n+1 }is shown inFIG. 8 , where L=4 and K=1. - [0061]
FIG. 8 is a schematic block diagram showing the selection of a branch metric**720**using survivor symbols, the ACS operation**730**and the survivor memory operation**740**, as performed by the reduced-state Viterbi detector ofFIG. 7 . Only the selection of one branch metric, and the ACS operation and survivor memory operation for one state are shown. The detector ofFIG. 7 would implement the circuits ofFIG. 8 for all required branch metrics and states. InFIG. 8 , λ_{n}(00xxx) is the selected branch metric for a transition from state 0_{n }that corresponds to the bit sequence b_{n}=0, b_{n−1}=0, b_{n−2}={circumflex over (b)}_{n−2}(0_{n}), b_{n−3}={circumflex over (b)}_{n−3}(0_{n}) and b_{n−4}={circumflex over (b)}_{n−4}(0_{n}). A 2^{L−K}-to-1 multiplexer**810**is required to select the correct branch metric among the precomputed ones. The critical path just comprises the multiplexer**810**for the branch metric selection and an add-compare-select**820**. This is significantly shorter compared to a conventional RSSE implementation, as the computation of branch metrics is outside the critical path. Except for the multiplexer**810**, the critical path in this RSSE architecture with precomputed branch metrics is the same as in a Viterbi detector that implements MLSE without any decision-feedback. - [0062]To reduce the hardware for the precomputation, the branch metrics do not have to be fully precomputed. Instead, ISI-free signal estimates can be precomputed, and branch metrics are then calculated using the correct ISI-free signal estimates that are selected based on past survivor symbols as shown in
FIG. 9 . An ISI-free signal estimate that corresponds to the bit sequence b_{n−1}, b_{n−2}, . . . b_{n−L }is given by$\begin{array}{cc}{q}_{n}\left({b}_{n-1}{b}_{n-2}\text{\hspace{1em}}\dots \text{\hspace{1em}}{b}_{n-L}\right)={r}_{n}-\sum _{i=1}^{L}{f}_{i}\xb7{a}_{n-i}.& \left(15\right)\end{array}$

In total, 2^{L }ISI-free signal estimates have to be precomputed, but only 2^{K+1 }branch metrics are calculated in this architecture. However, the Euclidean distance metric computation,

λ_{n}=(*q*_{n}*−f*_{0}*·a*_{n})^{2 }(16)

is in the critical path, while the computation of ISI-free signal estimates is outside the critical path. - [0063]
FIG. 9 is a schematic block diagram of a reduced-state Viterbi detector**900**incorporating pipelining of the ISI-free signal estimate computation in accordance with the present invention. As shown inFIG. 9 , the ISI-free signal estimate computation is removed from the critical path by precomputing ISI-free signal estimates at stage**910**. - [0064]The correct ISI-free signal estimates are selected at stage
**920**, discussed below in conjunction withFIG. 10 , based on survivor symbols. As previously indicated, the data rate can be increased by precomputing ISI-free signal estimates and selecting the appropriate ones based on past survivor symbols. As shown inFIG. 9 , a pipeline stage**915**can be inserted between the ISI-free signal estimate precomputation**910**and ISI-free signal estimate selection**920**. In the implementation shown inFIG. 9 , the critical path comprises the ISI-free signal estimate selection**920**, branch metrics computation**925**, ACSU**930**and SMU**940**. However, the computation of ISI-free signal estimates is not part of the critical path in contrast to the conventional reduced-state Viterbi detection implementation shown inFIG. 5 . - [0065]
FIG. 10 is a schematic block diagram showing the selection of an ISI-free signal estimate**920**using survivor symbols, the branch metric computation**925**, the ACS operation**930**and the survivor memory operation**940**, as performed by the reduced-state Viterbi detector ofFIG. 9 . Only the selection of one ISI-free signal estimate, the computation of one branch metric, and the ACS operation and survivor memory operation for one state are shown. The detector ofFIG. 9 would implement the circuits ofFIG. 10 for all required ISI-free signal estimates, branch metrics and states. InFIG. 10 , q_{n}(0xxx) is the selected ISI-free signal estimate for a transition from state 0_{n }that corresponds to the bit sequence b_{n−1}=0, b_{n−2}={circumflex over (b)}_{n−2}(0_{n}), b_{n−3}={circumflex over (b)}_{n−3}(0_{n}) and b_{n−4}={circumflex over (b)}_{n−4}(0_{n}). A 2^{L−K}-to-1 multiplexer**1010**is required to select the correct ISI-free signal estimate among the precomputed ones. The critical path just comprises the multiplexer**1010**for the ISI-free signal estimate selection, the branch metric computation**1015**and an add-compare-select**1020**. This is significantly shorter compared to a conventional RSSE implementation, as the computation of the ISI-free signal estimates is outside the critical path. - [0066]To reduce the hardware for the precomputation even further, ISI estimates instead of ISI-free signal estimates can be precomputed, and ISI-free signal estimates and branch metrics are then calculated using the correct ISI estimates that are selected based on past survivor symbols as shown in
FIG. 1 . The inverse of an ISI estimate that corresponds to the bit sequence b_{n−1}, b_{n−2}, . . . b_{n−L }is given by$\begin{array}{cc}{q}_{n}^{\prime}\left({b}_{n-1}{b}_{n-2}\text{\hspace{1em}}\dots \text{\hspace{1em}}{b}_{n-L}\right)=-\sum _{i=1}^{L}{f}_{i}\xb7{a}_{n-i}.& \left(17\right)\end{array}$

