PatentDe  


Dokumentenidentifikation EP0573621 27.01.1994
EP-Veröffentlichungsnummer 0573621
Titel NEUE EINRICHTUNG UND NEUES VERFAHREN ZUR VORKODIERUNG.
Anmelder Codex Corp., Mansfield, Mass., US
Erfinder EYUBOGLU, M., Vedat, Boston, MA 02215, US
Vertreter derzeit kein Vertreter bestellt
Vertragsstaaten DE, FR, GB
Sprache des Dokument En
EP-Anmeldetag 05.10.1992
EP-Aktenzeichen 929220754
WO-Anmeldetag 05.10.1992
PCT-Aktenzeichen US9208437
WO-Veröffentlichungsnummer 9313614
WO-Veröffentlichungsdatum 08.07.1993
EP-Offenlegungsdatum 15.12.1993
Veröffentlichungstag im Patentblatt 27.01.1994
IPC-Hauptklasse H04L 5/12

Beschreibung[en]
Field of the Invention

This invention relates generally to digital communication systems, and more particularly to precoding a digital data sequence for transmission in a digital communication system.

Background of the Invention

It has been shown that on strictly band-limited high-signal-to-noise ratio (SNR) channels with Gaussian noise, digital data can be reliably transmitted at rates approaching channel capacity by using a combination of ideal zero-forcing decision-feedback equalization (DFE) and known coded modulation and constellation shaping techniques designed for ideal channels free of intersymbol interference (ISI). However, ideal DFE is not realizable. Trellis precoding is a realizable combined coding, shaping and equalization technique that achieves the same performance as an ideal DFE along with coding and shaping.

One potential drawback of trellis precoding is that it is effective only for signal constellations whose signal points are uniformly distributed within a space-filling boundary region. Space-filling substantially means that a union of proper non-overlapping translations of the boundary region can cover (tile) the entire space. Stated in another way, the boundary region must be representable as a fundamental region of a lattice, typically referred to as a precoding lattice. To be compatible with known coded modulation techniques, a precoding lattice is typically chosen as a scaled version MZ2 of a two-dimensional integer lattice Z2 (where M is a scaling factor) such that the boundary region then has the shape of a square. In certain applications, square signal constellations are not desirable, since they have a higher two-dimensional peak-to-average power ratio (PAR) than constellations with more circular boundaries. More importantly, square constellations are not suitable for representing fractional bits per symbol and require a method known as constellation switching to allow fractional rate transmission, which further increases the two-dimensional PAR. In trellis precoding, it is possible to find precoding lattices whose Voronoi region is more circular than that of a square and which can accommodate certain fractional data rates. However, this approach is not very flexible, since it does not uniformly handle all fractional data rates and is more difficult to make invariant to 90° phase rotations, which is an important requirement in certain practical applications. Another drawback of trellis precoding is that to achieve shaping gain, the precoding operation must be combined with shaping operations, which increases the complexity of implementation.

There is a need for a flexible precoding method and device that can work with substantially any signal constellation at substantially any data rate and that can be implemented independently from constellation shaping while achieving an overall performance that is at least comparable to that of trellis precoding.

US 5,040,191 teaches the combination of trellis coding and precoding to provide the combined benefits of coding and precoding. The present invention teaches a novel method of precoding which can be combined with trellis coding as taught in the '191 patent. This invention is more closely related to Tomlinson precoding, which is a precoding method that is well-known in the art. The present invention is a new precoding technique which is different from Tomlinson precoding, and overcomes a shortcoming of Tomlinson precoding by decoupling constellation mapping and precoding.

Summary of the Invention

Accordingly the invention in a first aspect provides a device as set forth in claim 1.

According to a second aspect of the invention there is provided a digital communications receiver as set forth in claim 4.

According to a further aspect of the invention there is provided a method for precoding a digital data sequence as set forth in claim 7.

A further aspect of the invention provides a digital communication system as set forth in claim 8.

Brief Descriptions of the Drawings

FIG. 1 is a block diagram of a device in accordance with the present invention.

FIG. 2 is a more detailed block diagram illustrating a first embodiment of a device in accordance with the present invention.

FIG. 3 is a block diagram of a second embodiment of a device in accordance with the present invention.

FIG. 4 is a block diagram of a third embodiment of a device in accordance with the present invention.

FIG. 5 is a block diagram of a fourth embodiment of a device in accordance with the present invention.

FIG. 6 is a block diagram of a first embodiment of a digital communication system utilizing a device in accordance with the present invention.

FIG. 7 is a block diagram of a recovery unit of the digital communication system of FIG. 6, showing the recovery unit with more particularity.

FIG. 8 is a block diagram of a second embodiment with interleaving of a digital communication system utilizing a device in accordance with the present invention.

FIG. 9 is a flow diagram setting forth steps in accordance with the method of the present invention.

Detailed Description of a Preferred Embodiment

The method and device of the present invention permits precoding an arbitrary stream of signal points for transmission over a digital communication channel, performing particularly well on channels having severe attenuation distortion. Substantial benefits are obtained by utilizing the present invention: transmission at substantially any desired data rate without constellation switching, transmission with circular signal constellations, and simplification of shaping by completely separating shaping from precoding.

As illustrated in FIG. 1, a device in accordance with the present invention (precoding unit, 100) precodes a digital data sequence to provide a precoded sequence x(D) for transmission over a discrete time channel unit (defined below) with a complex impulse response h(D). A key characteristic of the invention is that the precoded sequence x(D) can be represented by the sum x(D) = u(D) + d(D) where u(D) is a signal point sequence from a translate of a trellis code C and represents the digital data sequence and d(D) is a dither sequence that may be represented as d(D) = c(D) - x(D)[h(D) - 1] where c(D) is a code sequence from a trellis code C. The code C can be any n-dimensional trellis code, where n is an integer, based on a lattice partition Λ/Λ', where Λ is a selected lattice, and Λ' is a selected sublattice of Λ. The individual signal points uk, k = 1, 2, ..., of u(D) are chosen from a translate of a two-dimensional lattice Λ2 and from within a finite two-dimensional region R.

Two key features distinguish the invention from prior art: (i) the sequence c(D) is not the all-zero sequence (when c(D) = 0, the technique reduces to a known transmitter linear equalization technique), and (ii) the selection of c(D) is based only upon the post-cursor ISI sequence p(D), in contrast to more conventional precoding techniques where c(D) is selected based upon both p(D) and u(D).

Typically, the present invention is utilized where the complex impulse response h(D) has no zeroes on the unit circle, or equivalently, when its inverse, 1/h(D), is stable. Therefore, the following embodiments utilize an h(D) that is a canonical response with a stable inverse. Note that h(D) may be an all-zero response such as h(D) = 1 + 0.75 D, an all-pole response such as h(D) = 1/(1 - 0.75 D), or a more general response that includes zeroes and poles. As in classical precoding techniques, it is assumed that the response h(D) has been determined and is known at the transmitter.

FIG. 2, numeral 200, is a more detailed block diagram of a first embodiment of a precoding device in accordance with the present invention. Here the digital data sequence is first mapped in a mapping unit (202) into a signal point sequence u(D) chosen from a translate of a trellis code C using any combination of known encoding, mapping and shaping techniques. This embodiment also comprises a first combining unit (204) and a filtering-slicing unit (206). The first combining unit (204), typically a summer, is operably coupled to receive the signal point sequence u(D), a code sequence c(D) and a post-cursor ISI sequence p(D) = x(D)[h(D) - 1] and provides the transmission sequence x(D) = u(D) + c(D) - p(D). The difference between x(D) and u(D) is the dither sequence d(D) = x(D) - u(D) = c(D) - p(D). A filtering unit (208) is operably coupled to receive the transmitted sequence x(D) to provide the post-cursor ISI sequence p(D).