In total, 2^{L }ISI estimates have to be precomputed, but only 2^{K }ISI-free signal estimates and 2^{K+1 }branch metrics are calculated in this architecture. An ISI-free signal estimate is then computed according to:

*q*_{n}*=r*_{n}*+q′*_{n}, (18)

and the corresponding branch metric is given by

λ_{n}=(*q*_{n}*−f*_{0}*·a*_{n})^{2}. (19) - [0067]
FIG. 11 is a schematic block diagram of an architecture that is a derivation of the architecture shown inFIG. 10 . InFIG. 11 , ISI estimates are precomputed, whereas inFIG. 10 ISI-free signal estimates are precomputed. InFIG. 11 , an ISI-free signal estimate and branch metric are computed based on a selected ISI estimate. Survivor symbols are used to select the correct ISI estimate. While the architecture ofFIG. 11 is associated with less hardware complexity than the architecture ofFIG. 10 , the critical path inFIG. 11 not only includes the branch metric computation**1115**, but also the computation of an ISI-free signal estimate**1112**. However, the critical path of this architecture is still shorter than the critical path of a conventional reduced-state Viterbi implementation shown inFIG. 6 . - [0068]It is noted that the improved data rate achieved by the present invention comes at the expense of increased hardware complexity, as shown, for example, in
FIGS. 8, 10 and**11**, as several branch metric, ISI-free signal estimate or ISI estimate candidates are precomputed per state transition in the trellis, while only one of these precomputed values is selected for the ACS operation. In contrast, the architecture ofFIG. 6 , that is associated with a significantly longer critical path, computes only one branch metric, ISI-free signal estimate and ISI estimate per state transition. - [0069]In the precomputation architectures of FIGS.
**8**(branch metrics),**10**(ISI-free signal estimates) and**11**(intersymbol interference estimates), a 2^{L−K}-to-1 multiplexer**810**,**1010**,**1110**lies in the critical path. Although the computation of the branch metrics, ISI-free signal estimates or ISI estimates is not part of the critical path anymore, the delay due to the multiplexer**810**,**1010**,**1110**still depends on the channel memory. In a straightforward tree-wise implementation of the 2^{L−K}-to-1 multiplexer**810**,**1010**,**1110**using 2^{L−K}-1 2-to-1 multiplexers, the delay is equal to the delay of L−K 2-to-1 multiplexers, potentially mitigating the speed-up achieved by precomputing branch metrics, ISI-free signal estimates or ISI estimates. - [0070]The present invention recognizes that when branch metrics, ISI-free signal estimates or ISI estimates are precomputed L−K time steps in advance, they can be selected using L−K levels of 2-to-1 multiplexers that are driven by ACS decisions and where each level is associated with a pipeline stage. However, only a single 2-to-1 multiplexer is part of the critical path, and the delay associated with the selection of correct values becomes independent of the channel memory.
- [0071]
FIG. 12 is a schematic block diagram of a reduced-state Viterbi detector**1200**incorporating pipelining of the branch metric selection. As shown inFIG. 12 , the branch metric computation is removed from the critical path by precomputing branch metrics at stage**1210**. The correct branch metrics are selected at stage**1220**, discussed below in conjunction withFIG. 13 , based on ACS decisions. As shown inFIG. 12 , a pipeline stage**1215**can be inserted between the branch metric precomputation**1210**and branch metric selection**1220**. - [0072]
FIG. 13 is a schematic block diagram showing the pipelined selection of a branch metric**1220**by the reduced-state Viterbi detector ofFIG. 12 using ACS decisions. The exemplary embodiment shown inFIG. 13 shows the pipelined selection of branch metrics with three pipeline stages for L=4 and K=1. The critical path in this architecture just includes a 2-to-1 multiplexer, such as a multiplexer in stage**1310**,**1320**or**1330**, and an add-compare**1340**in the ACSU. The critical path is significantly shorter compared toFIG. 8 , where an 8-to-1 multiplexer**810**lies in the critical path. In fact, it has the same length as in a Viterbi detector that implements MLSE without any decision-feedback. - [0073]In
FIG. 13 , branch metrics are precomputed L−K=3 time units in advance and then selected over three clock periods based on corresponding ACS decisions. As the ACS decision s_{n+1}(σ_{n+1}) determines the survivor symbol â_{n+K}(σ_{n+1}), the appropriate branch metric can be selected among candidates that correspond to the same symbol sequence a_{n+L−K}, a_{n+L−K−1}, . . . a_{n−K+1}, but a different past symbol a_{n−K}. InFIG. 13 , s_{n+1}(0_{n+1}) determines the surviving branch metric λ_{n+3}(0000x) in the top multiplexer in stage**1310**among the candidates λ_{n+3}(00000) and λ_{n+3}(00001), where x is a dummy variable for the symbol a_{n−1}. Similarly, s_{n+1}(0_{n+1}) determines the surviving branch metric λ_{n+2}(000xx) in the top multiplexer in stage**1320**among the candidates λ_{n+2}(0000x) and λ_{n+2}(0001x), where xx is a placeholder for the symbols a_{n−1}, and a_{n−2}. Finally, after selection by the multiplexer in stage**1330**, just one surviving branch metric per transition remains, which is used in the ACSU**1340**. The branch metric selection resembles the selection of survivor symbols in a SMU that is implemented according to the register exchange architecture. The total number of 2-to-1 multiplexers is 2^{L−K}−1 and thus equal to the number of 2-to-1 multiplexers required for a straightforward tree-wise implementation of the 8-to-1 multiplexer inFIG. 8 . Therefore, except for pipeline registers, the branch metric selection inFIG. 13 is associated with about the same complexity as the branch metric selection inFIG. 8 . - [0074]Generally, branch metrics are precomputed two or more time units in advance and then selected over two or more clock periods. At each stage
**1310**,**1320**,**1330**shown inFIG. 13 , one of C precomputed branch metric candidates is selected from a plurality of subgroups of B precomputed branch metrics, where each subgroup has an associated multiplexer and pipeline register and where B is less than C. It is noted that in the disclosed embodiments, C is a number of precomputed branch metric candidates associated with each trellis transition and B is equal to a number of transitions per state. As shown inFIG. 13 , at each stage, such as stage**1310**, the selected precomputed branch metrics from each subgroup are combined into one or more new subgroups for a following stage, such as stage**1320**. In the exemplary embodiment shown inFIG. 13 , the subgroups are combined based on the bit pattern of the selected precomputed branch metrics. For example, the selected precomputed branch metrics λ_{n+2}(0000x) and λ_{n+2}(0001x) have similar bit patterns, other than the second-to-least significant bit and are combined for selection in the second stage**1320**. Both branch metrics λ_{n+2}(0000x) and λ_{n+2}(0001x) correspond to a bit pattern that agrees in the first three bit positions. - [0075]
FIG. 14 is a schematic block diagram of a reduced-state Viterbi detector**1400**that incorporates pipelining of the ISI-free signal estimate selection. As shown inFIG. 14 , the ISI-free signal estimate computation is removed from the critical path by precomputing ISI-free signal estimates at stage**1410**. The correct ISI-free signal estimates are selected at stage**1420**, discussed below in conjunction withFIG. 15 , based on ACS decisions. As shown inFIG. 14 , a pipeline stage**1415**can be inserted between the ISI-free signal estimate precomputation**1410**and ISI-free signal estimate selection**1420**. - [0076]
FIG. 15 is a schematic block diagram showing the pipelined selection of an ISI-free signal estimate**1420**by the reduced-state Viterbi detector ofFIG. 14 using ACS decisions. The selection of ISI-free signal estimates resembles the selection of branch metrics, discussed above in conjunction withFIG. 13 . The exemplary embodiment shown inFIG. 15 shows the pipelined selection of an ISI-free signal estimate with three pipeline stages for L=4 and K=1. The critical path in this architecture just includes a 2-to-1 multiplexer, such as a multiplexer in stage**1510**,**1520**or**1530**, an add-compare**1540**in the ACSU and a branch metric computation**1535**. The critical path is significantly shorter compared toFIG. 10 , where an 8-to-1 multiplexer**1010**lies in the critical path. - [0077]In
FIG. 15 , ISI-free signal estimates are precomputed L−K=3 time units in advance and then selected over three clock periods based on corresponding ACS decisions. As the ACS decision s_{n+1}(σ_{n+1}) determines the survivor symbol â_{n−K}(σ_{n+1}), the appropriate ISI-free signal estimate can be selected among candidates that correspond to the same symbol sequence a_{n+L−K}, a_{n+L−K−1}, . . . a_{n−K+1}, but a different past symbol a_{n−K}. InFIG. 15 , s_{n+1}(0_{n+1}) determines the surviving intersymbol interference estimate q_{n+3}(000x) in the top multiplexer in stage**1510**among the candidates q_{n+3}(0000) and q_{n+3}(0001), where x is a dummy variable for the symbol a_{n−1}. Similarly, s_{n+1}(0_{n+1}) determines the surviving branch metric q_{n+2}(00xx) in the top multiplexer in stage**1520**among the candidates q_{n+2}(000x) and q_{n+2}(001x), where xx is a placeholder for the symbols a_{n−1 }and a_{n−2}. Finally, after selection by the multiplexer in stage**1530**, just one surviving ISI-free signal estimate per transition remains, which is used for the branch metric computation**1535**. - [0078]
FIG. 16 is a schematic block diagram showing the pipelined selection of an intersymbol interference estimate using ACS decisions according to the invention. InFIG. 16 , ISI estimates instead of ISI-free signal estimates are selected in a pipelined fashion. The ISI estimates are precomputed according to equation (17). The intersymbol interference estimate selection resembles the selection of the ISI-free signal estimate, discussed above in conjunction withFIG. 15 . The exemplary embodiment shown inFIG. 16 shows the pipelined selection of intersymbol interference estimates with three pipeline stages for L=4 and K=1. The critical path in this architecture just includes a 2-to-1 multiplexer, such as a multiplexer in stage**1610**,**1620**or**1630**, an ISI-free signal estimate computation**1632**, a branch metric computation**1635**, and an add-compare**1640**in the ACSU. The critical path is significantly shorter compared toFIG. 11 , where an 8-to-1 multiplexer**1110**lies in the critical path. - [0079]In
FIG. 16 , intersymbol interference estimates are precomputed L−K=3 time units in advance and then selected over three clock periods based on corresponding ACS decisions. As the ACS decision s_{n+1}(σ_{n+1}) determines the survivor symbol ân−K(σ_{n+1}), the appropriate intersymbol interference estimate can be selected among candidates that correspond to the same symbol sequence a_{n+L−K}, a_{n+L−K−1}, . . . a_{n−K+1}, but a different past symbol a_{n−K}. InFIG. 16 , s_{n+1}(0_{n+1}) determines the surviving intersymbol interference estimate q′_{n+3}(000x) in the top multiplexer in stage**1610**among the candidates q′_{n+3}(0000) and q′_{n+3}(0001), where x is a dummy variable for the symbol a_{n−1}. Similarly, s_{n+1}(0_{n+1}) determines the surviving branch metric q′_{n+2}(00xx) in the top multiplexer in stage**1620**among the candidates q′_{n+2}(000x) and q′_{n+2}(001X), where xx is a placeholder for the symbols a_{n−1 }and a_{n−2}. Finally, after selection by the multiplexer in stage**1630**, just one surviving intersymbol interference estimate per transition remains, which is used for the ISI-free signal estimate computation**1632**. - [0080]