In all embodiments that follow, the slicing unit (210) is utilized for slicing (on a symbol-by-symbol basis) the post cursor ISI sequence p(D) to an allowable code sequence c(D) selected from the trellis code C. The code sequence c(D) can be selected in many different ways. For a two-dimensional trellis code C, in one embodiment, the symbols ck of the sequence c(D) are selected from the sub lattice Λ'. This ensures that c(D) will belong to the trellis code C. In another embodiment, the symbols ck are selected on a symbol-by-symbol basis by following a single path in a trellis representation of C. The operation of the slicing unit (210) for multi-dimensional trellis codes will be described later.

Typically the symbols ck will be chosen on a symbol-by-symbol basis to minimize the instantaneous energy of the dither symbols dk = ck- pk. Since the filtering unit (208) has a delay of at least one time unit, the elements pk of the post-cursor ISI sequence p(D) depend only upon past values xi, i<k of the transmitted sequence x(D). Therefore, the current dither symbol dk will be statistically uncorrelated from the current signal point uk (assuming u(D) is itself an uncorrelated sequence). Therefore, the energy of the transmitted symbols Sx = E{|xk|2} will be the sum Su + Sd of the energies of u(D) and d(D), where E is a statistical expectation. The average energy Sx of the transmitted sequence x(D) will be approximately the same as the energy Su of the signal sequence u(D) as long as the average dither energy Sd is small. That means, the better the approximation c(D) ≈ p(D) is, the smaller will be the increase in average energy due to the dither sequence d(D).

It should be noted, however, that the invention is not limited to criteria that minimize the instantaneous dither energy, and any criterion can be used to select the code sequence c(D) as long as the selection of each ck is based only upon past values xi, i < k, of x(D). For example, in certain applications it may be desirable to limit the range of the channel output symbols yk = uk + ck. This can be achieved, at the expense of a higher dither energy Sd, by restricting the values of ck to a certain range. Also, when the present precoding method is used with a multi-level code, the selection of ck may be further restricted so that certain parity-check conditions are satisfied.

FIGs. 3-5 show three alternative embodiments of the present invention which are substantially equivalent to the first embodiment shown in FIG. 2. FIG. 3, numeral 300, sets forth a block diagram illustrating a second embodiment of a precoding device, in accordance with the present invention. In this embodiment the combining unit (304) is operably coupled to receive the signal point sequence u(D) from the mapping unit (202) and the code sequence c(D) from the slicing unit (210) to form the sequence s(D) s(D) = u(D) + c(D). An inverse filtering unit (306) which is operably coupled to the combining unit (304) to receive the sequence s(D) provides the transmitted sequence x(D) as an output of the precoding unit and as an input to the filtering unit (208) according to x(D) = s(D)/h(D). Since x(D)h(D) = x(D) + p(D) = s(D), it follows that x(D) = s(D)/h(D) = s(D) - p(D) = u(D) + c(D) - p(D).

Thus for the same input sequence u(D), this second embodiment will produce essentially the same transmitted sequence x(D) as that of the first embodiment.

A third embodiment of the precoding device is shown in FIG. 4, numeral 400. Here the filtering unit (408) is operably coupled to the first combining unit (304) such that the post-cursor ISI sequence p(D) is generated directly from the sequence s(D) instead of the transmitted sequence x(D), and the filtering unit (408) has the response {1 -1/h(D)}. It should be noted that since x(D) = s(D)/h(D), and p(D) = s(D){1 - 1/h(D)} = x(D){h(D) - 1} the filtering unit (408) in this embodiment will produce the same post-ISI sequence p(D) as the earlier embodiments, and therefore the same transmitted sequence x(D) will be generated as in the earlier embodiments, again assuming the same input sequence u(D). Other operable couplings for the third embodiment are as described for FIG. 3.

A fourth embodiment shown in FIG. 5, numeral 500, is similar to that in FIG. 4, except that here the inverse filtering operation x(D) = s(D)/h(D) is implemented according to x(D) = s(D) - p(D) where a second combining unit (512), operably coupled to the first combining unit (304) and to the filtering unit (408), substantially subtracts the post-cursor ISI sequence p(D) provided by the filtering unit (408) from the sequence s(D) provided by the first combining unit (304).

The above description utilizes an assumption that the channel is characterized by a discrete-time complex impulse response h(D). It is well-known in the state-of-the-art that any discrete time or continuous-time linear passband channel with additive noise can be represented by a canonical discrete-time equivalent channel with a causal (hk = 0, k <0), minimum-phase (all zeros outside or on the unit circle), monic (ho = 1) impulse response h(D) and additive white noise w(D). A canonical receiver front-end that includes a whitened matched filter and a sampler (in the case of continuous-time channels) operating at a selected symbol rate may be utilized to provide such an equivalent channel. It should be mentioned that in practice, typically, h(D) represents the combined effect of the filters in the transmitter, channel, the receiver, and a sampler. Similarly, w(D) represents the noise after it passes through the receive filters and the sampler. The whitened-matched filter reduces the strength of the distortion through proper filtering and therein lies the performance advantage of the present invention over conventional linear equalizations.

In practice, when h(D) is an all-zero response, a whitened matched filter can be determined adaptively using standard adaption techniques for decision-feedback equalizers. When it is desired that h(D) be an all-pole filter, then one can first determine adaptively an all-zero response h'(D) using the standard methods and then find h(D) = 1/g'(D) using well-known polynomial division techniques, where g'(D) is a finite polynomial approximately equal to g'(D) ≈ 1/h'(D).

A first embodiment of a device of the present invention incorporated into a digital communication system is illustrated in the block diagrams of FIG. 6, numeral 600, wherein at least one of a transmission unit and a receiving unit utilizes the present invention. The said system typically includes at least one of a transmission unit and a receiving unit wherein the transmission unit has a precoding unit (602) (typically in a transmitter) for transmitting a digital data sequence and a channel (604) obtained as described in the above paragraph, operably coupled to the precoding unit (602), for facilitating transmission of the precoded sequence x(D), and the receiving unit has a decoding unit (612), operably coupled to the channel unit (604), for receiving and decoding a received sequence r(D) to provide an estimated output sequence y&supand;(D), and a recovery unit (614), operably coupled to the decoding unit (612), for substantially recovering an estimate û(D) of the signal point sequence u(D). An estimate of the transmitted digital data sequence is then found from û(D) using an inverse map and shaping recovery (if constellation shaping is employed).

The equivalent channel (604), represented as set forth above, is substantially represented by a filter having a response h(D) (606), for receiving x(D) and producing an output sequence y(D) = x(D)h(D) substantially equal to s(D), defined earlier, an additive noise unit (608) for providing additive noise, and a combining unit (610), typically a summer, operably coupled to the h(D) unit (606) and to the additive noise unit (608).

The decoding unit (612) is typically a decoder for the trellis code C, as is known in the art. The decoding unit (612), typically receives and decodes a noisy received sequence r(D) which is of a form r(D) = x(D)h(D) + w(D) = y(D) + w(D) = [u(D) + c(D)] + w(D), to provide an estimate y&supand;(D) of the channel output sequence y(D) = x(D)h(D), and a recovery unit (614), operably coupled to the decoding unit (612), substantially recovers an estimate û(D) of the input sequence u(D), described more fully below.