FIG. 17 is a functional block diagram**1700**of the ACS operation performed inFIGS. 8, 10 ,**11**,**13**,**15**, and**16**. As shown inFIG. 17 , the ACS block**1700**includes an add function**1710**, compare function**1720**and select function**1730**. The exemplary add function**1710**includes two adders. The exemplary compare function**1720**could be implemented using a subtractor, where the sign bit of the subtractor output controls the selector**1730**. The select function**1730**comprises a multiplexer, controlled by the output of the compare function**1720**.FIG. 17 shows an exemplary two-way ACS implementation for a trellis with two transitions per state. A 4-way ACS structure for a trellis with four transitions per state is shown in United States patent application entitled “Method and Apparatus for Multiple Step Viterbi Detection with Local Feedback,” filed simultaneously herewith, assigned to the assignee of the present invention and incorporated by reference herein. - [0081]Among other benefits, the present invention allows for a VLSI implementation of reduced-state Viterbi detectors with local feedback for data rates that are significantly increased relative to conventional designs. Even larger data rate increases can be achieved when two or more trellis steps are processed within once clock period using a multi-step trellis. To achieve this additional speed advantage, the invention disclosed here can be combined with the multi-step detection method disclosed in United States patent application entitled “Method and Apparatus for Multiple Step Viterbi Detection with Local Feedback,” filed contemporaneously herewith and incorporated by reference herein. The invention uses an architecture that is very regular making it suitable for high-speed implementation. Viterbi detectors with local feedback can achieve better error rate performance than postprocessor-based structures in the magnetic recording application. Therefore, reduced-state Viterbi detection with local feedback is an attractive detector structure for future read channel chips. The use of reduced-state Viterbi detection with local feedback in the magnetic recording application is described in E. F. Haratsch, “Viterbi Detector Architectures for Magnetic Recording,” 2003 International Symposium on VLSI Technology, Systems, and Applications, 243-46, Oct. 6-8, 2003. Post-processor based detector structures are discussed in Z. A. Keirn et al., “On the Use of Redundant Bits for Magnetic Recording: Single Parity Codes and Reed-Solomon ECC,” IEEE Transactions on Magnetics, 225-30 (January 2004), and the references therein.
- [0082]Another benefit of the invention is that ACS decisions can be used to select precomputed branch metrics, ISI-free signal estimates or ISI estimates as shown in
FIGS. 12-16 . When ACS decisions are used to select precomputed values, the SMU**1240**and**1440**can be implemented using a trace-back structure, as survivor symbols are not used for local feedback anymore. The details of a trace-back survivor memory architecture can be read, e.g. in R. Cypher and C. B. Shung, “Generalized Trace-Back Techniques for Survivor Memory Management in the Viterbi Algorithm,” Journal of VLSI Signal Processing, 85-94 (1993), or in H.-L. Lou, “Implementing the Viterbi algorithm”, IEEE Signal Processing Magazine, 42-52 (September 1995), or in O. J. Joeressen and H. Meyr, “Viterbi Decoding with Dual Timescale Traceback Processing,” IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, 213-217 (September 1995), each incorporated by reference herein. - [0083]In a register-exchange survivor memory implementation, survivor symbols for each state are stored and updated at each detection step. In a trace-back implementation, however, ACS decisions are stored as pointers in a memory, and the detected symbols are obtained by tracing back the pointers that correspond to a survivor path. As the trace-back architecture does not require the updating of all survivor symbols at each detection step, it is associated with less power consumption than the register-exchange architecture. However, the trace-back architecture is associated with larger detection latency and therefore not suitable for the reduced-state Viterbi detector shown in
FIG. 6 , where zero delay survivor symbols are required for the local feedback to compute ISI estimates and branch metrics. However, the disclosed architectures shown inFIGS. 12-16 use ACS decision to select precomputed branch metrics, ISI-free signal estimates or ISI estimates, therefore making it possible to implement the survivor memory SMU using a trace-back architecture. In this case, the trace-back SMU will be associated with significantly less power consumption than a corresponding register-exchange SMU implementation. - [0084]The techniques described herein can be employed, e.g., to detect data in the presence of intersymbol interference and noise in magnetic recording read channels. The disclosed reduced-state Viterbi detectors with local feedback improve the detection of read data bits compared to post-processor based structures. In particular, the invention can be used to implement a read channel that performs noise-predictive data detection and achieves the ever increasing high data rates that are required by evolving storage applications. For a discussion of noise-predictive detection in magnetic recording, see, e.g., R. D. Cideciyan et al., “Noise Predictive Maximum Likelihood Detection Combined With Parity-Based Post-Processing,” IEEE Trans. on Magnetics, 714-20 (March 2001), and E. F. Haratsch, “Viterbi Detector Architectures for Magnetic Recording,” International Symposium on VLSI Technology, Systems, and Applications, 243-46 (October 2003).
- [0085]The simplified block diagram for a read channel incorporating noise-predictive reduced-state Viterbi detection is shown in
FIG. 18 , where signals received at the input of the finite response (FIR) equalizer are in fact signals that have been processed by the analog front-end, which typically includes a variable gain amplifier, continuous time filter and A/D converter. The FIR equalizer**1810**shapes the channel impulse response such that the signals at the output of the FIR equalizer y_{n }can be described by the equation:$\begin{array}{cc}{y}_{n}=\sum _{i=0}^{M}{h}_{i}\xb7{a}_{n-i}+{v}_{n},& \left(20\right)\end{array}$