Since c(D) belongs to the trellis code C, the sequence y(D) = u(D) + c(D) must be a sequence in the same translate of C as the signal point sequence u(D). Of course, since ck can be large, the symbols yk = uk + ck of y(D) will lie within a boundary region that is larger than the region R defined earlier for the symbols uk of u(D). That means that the sequence y(D) can be estimated by a conventional decoder (612) for C, as is known in the art, with the provision that it searches over a larger (possibly infinite) range of signal points.

Note that a conventional receiver will not take advantage of the correlation between successive channel output signals yk. To see this, note that it is possible to construct augmented trellis diagrams using the fact that yk = uk + ck, and ck depends on past values of yk. If for example, h(D) is an all-pole response of order m, then ck will depend on a finite number of recent symbols yk-i, i = 1 ,...,m (where m is a selected integer). Therefore, a sequence estimator can be defined by combining the state of the input sequence u(D) and the channel state sk = [yk-1, yk-2,...,yk-m]. Since such a trellis can have an excessively large number of states, and therefore will be difficult to search, reduced-complexity search techniques can be employed to achieve nearly the same performance at substantially reduced complexity. The number of states in an ML trellis can be reduced using state merging techniques for reduced-state sequence estimation (RSSE).

A simplest form RSSE involves a search over the trellis of the trellis code C, and operates like a conventional decoder except it uses the path history of each surviving sequence to find an estimate c&supand;k of the code variables ck. It is important to note that a different estimate c&supand;k is used for each state of the trellis, and the estimate associated with each state is used to compute the branch metrics for all branches that leave that state. Given an estimate c&supand;k allows determination of an estimated range R + c&supand;k for the signal points yk = uk + c&supand;k and this limited range is used in the all branch metric computations for that estimate. The disadvantage of this technique is that it increases the computational complexity since it requires considerably more branch metric calculations. The complexity can reduced, however, by using only a more likely subset of the possible states to form the estimates c&supand;k.

The recovery unit (614), illustrated with more particularity in the block diagram of FIG. 7, numeral 700, typically includes at least a recovery filtering unit (702), operably coupled to the estimated output sequence y&supand;(D), for filtering y&supand;(D) to obtain an estimate p&supand;(D) of the precursor ISI sequence p(D), substantially of a form p&supand;(D) = y&supand;(D){1-1/h(D)}, a recovery slicing unit (704), operably coupled to the recovery filtering unit (702), for slicing p&supand;(D) to provide a recovery code sequence c&supand;(D) from the trellis code C, in a manner that is substantially the same as that used in the precoding unit at the transmitter, and a recovery combining unit (706), operably coupled to the recovery slicing unit (704) and to the decoding unit (612), for substantially determining a difference between the output sequence y&supand;(D) and the sequence c&supand;(D) to obtain the estimate û(D) of the original input sequence u(D). As long as there are no decision errors (y&supand;(D) = y(D)), and the operations in the transmitter and receiver are substantially symmetrical, the original sequence u(D) will be correctly recovered. Other equivalent implementations of the recovery circuit are also possible.

To summarize, the recovery filtering unit (702) is utilized to reconstruct an estimate p&supand;k of a post-cursor intersymbol interference variable pk, then the recovery slicing unit (704) is utilized to determine a symbol c&supand;k belonging to a sublattice Λ' (or an allowable coset of Λ') on a symbol-to-symbol basis to form a sequence c&supand;(D) that substantially correlates with c(D) in the precoding unit (100) at the transmitter, and then utilize the recovery combining unit (706) to provide û(D) = y&supand;(D) - c&supand;(D).

Of course, there will be occasional errors in y&supand;(D) due to channel noise, and these may lead to error propagation. However, since 1/h(D) is stable, the error propagation in the filter 1 - 1/h(D) will never be catastrophic. Moreover, if h(D) is an all-pole response of order m (where m is a selected integer), then error propagation will be strictly limited to at most m symbols.

In an embodiment where the elements ck of the sequence c(D) are chosen on a symbol-by-symbol basis at the transmitter by following a path through the trellis of the code C, during recovery, occasional errors may cause loss of the correct state thereby causing loss of synchronization, which may continue for a long time if not corrected. One way to circumvent this problem is to force c(D) in the transmitter to a known state of the trellis at every Lth symbol, where L is a preselected integer, to reestablish synchronization in the receiver at an expense of a slight increase in the transmit energy.

Where an estimate ûk, a kth variable of the recovered sequence u(D), falls outside the allowed range R of a kth variable uk of the input sequence u(D), such a range violation indicates that a decision error has occurred either in the current symbol yk, or ck is in error because of an error in some recent symbol yk-i, i > 0. When such range violations are detected, one can try to correct them by adjusting the estimates yk or yk-i. Thus, by monitoring the range violations, some degree of error correction can be achieved. Such an error detection capability can also be useful for monitoring the performance of the transmission system.

A second embodiment of a digital communication system having a device in accordance with the present invention, the system including interleaving, is illustrated in FIG. 8, numeral 800. In certain applications, an interleaver INT (802) may be included in a precoding unit (810). The precoding unit (810) is the same as the precoding unit (100), except that the precoding unit (810) further includes an interleaver such that the signal point sequence u(D) is now replaced by an interleaved signal point sequence u'(D) which is obtained by passing u(D) through the interleaver (802). The interleaver will allow the removal of potential correlation between successive noise samples at the output of the channel. The interleaver (802) can be any device with a realizable inverse (deinterleaver) that changes the ordering of the input symbols. For example, in a periodic interleaver, input samples are delayed according to a sequence of delays that are periodic with same period P. As described above for FIG. 6, x(D) and r(D) are generated.

In the receiver, the received noisy sequence r(D) = u'(D) + c(D) + w(D) is deinterleaved prior to decoding by passing r(D) through a deinterleaver DEINT (804), operably coupled to receive r(D), to obtain the deinterleaved received sequence r'(D) = u(D) + c'(D) + w'(D), where w'(D) is the noise sequence whose order has been shuffled in the deinterleaver (804). It is this shuffling of the noise that produces improved performance when the channel noise is bursty or correlated. Note that the deinterleaver (804) recovers the original ordering for the signal point sequence u(D), but in the process it changes the ordering of the code symbols ck added in the precoding device at the transmitter. It is essential that the shuffled code sequence c'(D) is a code sequence in the trellis code C, so that the decoding unit (612), typically a conventional Viterbi decoder for C, operably coupled to the deinterleaver (804), can recover the original sequence y(D) = u(D) + c'(D). For a two-dimensional trellis code C this is always ensured if the elements ck are chosen from the sublattice Λ' on a symbol-by-symbol basis and with no memory.

To recover the original input sequence u(D), the estimated sequence y&supand;(D) from the decoder for C must be interleaved once again by passing y&supand;(D) through a second interleaver INT (806), which is operably coupled to the decoding unit (612) and which is substantially equivalent to the interleaver (802) in the transmitter, to recover the original ordering of the code symbols c(D). Interleaving provides y&supand;'(D) = û'(D) + c&supand;(D). The recovery unit (614), as further illustrated in FIG. 7, operably coupled to the second interleaver (806), receives y&supand;'(D) and provides an estimate û'(D) for the interleaved sequence u'(D). An estimate û(D) of the original signal point sequence u(D) is then obtained by passing û'(D) through a second deinterleaver DEINT (808) that is included in the recovery unit (614). Recovery unit (812) shown in FIG. 8 is the same as recovery unit (614) except that it includes the deinterleaver (808) that is operably coupled to the recovery combining unit (706) and provides a deinterleaved estimated signal point sequence û(D) for the inverse mapping device (708).