where a_{n }are the data symbols defined as in equation (1), h_{i }are the equalization target coefficients, M is the equalization target order, and v_{n }is the noise at the output of the FIR equalizer. The equalization target is chosen such that its frequency spectrum matches the characteristics of the read channel well. The impulse response associated with the equalization target can be described by the equation:

*H*(*D*)=*h*_{0}*+h*_{1}*·D+h*_{2}*·D*^{2}*+ . . . +h*_{M}*·D*^{M}. (21) - [0086]The error rate performance of a read channel can be improved by employing a noise-predictive FIR (NP-FIR) filter
**1820**after the FIR equalizer**1810**that whitens the noise. The impulse response associated with the NP-FIR can be characterized with the polynomial:

*P*(*D*)=*p*_{0}*+p*_{1}*·D+p*_{2}*·D*^{2}*+ . . . +p*_{N}*·D*^{N}, (22)

where p_{i}, 0≦i≦N are the coefficients and N is the order of the NP-FIR filter. - [0087]The subsequent reduced-state Viterbi detector considers a channel response with the polynomial:

*F*(*D*)=*f*_{0}*+f*_{1}*·D+f*_{2}*·D*^{2}*+ . . . +f*_{M+N}*·D*^{M+N}*=H*(*D*)·*P*(*D*), (23)

and the signals at the input of the reduced-state Viterbi detector are given by:$\begin{array}{cc}{r}_{n}=\sum _{i=0}^{L}{f}_{i}\xb7{a}_{n-i}+{w}_{n},& \left(24\right)\end{array}$

where f_{i}, 0≦i≦L, are the channel coefficients of the channel seen by the reduced-state Viterbi detector, L is the channel memory, and w_{n }is the remaining noise at the output of the NP-FIR filter**1820**. The channel memory L would be typically equal to L=M+N, but the reduced-state detector could also consider a channel with shorter channel memory, i.e. L<M+N. The channel coefficients f_{i }are given by the convolution of the equalization target and the impulse response of the NP-FIR filter**1820**(see equation (23)). Equation (24) is equivalent to equation (2). Therefore, the features of the invention can be applied to the read channel application, i.e., a reduced-state can be defined as in equation (8), branch metrics can be precomputed as in equation (14) using the channel coefficients defined by (23), and an ACS operation can be performed as in equation (12). A correct branch metric for a transition in the trellis can be selected using the architectures shown inFIG. 8 orFIG. 13 . Instead of branch metrics, ISI-free signal estimates or ISI estimates can be precomputed according to equations (15) and (17) respectively using the channel coefficients defined by (23), and a correct ISI-free signal estimate or ISI estimate can be selected as shown inFIGS. 10, 11 ,**15**and**16**respectively. Therefore, the reduced-state Viterbi detector**1830**can be implemented using the architectures shown inFIGS. 7-16 . - [0088]The invention can also be applied when a parity check code is used to achieve coding gain. For example, when a one-bit parity check code is used, a state σ
_{n }in the full-state trellis is given by:

σ_{n}=(*s*_{n−1}*; b*_{n−1}*, b*_{n−2}*, . . . , b*_{n−L}), (25)

where s_{n }is the running parity syndrome given by the XOR-sum:

*s*_{n}*=b*_{n}*⊕s*_{n−1}. (26)

The total number of states in the reduced-state trellis that accounts for the parity check code is given by:

2×2^{L}. (27) - [0089]Analogous to equation (8), a state σ
_{n }in the reduced-state trellis can be defined by considering only the past K information bits or symbols:

σ_{n}=(*s*_{n−1}*; b*_{n−1}*, b*_{n−2}*, . . . , b*_{n−K}), (28)

and the number of states in the reduced-state trellis is equal to:

2×2^{K}. (29) - [0090]A conventional implementation of a reduced-state Viterbi detector that considers the reduced-state trellis defined according to equation (28) would use equations (10)-(12) to compute ISI estimates, branch and path metrics. However, due to the local feedback required for the computation of ISI estimates, it is challenging to achieve very high data rates. However, the maximum achievable data rate can again be increased significantly by precomputing branch metrics and selecting correct ones as described above. Alternatively, ISI-free signal estimates or ISI estimates can be precomputed and correct ones can be selected as described above.
- [0091]Branch metrics can be precomputed for possible bit sequences according to equation (14). As described above for L=4 and K=1, the required branch metric λ
_{n}(b_{n}b_{n−1}xxx) for a transition from state σ_{n}=(s_{n−1}; b_{n−1}) is selected among the precomputed branch metrics λ_{n}(b_{n}b_{n−1}000), λ_{n}(b_{n}b_{n−1}001), λ_{n}(b_{n}b_{n−1}010), λ_{n}(b_{n}b_{n−1}011), λ_{n}(b_{n}b_{n−1}100), λ_{n}(b_{n}b_{n−1}101), λ_{n}(b_{n}b_{n−1}110), and λ_{n}(b_{n}b_{n−1}111) based on survivor symbols or ACS decisions. - [0092]The invention can also be applied to signal-dependent detection, which is sometimes referred to as data-dependent detection and explained in detail in the co-pending United States patent application entitled “Method and Apparatus for Generating Filter Tap Weights and Biases for Signal Dependent Branch Metric Computation,” incorporated by reference herein. In signal-dependent detection, more than one signal-dependent (SD) NP-FIR filters operate in parallel to whiten the noise.
FIG. 19 illustrates this for the case that two SD NP-FIR filters**1920**-**1**and**1920**-**2**are used. The invention can easily be used when there are more than two NP-FIR filters**1920**. The reduced-state Viterbi detector**1930**can be implemented using the architectures shown inFIGS. 7-16 . - [0093]In
FIG. 19 , the output of the FIR equalizer**1910**is supplied to two SD NP-FIR filters**1920**-**1**and**1920**-**2**to produced two signals r_{n}(1) and r_{n}(2). Each SD NP-FIR filter**1920**implements the impulse response defined by equation (22) with a different set of filter coefficients. For example, the first SD NP-FIR filter**1920**-**1**that produces r_{n}(1) uses a first set of coefficients p_{i}(1), 0≦i≦N, whereas the second SD NP-FIR filter**1920**-**2**that produces r_{n}(2) uses a second set of coefficients p_{i}(2), 0≦i≦N that can differ from the first set of NP-FIR filter coefficients. The corresponding polynomials that describe the SD-NP FIR filters**1920**are denoted P(D;1) and P(D;2), e.g.,

*P*(*D;*1)=*p*_{0}(1)+*p*_{1}(1)·*D+ . . . p*_{N}(1)·*D*^{N}. (30)

The filter coefficients of the different SD NP-FIR filters**1920**can differ, as in a signal-dependent channel the noise statistics depend on the transmitted data or bit sequence. The generation of filter coefficients for the SD NP-FIR filters is described in co-pending United States patent application entitled “Method and Apparatus for Generating Filter Tap Weights and Biases for Signal Dependent Branch Metric Computation,” incorporated by reference herein. - [0094]For the considered channel with two SD NP-FIR filters
**1920**, the reduced-state Viterbi detector**1930**would compute branch metrics considering two different channel impulse responses with the polynomials F(D;1) and F(D;2) that are given by:

*F*(*D;*1)=*f*_{0}(1)+*f*_{1}(1)·*D+ . . . f*_{M+N}(1)·*D*^{M+N}*=H*(*D*)·*P*(*D;*1), and (31)

*F*(*D;*2)=*f*_{0}(2)+*f*_{1}(2)·*D+ . . . f*_{M+N}(2)·*D*^{M+N}*=H*(*D*)·*P*(*D;*2). (32) - [0095]In a signal-dependent channel, the filter coefficients f
_{i }that are used to compute a branch metric depend on the transmitted data or bit sequence. For two SD NP-FIR filters**1920**, branch metrics are precomputed according to equation (14) for a first group of bit sequences (b_{n−1}b_{n−2 }. . . b_{n−L}) using filter coefficients f_{i}(1), and for a second group of bit sequences (b_{n−1}b_{n−2 }. . . b_{n−L}) the filter coefficients f_{i}(2) are used. - [0096]For example, signal-dependent branch metrics for transitions from states σ
_{n }to σ_{n+1 }that correspond to all bit sequences starting with (b_{n}b_{n−1})=(00) or (b_{n}b_{n−1})=(11) are computed using channel coefficients f_{i}(1) and the sample r_{n}(1):$\begin{array}{cc}{\lambda}_{n}\left(00{b}_{n-2}{b}_{n-3}\text{\hspace{1em}}\dots \text{\hspace{1em}}{b}_{n-L}\right)={\left({r}_{n}\left(1\right)-\sum _{i=0}^{L}{f}_{i}\left(1\right)\xb7{a}_{n-i}\right)}^{2},& \left(33\right)\\ {\lambda}_{n}\left(11{b}_{n-2}{b}_{n-3}\text{\hspace{1em}}\dots \text{\hspace{1em}}{b}_{n-L}\right)={\left({r}_{n}\left(1\right)-\sum _{i=0}^{L}{f}_{i}\left(1\right)\xb7{a}_{n-i}\right)}^{2}.& \left(34\right)\end{array}$ - [0097]Continuing this example, signal-dependent branch metrics for all bit sequences that start with (b
_{n}b_{n−1})=(01) or (b_{n}b_{n−1})=(10) are computed using the second of channel coefficients f_{i}(2) and the second sample r_{n}(2):$\begin{array}{cc}{\lambda}_{n}\left(01{b}_{n-2}{b}_{n-3}\text{\hspace{1em}}\dots \text{\hspace{1em}}{b}_{n-L}\right)={\left({r}_{n}\left(2\right)-\sum _{i=0}^{L}{f}_{i}\left(2\right)\xb7{a}_{n-i}\right)}^{2},& \left(35\right)\\ {\lambda}_{n}\left(10{b}_{n-2}{b}_{n-3}\text{\hspace{1em}}\dots \text{\hspace{1em}}{b}_{n-L}\right)={\left({r}_{n}\left(2\right)-\sum _{i=0}^{L}{f}_{i}\left(2\right)\xb7{a}_{n-i}\right)}^{2}.& \left(36\right)\end{array}$ - [0098]Comparing the signal-dependent branch metric equations (33)-(36) with the non signal-dependent branch metric equation (14), signal-dependent branch metrics are computed using the same underlying function, but the choice of the sample r
_{n }and channel coefficients f_{i }depends on the bit sequence for which the branch metric is computed. The precomputation of signal-dependent branch metrics was illustrated here using two signal-dependent NP-FIR filters and a particular grouping of bit sequences, but it is apparent how signal-dependent branch metrics are precomputed for more than two signal-dependent NP-FIR filters and other groupings. E.g., all possible bit sequences of length L can be divided into more than two groups, for which separate samples r_{n }and separate sets of channel coefficients f_{i }would be used to precompute branch metrics. - [0099]The selection of the correct branch metric, the ACS operation and the SMU are implemented as described above for the non-signal dependent detector. Therefore all the benefits of the invention apply to signal-dependent detection as well. In a further variation of the present invention, signal-dependent ISI-free signal estimates or intersymbol interference estimates can be precomputed, instead of branch metric, as would be apparent to a person of ordinary skill in the art based on the present disclosure. Then, signal-dependent branch metrics would be computed based on selected signal-dependent ISI-free signal estimates or ISI estimates as described above.
- [0100]It is to be understood that the embodiments and variations shown and described herein are merely illustrative of the principles of this invention and that various modifications may be implemented by those skilled in the art without departing from the scope and spirit of the invention.

Patent Citations

Cited Patent | Filing date | Publication date | Applicant | Title |
---|---|---|---|---|

US5136593 * | Oct 30, 1989 | Aug 4, 1992 | Carnegie-Mellon University | Apparatus and method for fixed delay tree search |

US5220570 * | Nov 30, 1990 | Jun 15, 1993 | The Board Of Trustees Of The Leland Stanford Junior University | Programmable viterbi signal processor |

US5291499 * | Mar 16, 1992 | Mar 1, 1994 | Cirrus Logic, Inc. | Method and apparatus for reduced-complexity viterbi-type sequence detectors |

US5844946 * | May 24, 1996 | Dec 1, 1998 | Mitsubishi Denki Kabushiki Kaisha | Soft-decision receiver and decoder for digital communication |

US5870433 * | Dec 4, 1996 | Feb 9, 1999 | Ke Kommunikations-Eletronik Gmbh & Co. | Method of processing signals in a viterbi decoder according to delayed decision feedback sequence estimation (DDFSE) algorithm |

US5881106 * | Aug 30, 1995 | Mar 9, 1999 | Sgs-Thomson Microelectronics S.A. | Signal processing circuit to implement a Viterbi algorithm |

US5910968 * | Feb 18, 1997 | Jun 8, 1999 | U.S. Philips Corporation | Continuous phase modulation digital communication system |

US5970104 * | Mar 19, 1997 | Oct 19, 1999 | Cadence Design Systems, Inc. | Method and apparatus for generating branch metrics and branch indices for convolutional code Viterbi decoders |

US6035006 * | Jun 30, 1997 | Mar 7, 2000 | Nec Corporation | Estimator having a feedback loop |

US6088404 * | Apr 22, 1998 | Jul 11, 2000 | Daewoo Electronics Co., Ltd. | Method and apparatus for decoding trellis code data |

US6201831 * | Nov 12, 1999 | Mar 13, 2001 | Broadcom Corporation | Demodulator for a multi-pair gigabit transceiver |

US6291523 * | Aug 25, 1998 | Sep 18, 2001 | Novartis Ag | Certain 5-alkyl-2-arylaminophenylacetic acids and derivatives |