Thus, a digital communications receiver may be utilized in accordance with the present invention for receiving a digital data sequence that was mapped into a signal point sequence x(D) and transmitted over a channel characterized by a non ideal response h(D) using a trellis code C, providing a received sequence r(D), comprising at least decoding means, operably coupled to receive r(D), for decoding the received transmission sequence r(D) to provide an estimated output sequence y&supand;(D), and recovery means, operably coupled to the decoding means, for substantially recovering an estimated sequence û(D) for a sequence u(D) for a transmitted signal point sequence x(D), selected from a subset of all possible signal point sequences that are of a form u(D) + d(D), wherein u(D) is a signal point sequence from a translate of said trellis code C and uniquely represents said digital data sequence and wherein d(D) represents a nonzero difference between a selected nonzero code sequence c(D) from said trellis code C and a post-cursor intersymbol interference (ISI) sequence p(D) substantially of a form p(D) = x(D)[h(D)-1], such that c(D) is selected based only upon p(D).

As illustrated in FIG. 7, one embodiment of the recovery means utilizes a recovery filtering unit (702), operably coupled to receive the estimated output sequence y&supand;(D), for providing an estimated post-cursor intersymbol interference (ISI) sequence p&supand;(D), a recovery slicing unit (704), operably coupled to the recovery filtering means, for providing an estimated nonzero code sequence c&supand;(D) from said trellis code C that substantially correlates with c(D) utilized for providing the transmission sequence x(D), a third combining unit (706) (typically a summer), operably coupled to receive the estimated output sequence y&supand;(D) and to the recovery slicing means, for determining the estimated sequence û(D), substantially of a form û(D) = y&supand;(D) - c&supand;(D), and an inverse mapping unit (708), operably coupled to the third combining means, for inverse mapping the estimated sequence û(D) to provide a recovered digital data sequence substantially equal to the transmitted digital data sequence. The receiver further includes, where the signal point sequence u(D) was interleaved before transmission, a first deinterleaving unit DEINT (804), operably coupled to receive r(D), for providing a deinterleaved r(D) sequence to the decoding unit (612), an interleaving unit INT (806), operably coupled to the decoding unit (612), for providing an interleaved estimate of the decoded sequence, y&supand;'(D), to the recovery unit (812), and wherein the recovery unit (812) further includes a second deinterleaving unit (808). The digital commmunications receiver is utilized as further described above.

In addition the digital communications receiver may be selected such that the decoding unit (612) further includes a reduced complexity sequence estimator unit that utilizes a correlation between successive symbols yk. In one implementation, the reduced complexity sequence estimator unit utilizes a sequence estimator having a reduced number of states that are determined utilizing state merging techniques for reduced-state sequence estimation (RSSE).

Where desired, the recovery unit may be selected to include a range violation determiner unit. When a kth variable ûk of the recovered sequence û(D) is outside a range R (a range violation), this unit adjusts at least one of an estimate y&supand;k and a past estimate y&supand;k-i (where i is a positive integer) to substantially correct the range violation.

The present invention may also be utilized with multi-dimensional codes. For example, if u(D) is an arbitrary complex sequence from a translate of a four-dimensional trellis code C based on a four-dimensional (4D) lattice partition Z4/RD4, the sequence c(D) can be selected as follows. First it should be noted that the sublattice RD4 can be represented as a union of the 4D lattice 2Z4 with its coset 2Z4 + (1,1,1,1). Moreover, the 4D lattice 2Z4 can be obtained by taking a Cartesian product of the two-dimensional (2D) lattice 2Z2 which consists of all pairs of even integers. Therefore RD4 can be represented as RD4 = (2Z2x2Z2) U [2Z2 + (1,1)]x[2Z2 + (1,1)], where U represents the union and x represents the Cartesian product. The union of the 2D lattice 2Z2 with its coset 2Z2 + (1,1) forms the 2D lattice RZ2.

Therefore, the slicing unit (202) for precoding may select the code sequence c(D) by selecting, in the even symbol interval k, its symbol ck from RZ2. If ck belongs to 2Z2 then in the following odd symbol interval, the second symbol ck+1 is selected from the even integer lattice 2Z2. If ck belongs to the coset 2Z2+ (1,1), however, then in the next odd symbol interval, the second symbol ck+1 is selected from the coset 2Z2 + (1,1). This way it is ensured that the 4D symbol (ck, ck+1) will belong to RD4.

Alternatively, the sequence c(D) may be selected by following the best path through the trellis of the multi-dimensional code. For example, in the case of a popular 4D 16-state trellis code disclosed by Wei (Trellis-coded Modulation With Multi-dimensional Constellations, by L.F. Wei, IEEE Transactions on Information Theory, Vol. IT-33, pp. 483-501, July 1987), in the even symbol interval, the closest symbol ck on the integer lattice Z2 can be selected. In the following odd symbol interval, the closest symbol in either RZ2 or its coset RZ2 + (1,0) is selected, depending on the current state of the path and depending on whether ck belongs to RZ2 or its coset RZ2 + (1,0). This way, the dither variable dk is forced to lie inside the Voronoi region of Z2 for the first symbol, and the Voronoi region of RZ2 for the second symbol.

In the alternative c(D) selection described above, a single error can propagate indefinitely in the reconstruction causing errors in the selection of the second symbol ck+1, since state information may be lost, and it may not be possible to correctly determine whether ck+1 will lie in RZ2 or its coset RZ2 + (1,0). As stated above, this problem can be avoided by forcing the trellis to the all-zero state every Lth (L even) symbol interval. During the symbol intervals, where the forcing takes place, the symbols will be selected from specific cosets of 2Z2 and therefore the dither variable dk will lie in the Voronoi region of 2Z2 in the second symbol, thereby slightly increasing the dither energy.

FIG. 9, numeral 900, sets forth a flow diagram illustrating steps in accordance with the method of the present invention for precoding a stream of signal points for transmission in a digital communication system. The method provides for precoding a digital data sequence to generate a sequence x(D) for transmission over a discrete-time channel with a impulse response h(D). A stream of signal points u(D) chosen from some translate of a trellis code C is transmitted as x(D) = u(D) + d(D), where d(D) is a dither sequence of a form d(D) = c(D) - p(D), where p(D) represents a post-cursor intersymbol interference (ISI), and c(D) is a code sequence, where c(D) is different from an all-zero sequence and is obtained from an untranslated version of the trellis code C based only upon p(D). In one embodiment, the method comprises the steps of mapping the digital data sequence to a signal point sequence u(D) (902), summing u(D), a selected code sequence c(D) and a post-cursor ISI sequence p(D) to obtain the transmission sequence x(D) = u(D) + c(D) - p(D) (904), filtering x(D) to obtain (906) p(D) substantially of a form: p(D) = x(D)[h(D) - 1], and slicing p(D) on a symbol-by-symbol basis to obtain the code sequence c(D) (908). Further modifications of the method may be utilized in accordance with the modifications described more fully above for the device of the present invention.