US6374387 * | Aug 19, 1999 | Apr 16, 2002 | Institute Of Microelectronics | Viterbi decoding of punctured convolutional codes without real-time branch metric computation |

US6396254 * | Dec 6, 1999 | May 28, 2002 | Cirrus Logic, Inc. | Read channel |

US6467064 * | Dec 15, 1999 | Oct 15, 2002 | Fujitsu Limited | Viterbi decoder |

US6690739 * | Dec 26, 2000 | Feb 10, 2004 | Shou Yee Mui | Method for intersymbol interference compensation |

US6701483 * | Aug 18, 2000 | Mar 2, 2004 | Alantro Communications, Inc. | Fast search-based decoding scheme |

US6954841 * | Sep 13, 2002 | Oct 11, 2005 | International Business Machines Corporation | Viterbi decoding for SIMD vector processors with indirect vector element access |

US6999521 * | Dec 23, 1999 | Feb 14, 2006 | Lucent Technologies Inc. | Method and apparatus for shortening the critical path of reduced complexity sequence estimation techniques |

US7000175 * | Apr 13, 2001 | Feb 14, 2006 | Agere Systems Inc. | Method and apparatus for pipelined joint equalization and decoding for gigabit communications |

US20020073377 * | Dec 7, 2000 | Jun 13, 2002 | Ara Patapoutian | Ameliorating the adverse impact of burst errors on the operation of ISI detectors |

US20020083396 * | Apr 13, 2001 | Jun 27, 2002 | Kameran Azadet | Method and apparatus for pipelined joint equalization and decoding for gigabit communications |

US20030120993 * | Dec 21, 2001 | Jun 26, 2003 | Texas Instruments Incorporated | Viterbi decoder using restructured trellis |

US20030123585 * | Dec 19, 2002 | Jul 3, 2003 | Kuang-Yu Yen | Receiver with DFE and viterbi algorithm for block code transmission |

US20040032683 * | Mar 28, 2003 | Feb 19, 2004 | Ashley Jonathan J. | Method and apparatus for calibrating data-dependent noise prediction |

US20040133843 * | Dec 22, 2003 | Jul 8, 2004 | Kabushiki Kaisha Toshiba | Digital signal decoding device and digital signal decoding method |

US20050105658 * | Oct 8, 2004 | May 19, 2005 | Haratsch Erich F. | Pipelined decision-feedback unit in a reduced-state viterbi detector with local feedback |

US20050268211 * | May 25, 2004 | Dec 1, 2005 | Haratsch Erich F | Method and apparatus for precomputation and pipelined selection of branch metrics in a reduced-state Viterbi detector |

Referenced by

Citing Patent | Filing date | Publication date | Applicant | Title |
---|---|---|---|---|

US7673224 | Sep 12, 2006 | Mar 2, 2010 | Agere Systems Inc. | Low power viterbi decoder using a novel register-exchange architecture |

US7702991 * | Mar 27, 2007 | Apr 20, 2010 | Agere Systems Inc. | Method and apparatus for reduced-state viterbi detection in a read channel of a magnetic recording system |

US8149529 * | Jul 28, 2010 | Apr 3, 2012 | Lsi Corporation | Dibit extraction for estimation of channel parameters |

US8867154 * | Jun 6, 2013 | Oct 21, 2014 | Lsi Corporation | Systems and methods for processing data with linear phase noise predictive filter |

US8873177 | Sep 30, 2011 | Oct 28, 2014 | Lsi Corporation | Hardware-based methods and apparatus for inter-track interference mitigation in magnetic recording systems |

US8902524 | Mar 15, 2013 | Dec 2, 2014 | Lsi Corporation | Inter-track interference mitigation in magnetic recording systems using averaged values |

US20070189424 * | Mar 27, 2007 | Aug 16, 2007 | Agere Systems Inc | Method and apparatus for reduced-state viterbi detection in a read channel of a magnetic recording system |

US20080072127 * | Sep 12, 2006 | Mar 20, 2008 | Tuhin Subhra Chakraborty | Low power viterbi decoder using a novel register-exchange architecture |

EP2590169A1 * | Sep 26, 2012 | May 8, 2013 | LSI Corporation | Hardware-based methods and apparatus for inter-track interference mitigation in magnetic recording systems |

Classifications

U.S. Classification | 360/39, G9B/20.046 |

International Classification | H03M13/41, G11B20/18, G11B20/10 |

Cooperative Classification | H04L25/03235, G11B20/18, H04L25/03191, G11B2020/1863, H03M13/41 |

European Classification | H04L25/03B7C9, H04L25/03B7A1, G11B20/18 |

Legal Events

Date | Code | Event | Description |
---|---|---|---|

May 25, 2004 | AS | Assignment | Owner name: AGERE SYSTEMS INC., PENNSYLVANIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:HARATSCH, ERICH FRANZ;REEL/FRAME:015398/0296 Effective date: 20040525 |

Nov 16, 2012 | AS | Assignment | Owner name: AGERE SYSTEMS LLC, PENNSYLVANIA Free format text: CERTIFICATE OF CONVERSION;ASSIGNOR:HARATSCH, ERICH FRANZ;REEL/FRAME:029310/0718 Effective date: 20040525 |

Rotate