The present invention may be implemented in a digital communication system where a digital signal processor is utilized to precode a digital data sequence to obtain a sequence x(D) for transmission over a discrete-time channel with an impluse response h(D). The processor typically comprises a program storage medium having a computer program to be executed by the digital signal processor, the program comprising a unit for generating a sequence x(D) wherein x(D) can be represented as the sum u(D) + d(D) of a stream of signal points u(D) chosen from a translate of a trellis code C and a dither sequence d(D) = c(D) - p(D), where c(D) is a sequence from a translate of a trellis code C and where p(D) represents a post-cursor intersymbol interference (ISI) sequence of a form p(D) = x(D){h(D) - 1}. The code sequence c(D) is different from an all-zero sequence and is determined based upon only the post-cursor ISI sequence p(D). Further description of the operation of the processor follows that described above.

The present invention relies on past channel output signals to remove a dither sequence d(D) that is added to a input sequence u(D) at the transmitter to form a transmitted sequence, x(D) = u(D) + d(D), the dither sequence being substantially a difference between a post-cursor intersymbol interference p(D) and an appropriate code sequence, c(D), from a trellis code from which the input sequence C is chosen. The present invention may be utilized with virtually any signaling method and at any data rate. Further, the present invention may be utilized independently of constellation shaping techniques; that means u(D) may represent an already shaped sequence whose signal points have a nonuniform Gaussian-like probability distribution.

In the present invention, the dither sequence may increase the average transmit energy. Since in practice, the average transmit energy must be kept constant, the signal x(D) must be scaled down to maintain the same average energy. The increase in the average transmit energy is referred to herein as a dithering loss.

The dithering loss depends on the average energy Su of the signal point sequence u(D), on whether u(D) is coded or uncoded, and also on the method used in selecting the code sequence c(D). For example, if u(D) is a sequence of symbols selected from an uncoded M x M quadrature amplitude-modulated (QAM) signal constellation (where M is a selected integer, typically a power of two) with an average energy Su = (M2 - 1)/6 , the transmission rate will be r = log2 M2 bits per symbol, and the dither sequence d(D) will have dither variables dk that are uniformly distributed inside a square of side length one, so that Sd ≈ 1/6 and Sx ≈ Su + Sd ≈ M2/6, where Sd is a dither energy. Thus, the dithering loss is approximately 0.28 dB for M = 4 (r = 4) and about 0.02 dB for M = 6 (r = 8), and rapidly approaches zero as M grows. Here, dithering loss is substantially the same as a similar loss known to occur in Tomlinson precoding.

In a second exemplary embodiment, for a sequence of data symbols from an M x M QAM signal constellation, where a sequence u(D) is a sequence of code symbols from a translate C + (0.5)2 of a two-dimensional trellis code C that is based on the lattice partition Z2/2Z2 and a rate-1/2 convolutional code, the transmission rate is r = log2 M2 - 1 bits per symbol, and the dither variables dk are distributed inside a Voronoi region of Λ' = 2Z2, which is a square of side 2, the dither energy is now four times higher; i.e., Sd ≈ 2/3 and Sx ≈ (M2 + 3)/6. Thus, the dithering loss is about 1.02 dB for M = 4 (r = 3), and about 0.07 dB for M = 16 (r = 7).

Clearly, for the same signal constellation (same M), the dithering loss is higher in the trellis-coded QAM system (second exemplary embodiment) than in the uncoded system (first exemplary embodiment). However, as M gets large, the dithering loss becomes very small in both cases. Also, it should be noted that when compared at a same data rate, the difference in the dithering loss is smaller.

In a third exemplary embodiment, for a sequence u(D) of data symbols from an M x M QAM signal constellation that is generated by a trellis shaping system with about 1 dB shaping gain and a shaping constellation expansion factor of 2, wherein a transmission rate for the otherwise uncoded system is r = log2 M2 - 1 bits per symbol. The dithering energy is substantially Sd ≈ 1/6 and Su ≈ (M2 - 1)/15. Thus, Sx ≈ M2/15 + 0.1, and the dithering loss is substantially 0.67 dB for M = 4 (r3 = 3) and 0.04 for M = 16 (r3 = 7). Hence, shaping increases the dithering loss only very slightly.

A fourth exemplary embodiment illustrates implementation of the present invention to provide reduction of dithering loss while incurring slightly higher computational complexity. In this case, the input sequence is the same as in the second exemplary embodiment, but the method of slicing for determining c(D) is different. Here c(D) is selected on a symbol-by-symbol basis as described above from appropriate cosets of the so-called time-zero lattice Λ0 = RZ2 by following the trellis associated with the trellis code. In this case, the dithering variable is uniformly distributed inside the Voronoi region of RZ2, and hence Sd ≈ 1/3, and therefore the dithering loss will be only 0.54 dB for M = 4 and 0.03 dB for M = 16.

Although several exemplary embodiments are described above, it will be obvious to those skilled in the art that many alterations and modifications may be made without departing from the invention. For example, even though primarily trellis codes are described above, the method can be used with block or lattice codes as well. The trellis code can be trivial as in an uncoded system. It can also be used with selected multi-level trellis codes. Also, although two-dimensional (passband, quadrature) transmission systems are emphasized, the methods can also be applied to one-dimensional (baseband) or higher-dimensional (parallel channels) transmission systems. Accordingly, it is intended that all such alterations and modifications be included within the scope of the invention as defined in the appended claims.


Anspruch[de]
  1. Einrichtung zur Abbildung einer digitalen Datenfolge in eine Signalpunktfolge x(D) für die Übertragung über einen Kanal, der eine nicht ideale Antwort h(D) hat, durch Verwendung eines Trelliskodes C, wobei die Einrichtung enthält:
    • eine Vorkodierungseinheit (100) zur Auswahl dieser Signalpunktfolge x(D) aus einer Untermenge von allen möglichen Signalpunktfolgen, die eine Form u(D)+d(D) haben, wobei u(D) eine Signalpunktfolge aus einer Umsetzung dieses Trelliskodes C ist und diese digitale Datenfolge eindeutig darstellt, und wobei d(D) eine Nicht-Null Differenz darstellt zwischen einer ausgewählten Nicht-Null Kodefolge c(D) von diesem Trelliskode C und einer nachträglichen Zwischenzeicheninterferenz-(ISI) Folge p(D), die im wesentlichen eine Form p(D)=x(D)[h(D)-1] hat, so daß c(D) nur auf p(D) basierend ausgewählt ist.
  2. Einrichtung nach Anspruch 1, wobei die Vorkodierungseinheit (100) enthält:
    • Abbildungsmittel (202) zur Abbildung der digitalen Datenfolge in eine Signalpunktfolge u(D), die aus einer Umsetzung eines Trelliskodes C ausgewählt wird;
    • Verknüpfungsmittel (204, 304), die auf die Signalpunktfolge u(D) reagieren zur Verknüpfung der Signalpunktfolge u(D) und einer Kodefolge c(D), um eine Verknüpferausgangsfolge s(D) bereitzustellen;
    • Mittel (204, 306, 512) zur Erzeugung der Signalpunktfolge x(D), die im wesentlichen die Form s(D)-p(D hat, im wesentlichen durch die Subtraktion einer nachträglichen ISI-Folge p(D) von der Verknüpferausgangsfolge s(D);
    • Filtermittel (208, 408), die schaltbar gekoppelt sind, um zu empfangen:
      • die Verknüpferausgangsfolge s(D); oder
      • die Signalpunktfolge u(D),
      wobei diese Filtermittel die nachträgliche ISI-Folge p(D) bereitstellen; und
    • Zerteilungsmittel (210), die schaltbar gekoppelt sind, um die nachträgliche ISI-Folge p(D) zu empfangen zur Zerteilung der nachträglichen Folge p(D) in die ausgewählte Kodefolge der Signalpunkte c(D), die aus dem Trelliskode C ausgewählt ist.
  3. Einrichtung nach Anspruch 2, wobei die Abbildungsmittel Mittel zur Konstellationsformung enthalten.
  4. Digitaler Nachrichtenempfänger, der Entschlüsselungs- und Wiederherstellungsmittel hat zum Empfang einer Empfangsfolge r(D) einer digitalen Datenfolge, die in eine Signalpunktfolge x(D) abgebildet wurde, und durch Verwendung eines Trelliskodes C über einen Kanal übertragen wurde, der eine nicht ideale Antwort h(D) besitzt, dadurch gekennzeichnet, daß:
    • die Entschlüsselungsmittel (612) auf die Empfangsfolge r(D) zur Bereitstellung einer abgeschätzten Ausgangsfolge dy(D) reagieren; und
    • die Wiederherstellungsmittel (614) auf die abgeschätzte Ausgangsfolge dy(D) im wesentlichen zur Wiederherstellung einer abgeschätzten Folge du(D) einer Folge u(D) für eine übertragene Signalpunktfolge x(D) reagieren, die aus einer Untermenge von allen möglichen Signalpunktfolgen ausgewählt wird, die eine Form von u(D)+d(D) haben, wobei u(D) eine Signalpunktfolge aus einer Umsetzung dieses Trelliskodes C ist und diese digitale Datenfolge eindeutig darstellt, und wobei d(D) eine Nicht-Null Differenz ist zwischen einer ausgewählten Nicht-Null Kode folge c(D) aus diesem Trelliskode C und einer nachträglichen Zwischenzeicheninterferenz-(ISI) Folge p(D), die im wesentlichen eine Form p(D)=x(D)[h(D)-1] hat, so daß c(D) nur auf p(D) basierend ausgewählt ist.
  5. Digitaler Nachrichtenempfänger nach Anspruch 4, wobei die Wiederherstellungsmittel enthalten:
    • Wiederherstellungsfiltermittel (702), die auf die abgeschätzte Ausgangsfolge dy(D) reagieren zur Bereitstellung einer abgeschätzten nachträglichen Zwischenzeicheninterferenz-(ISI) Folge dp(D);
    • Wiederherstellungszerteilungsmittel (704), die auf die abgeschätzte nachträgliche Zwischenzeicheninterferenz-(ISI) Folge dp(D) reagieren zur Bereitstellung einer abgeschätzten Nicht-Null Kodefolge dc(D) aus diesem Trelliskode C, die im wesentlichen mit der Folge c(D) korreliert, die zur Bereitstellung der Übertragungsfolge x(D) verwendet wird;
    • Wiederherstellungsverknüpfungsmittel (706), die auf die abgeschätzte Ausgangsfolge dy(D) und auf die abgeschätzte Nicht-Null Kodefolge dc(D) reagieren zur Bestimmung der abgeschätzten Folge du (D), die im wesentlichen eine Form von du(D)=dy(D)-dc(D) hat; und
    • Umkehrabbildungsmittel (708), die auf die abgeschätzte Folge du(D) reagieren zur Umkehrabbildung der abgeschätzten Folge du(D), um eine wiederhergestellte digitale Datenfolge bereitzustellen.
  6. Digitaler Nachrichtenempfänger nach Anspruch 5, wobei die Umkehrabbildungsmittel Mittel zur Wiederherstellung der Konstellationsformung enthalten.
  7. Verfahren zur Vorkodierung einer digitalen Datenfolge, dadurch gekennzeichnet, daß das Verfahren eine Signalpunktfolge u(D) verwendet, die aus einer Umsetzung eines Trelliskodes C ausgewählt wird, zur Übertragung über einen zeitdiskreten Kanal mit einer Impulsantwort h(D), wobei p(D) eine nachträgliche Zwischenzeicheninterferenz-(ISI) Folge darstellt und c(D) eine Nicht-Null Kodefolge von Signalpunkten c(D) ist, wobei c(D) aus einer nicht umgesetzten Version des Trelliskodes C gewonnen wird, wobei das Verfahren die folgenden Schritte umfaßt:
    • Abbildung (902) der digitalen Datenfolge in eine Signalpunktfolge u(D);
    • Summierung (904) der Folgen u(D), c(D) und p(D), um eine Übertragungsfolge x(D) bereitzustellen, die im wesentlichen eine Form x(D)=u(D)+c(D)-p(D) hat;
    • Filterung (906) der Folge x(D), um eine nachträgliche Zwischenzeicheninterferenz-(ISI) Folge p(D) zu erhalten, die im wesentlichen die Form p(D)=x(D)[h(D)-1] hat; und
    • zeichenweise Zerteilung (908) der nachträglichen ISI-Folge p(D), um die Kodefolge c(D) zu erhalten.
  8. Digitales Datenübertragungssystem (600), das eine Übertragungseinheit hat zur Abbildung einer digitalen Datenfolge in eine Signalpunktfolge x(D) zur Übertragung über einen Kanal, der eine nicht ideale Antwort h(D) hat, indem ein Trelliskode C verwendet wird, und eine Empfängereinheit zum Empfang einer Empfangsfolge r(D) von diesem Kanal, dadurch gekennzeichnet, daß:
    • die Übertragungseinheit eine Vorkodierungseinheit zur Auswahl dieser Signalpunktfolge x(D) aus einer Untermenge von allen möglichen Signalpunktfolgen hat, die eine Form von u(D)+d(D) haben, wobei u(D) eine Signalpunktfolge aus einer Umsetzung dieses Trelliskodes C ist und diese digitale Datenfolge eindeutig darstellt, und wobei d(D) eine Nicht-Null Differenz ist zwischen einer ausgewählten Nicht-Null Kodefolge c(D) aus diesem Trelliskode C und einer nachträglichen Zwischenzeicheninterferenz-(ISI) Folge p(D), die im wesentlichen eine Form p(D)=x(D)[h(D)-1] hat, so daß c(D) nur auf p(D) basierend ausgewählt ist; und
    • die Empfangseinheit enthält:
      • Entschlüsselungsmittel, die auf die Empfangsfolge r(D) reagieren zur Bereitstellung einer abgeschätzten Ausgangsfolge y(D); und
      • Wiederherstellungsmittel, die auf die abgeschätzte Ausgangsfolge y(D) reagieren, im wesentlichen zur Wiederherstellung einer abgeschätzten Folge du(D) einer Folge u(D) für eine übertragene Signalpunktfolge x(D), die aus einer Untermenge von allen möglichen Signalpunktfolgen ausgewählt wird, die eine Form von u(D)+d(D) haben, wobei u(D) eine Signalpunktfolge aus einer Umsetzung dieses Trelliskodes C ist und diese digitale Datenfolge eindeutig darstellt, und wobei d(D) eine Nicht-Null Differenz ist zwischen einer ausgewählten Nicht-Null Kode folge c(D) aus diesem Trelliskode C und einer nachträglichen Zwischenzeicheninterferenz-(ISI) Folge p(D), die im wesentlichen eine Form p(D)=x(D)[h(D)-1] hat, so daß c(D) nur auf p(D) basierend ausgewählt ist.
Anspruch[en]
  1. A device for mapping a digital data sequence into a signal point sequence x(D) for transmission over a channel having a nonideal response h(D) using a trellis code C, the device comprising:

       a precoding unit (100) for selecting said signal point sequence x(D) from a subset of all possible signal point sequences that are are of a form u(D) + d(D), wherein u(D) is a signal point sequence from a translate of said trellis code C and uniquely represents said digital data sequence and wherein d(D) represents a nonzero difference between a selected nonzero code sequence c(D) from said trellis code C and a post-cursor intersymbol interference (ISI) sequence p(D) substantially of a form p(D) = x(D)[h(D)-1], such that c(D) is selected based only upon p(D).
  2. The device of claim 1 wherein the precoding unit (100) comprises:
    • mapping means (202) for mapping the digital data sequence into a signal point sequence u(D) selected from a translate of a trellis code C;
    • combining means (204, 304), responsive to the signal point sequence u(D), for combining the signal point sequence u(D) and a code sequence c(D) to provide a combiner output sequence s(D);
    • means (204, 306, 512) for generating the signal point sequence x(D) substantially of the form s(D) - p(D) by substantially subtracting a post-cursor ISI sequence p(D) from the combiner output sequence s(D);
    • filtering means (208, 408), operably coupled to receive one of:
      • the combiner output sequence s(D); and
      • the signal point sequence x(D), said filtering means providing the post-cursor ISI sequence p(D); and
    • slicing means (210), operably coupled to receive the post-cursor ISI sequence p(D), for slicing the post-cursor sequence p(D) into the selected code sequence of signal points c(D) selected from the trellis code C.
  3. The device of claim 2 wherein the mapping means includes means for constellation shaping.
  4. A digital communications receiver having decoding and recovery means for receiving a received sequence r(D) of a digital data sequence that was mapped into a signal point sequence x(D) and transmitted over a channel having a nonideal response h(D) using a trellis code C, characterized in that:
    • the decoding means (612) is responsive to the received sequence r(D) for providing an estimated output sequence dy(D); and
    • the recovery means (614) is responsive to the estimated output sequence y&supand;(D) for substantially recovering an estimated sequence û(D) for a sequence u(D) for a transmitted signal point sequence x(D), selected from a subset of all possible signal point sequences that are of a form u(D) + d(D), wherein u(D) is a signal point sequence from a translate of said trellis code C and uniquely represents said digital data sequence and wherein d(D) represents a nonzero difference between a selected nonzero code sequence c(D) from said trellis code C and a post-cursor intersymbol interference (ISI) sequence p(D) substantially of a form p(D) = x(D)[h(D)-1], such that c(D) is selected based only upon p(D).
  5. The digital communications receiver of claim 4 wherein the recovery means comprises:
    • recovery filtering means (702), responsive to the estimated output sequence y&supand;(D), for providing an estimated post-cursor intersymbol interference (ISI) sequence p&supand;(D);
    • recovery slicing means (704), responsive to the estimated post-cursor intersymbol interference (ISI) sequence p&supand;(D), for providing an estimated nonzero code sequence c&supand;(D) from said trellis code C that substantially correlates with c(D) utilized for providing the transmission sequence x(D);
    • recovery combining means (706), responsive to the estimated output sequence dy(D) and to the estimated nonzero code sequence c&supand;(D), for determining the estimated sequence û(D), substantially of a form û(D) = y&supand;(D) - c&supand;(D); and
    • inverse mapping means (708), responsive to the estimated sequence û(D), for inverse mapping the estimated sequence û(D) to provide a recovered digital data sequence.
  6. The digital communications receiver of claim 5 wherein said inverse mapping means includes means for constellation shaping recovery.
  7. A method for precoding a digital data sequence characterized in that the method utilizes a signal point sequence u(D) chosen from a translate of a trellis code C for transmission over a discrete-time channel with an impulse response h(D) where p(D) represents a post-cursor intersymbol interference (ISI), and c(D) is a nonzero code sequence of signal points c(D), where c(D) is obtained from an untranslated version of the trellis code C, the method comprising the steps of:
    • mapping (902) the digital data sequence to a signal point sequence u(D);
    • summing (904) the sequences u(D), c(D) and p(D) to provide a transmission sequence x(D) substantially of a form x(D) = u(D) + c(D) - p(D);
    • filtering (906) the sequence x(D) to obtain a post-cursor intersymbol interference (ISI) sequence p(D) substantially of a form p(D) = x(D)[h(D) - 1]; and
    • slicing (908) the post-cursor ISI sequence p(D) on a symbol-by-symbol basis to obtain the code sequence c(D).
  8. A digital communication system (600) having a transmission unit for mapping a digital data sequence into a signal point sequence x(D) for transmission over a channel having a nonideal response h(D) using a trellis code C and a raceiving unit for receiving a received sequence r(D) from said channel, characterized in that:
    • the transmission unit has a precoding unit for selecting said signal point sequence x(D) from a subset of all possible signal point sequences that are of a form u(D) + d(D), wherein u(D) is a signal point sequence from a translate of said trellis code C and uniquely represents said digital data sequence and wherein d(D) represents a nonzero difference between a selected nonzero code sequence c(D) from said trellis code C and a post-cursor intersymbol interference (ISI) sequence p(D) substantially of a form p(D) = x(D)[h(D)-1], such that c(D) is selected based only upon p(D); and
    • the receiving unit has:
      • decoding means, responsive to the received sequence r(D), for providing an estimated output sequence y(D), and
      • recovery means, responsive to the estimated output sequence y(D), for substantially recovering an estimated sequence u(D) for a sequence u(D) for a transmitted signal point sequence x(D), selected from a subset of all possible signal point sequences that are of a form u(D) + d(D), wherein u(D) is a signal point sequence from a translate of said trellis code C and uniquely represents said digital data sequence and wherein d(D) represents a nonzero difference between a selected nonzero code sequence c(D) from said trellis code C and a post-cursor intersymbol interference (ISI) sequence p(D) substantially of a form p(D) = x(D)[h(D)-1], such that c(D) is selected based only upon p(D).
Anspruch[fr]
  1. Dispositif permettant de mapper une séquence de données numériques en une séquence de points de signaux x(D) à transmettre par un canal doté d'une réponse non idéale h(D) utilisant un code en treillis C, le dispositif comprenant :

       une unité de précodage (100) permettant de sélectionner ladite séquence de points de signaux x(D) à partir d'un sous-ensemble composé de toutes les séquences de points de signaux possibles présentées sous une forme u(D) + d(D), dans laquelle u(D) est une séquence de points de signaux provenant d'une translation dudit code en treillis C et représente uniquement ladite séquence de données numériques et dans laquelle d(D) représente une différence non nulle entre une séquence de code non nulle sélectionnée c(D) à partir dudit code en treillis C et une séquence de brouillages intersymboles (ISI) post-curseur p(D) présentée sensiblement sous une forme p(D) = x(D)[h(D) - 1], telle que c(D) est sélectionnée uniquement sur la base de p(D).
  2. Dispositif selon la revendication 1 dans lequel l'unité de précodage (100) comprend :
    • des moyens de mappage (202) permettant de mapper la séquence de données numériques en une séquence de points de signaux u(D) sélectionnée à partir d'une translation d'un code en treillis C ;
    • des moyens combinatoires (204, 304) sensibles à la séquence de points de signaux u(D), permettant de combiner la séquence de points de signaux u(D) et une séquence de code c(D) afin de fournir une séquence de sortie de multiplexeur s(D) ;
    • des moyens (204, 306, 512) permettant de générer la séquence de points de signaux x(D) sensiblement sous la forme s(D) - p(D) en soustrayant sensiblement une séquence de brouillages intersymboles (ISI) post-curseur p(D) de la séquence de sortie de multiplexeur s(D) ;
    • des moyens de filtrage (208, 408), couplés de manière fonctionnelle pour recevoir l'une des séquences parmi :
    • la séquence de sortie du multiplexeur s(D) ; et
    • la séquence de points de signaux x(D), lesdits moyens de filtrage fournissant la séquence de brouillages intersymboles (ISI) post-curseur p(D) ; et
    • des moyens de découpage (210), couplés de manière fonctionnelle pour recevoir la séquence de brouillages intersymboles (ISI) post-curseur p(D), permettant de découper la séquence post-curseur p(D) en séquence de code sélectionnée de points de signaux c(D) sélectionnée à partir du code en treillis C.
  3. Dispositif selon la revendication 2, dans lequel les moyens de mappage comprennent des moyens de mise en forme de constellation.
  4. Récepteur de communications numérique doté de moyens de décodage et de récupération permettant de recevoir une séquence reçue r(D) d'une séquence de données numériques qui avait été mappée en une séquence de points de signaux x(D) et transmise par un canal doté d'une réponse non idéale h(D) utilisant un code en treillis C, caractérisé en ce que :
    • les moyens de décodage (612) sont sensibles à la séquence reçue r(D) pour fournir une séquence de sortie estimée dy(D) ; et
    • les moyens de récupération (614) sont sensibles à la séquence de sortie estimée dy(D) pour récupérer sensiblement une séquence estimée du(D) pour une séquence u(D) pour une séquence de points de signaux transmis x(D), sélectionnée à partir d'un sous-ensemble composé de toutes les séquences de points de signaux possibles présentées sous une forme u(D) + d(D), dans laquelle u(D) est une séquence de points de signaux provenant d'une translation dudit code en treillis C et représente uniquement ladite séquence de données numériques et dans laquelle d(D) représente une différence non nulle entre une séquence de code non nulle sélectionnée c(D) à partir dudit codé en treillis C et une séquence de brouillages intersymboles (ISI) post-curseur p(D) présentée sensiblement sous une forme p(D) = x(D)[h(D) -1], telle que c(D) est sélectionnée uniquement sur la base de p(D).
  5. Récepteur de communications numériques selon la revendication 4, dans lequel les moyens de récupération comprennent :
    • des moyens de filtrage de récupération (702), sensibles à la séquence de sortie estimée dy(D), permettant de fournir une séquence de brouillages intersymboles (ISI) post-curseur estimée dp(D) ;
    • des moyens de découpage de récupération (704) sensibles à la séquence de brouillages intersymboles (ISI) post-curseur estimée dp(D), permettant de fournir une séquence de code non nulle estimée dc(D) à partir dudit code en treillis C qui soit sensiblement en corrélation avec c(D) utilisée pour fournir la séquence de transmission x(D) ;
    • des moyens combinatoires de récupération (706), sensibles à la séquence de sortie estimée dy(D) et à la séquence de code non nulle estimée dc(D), permettant de déterminer la séquence estimée du(D), sensiblement présentée sous une forme du(D) = dy(D) - dc(D) ; et
    • des moyens de mappage inverse (708) sensibles à la séquence estimée du(D), permettant de mapper en inverse la séquence estimée du(D) pour fournir une séquence de données numériques récupérées.
  6. Récepteur de communications numérique selon la revendication 5, dans lequel lesdits moyens de mappage inverses comprennent des moyens de récupération de mise en forme de constellation.
  7. Procédé permettant de précoder une séquence de données numériques, caractérisée en ce que le procédé utilise une séquence de points de signaux u(D) choisie à partir d'une translation d'un code en treillis C à transmettre par une voie temporelle discrète dotée d'une réponse impulsionnelle h(D), dans laquelle p(D) représente une séquence de brouillages intersymboles (ISI) post-curseur, et c(D) est une séquence de code non nulle de points de signaux c(D), dans laquelle c(D) est obtenue à partir d'une version non translatée du code en treillis C, le procédé comprenant les étapes consistant à :
    • mapper (902) la séquence de données numériques en une séquence de points de signaux u(D) ;
    • totaliser (904) les séquences u(D), c(D) et p(D) pour fournir une séquence de transmission x(D) présentée sensiblement sous une forme x(D) = u(D) + c(D) - p(D) ;
    • filtrer (906) la séquence x(D) pour obtenir une séquence de brouillages intersymboles (ISI) post-curseur p(D) présentée sensiblement sous une forme p(D) = x(D)[h(D) - 1] ; et
    • découper (908) la séquence de brouillages intersymboles (ISI) post-curseur p(D) sur une base symbole-par-symbole pour obtenir la séquence de code c(D).
  8. Système de communications numériques (600) doté d'une unité de transmission permettant de mapper une séquence de données numériques en une séquence de points de signaux x(D) à transmettre par une voie dotée d'une réponse non idéale h(D) utilisant un code en treillis C et une unité de réception permettant de recevoir une séquence reçue r (D) dudit canal, caractérisé en ce que :
    • l'unité d'émission est dotée d'une unité de précodage permettant de sélectionner ladite séquence de points de signaux x(D) à partir d'un sous-ensemble composé de toutes les séquences de points de signaux possibles présentées sous une forme u(D) + d(D), dans laquelle u(D) est une séquence de points de signaux provenant d'une translation dudit code en treillis C et représente uniquement ladite séquence de données numériques et dans laquelle d(D) représente une différence non nulle entre une séquence de code non nulle sélectionnée c(D) à partir dudit code en treillis C et une séquence de brouillages intersymboles (ISI) post-curseur p(D) présentée sensiblement sous une forme p(D) = x(D)[h(D) - 1], telle que c(D) est sélectionnée uniquement sur la base de p(D) ; et
    • l'unité de réception est dotée de :
    • moyens de décodage, sensibles à la séquence reçue r(D), permettant une séquence de sortie estimée y(D), et
    • des moyens de récupération, sensibles à la séquence de sortie estimée y(D), permettant de récupérer sensiblement une séquence estimée u(D) pour une séquence u(D) pour une séquence de points de signaux transmise x(D), sélectionnée à partir d'un sous-ensemble composé de toutes les séquences de points de signaux possibles présentées sous une forme u(D) + d(D), dans laquelle u(D) est une séquence de points de signaux provenant d'une translation dudit code en treillis C et représente uniquement ladite séquence de données numériques et dans laquelle d(D) représente une différence non nulle entre une séquence de code non nulle sélectionnée c(D) à partir dudit code en treillis C et une séquence de brouillages intersymboles (ISI) post-curseur p(D) présentée sensiblement sous une forme p(D) = x(D)[h(D) - 1], telle que c(D) est sélectionnée uniquement sur la base de p(D).






IPC
A Täglicher Lebensbedarf
B Arbeitsverfahren; Transportieren
C Chemie; Hüttenwesen
D Textilien; Papier
E Bauwesen; Erdbohren; Bergbau
F Maschinenbau; Beleuchtung; Heizung; Waffen; Sprengen
G Physik
H Elektrotechnik

Anmelder
Datum

Patentrecherche

Patent Zeichnungen (PDF)

Copyright © 2008 Patent-De Alle Rechte vorbehalten. eMail: info@patent-de.com