HK1103184A - Fast soft value computation methods for gray-mapped qam - Google Patents

Description

Fast soft value calculation method for Gray-mapping QAM

Technical Field

The present invention relates generally to digital communication systems and more particularly to the use of QAM communication signals in such systems.

Background

Evolving wireless communication standards are increasingly focusing on achieving higher data rates while maintaining acceptable communication reliability. These efforts typically involve the use of higher order modulation methods, which are more complex than the modulation standards used in earlier systems. For example, the evolving wideband cdma (wcdma) standard employs 16-ary quadrature amplitude modulation (16QAM) for use in the high speed downlink shared channels (HS-DSCHs) defined by those standards, in contrast to the relatively simple constant envelope frequency modulation employed in the original Analog Mobile Phone System (AMPS). Other developing third generation ("3G") and fourth generation ("4G") radio communication systems also employ some form of higher order QAM, with some systems using or studying the use of 64QAM and higher orders.

A receiver, such as a wireless communication terminal device, receiving such a signal must "map" the received symbols to a defined modulation constellation corresponding to the particular degree of QAM used. For example 16QAM defines sixteen constellation points, each defined by a unique pair of phase and amplitude, and each representing a unique four-bit value. Thus, the source information bits are mapped four bits at a time to corresponding 16QAM modulation symbols, which are ultimately transmitted over the associated carrier frequency signal. The simplified form of receiver operation thus becomes to determine what symbols were received based on where the received symbols fall within the defined modulation constellation as estimated from the amplitude and phase of the received symbols. Each nominal 16QAM constellation comprises four rows of four constellation points, symmetrically distributed about an x-y (real-imaginary) origin at desired point intervals.

In one type of coding for 16QAM, the modulation symbols are gray coded, where the binary representations of each adjacent modulation symbol differ by one bit. There are various methods for demodulating gray coded QAM. In general, rather than using "hard" decoding decisions, such as "1" or "0" per bit decisions, the receiver prefers some form of "soft" decoding in which individual bits transmitted by received QAM symbols are estimated or otherwise assigned a "confidence" weighting indicative of the quality of the decision per bit. In the context of gray coded QAM, such bit soft value calculations may be performed using region-specific equations, where performing the calculation of a given bit soft value depends on the particular region of the modulation constellation in which the received symbol is located. Such an approach results in computational inefficiency due to the selection logic overhead associated with selecting the appropriate equation for each region.

Previous approaches have overcome the low performance problem of the region scheme by proposing simplified soft value equations that generate two or more constellation regions, thereby obviating the need for soft value equations per region. However, since these methods are based on simplified approximations, there is no exact solution in that the bit soft values resulting from performing the simplified equations do not exactly match the results from performing the full region-specific equations.

Disclosure of Invention

The present invention includes methods and apparatus that provide computationally efficient but accurate bit soft value computation for gray coded Quadrature Amplitude Modulation (QAM). By way of non-limiting example, the present invention may be advantageously applied in mobile terminals and other wireless communication receivers operating in wideband CDMA communication networks that use 16-ary QAM (16QAM) over high speed downlink shared channels (HS-DSCHs).

More broadly, in one exemplary embodiment, the present invention includes a method of calculating bit soft values for gray coded modulation symbols in a received communication signal. The method includes providing a set of uniform equations, each corresponding to a bit position of a bit soft value to be calculated, and generating for the bit position a solution as accurate as an available region equation selected from a set of region equations associated with a nominal modulation constellation; generating symbol samples of gray coded modulation symbols in the received communication signal, each symbol sample comprising a real part and an imaginary part; the symbol samples or the nominal modulation constellation are scaled (scaling) to compensate for the amplitude of the received modulation symbols. For each bit position to be determined, soft bit values are then determined for each scaled symbol sample based on calculating the soft bit values by solving corresponding uniform equations using the real or imaginary parts of the bit position dependent symbol samples.

The above-described methods, or variations of embodiments thereof, may be implemented in an ASIC, FPGA or other such logic circuit, and may be implemented as stored program instructions executed by a DSP or other microprocessor. For example, a baseband DSP in a mobile terminal or other wireless communication receiver may be configured to perform the unified equations described above, thereby enabling a computationally efficient mechanism for obtaining bit soft values corresponding to received gray-coded QAM symbols. The invention is, of course, not limited thereto and additional features and advantages will become apparent to those skilled in the art upon review of the following description and by reference to the accompanying drawings.

Drawings

Fig. 1 is a schematic diagram of a wireless communication network constructed in accordance with an exemplary embodiment of the invention;

FIG. 2 is a schematic diagram of a mobile terminal constructed in accordance with an exemplary embodiment of the invention;

fig. 3A is a schematic diagram of an example nominal modulation constellation for 16 QAM; and

FIG. 3B is a diagram of a corresponding exemplary Gray code bit mapping;

fig. 4 is a diagram of exemplary processing logic configured to perform bit soft value calculations for received QAM symbols in accordance with one or more exemplary embodiments of the present invention;

fig. 5 is a schematic diagram of an example nominal modulation constellation for 64 QAM.

Detailed Description

Fig. 1 IS a schematic diagram of an exemplary communication network 10 that may be constructed in accordance with various wireless communication network standards including WCDMA, IS-2000(cdma2000), and the like. Those skilled in the art will appreciate that the illustrations simplify certain details of the network 10, but such details are not necessary to an understanding or discussion of the present invention. In actual implementations, network 10 may include more or different entities than shown, and further, some or all of the terms may vary depending on the particular network standard involved.

Regardless of the exemplary network 10 communicatively coupling a plurality of mobile terminals (one mobile terminal 12 is shown for simplicity) to one or more external networks 16, the external networks 16 may include the internet and/or other Public Data Networks (PDNs), Public Switched Telephone Networks (PSTNs), and other communication/data networks. A Radio Access Network (RAN)16 cooperates with one or more "core networks" (CNs) 18 to communicate data to and from the mobile stations 12. These CNs may include a Packet Switched Core Network (PSCN) configured to carry packet data in and out of network 10 and/or a Circuit Switched Core Network (CSCN) configured to carry circuit switched data, such as 64 kbit PCM voice and data, in and out of network 10.

Regardless of the specific details of the network 10, it is contemplated that the RAN 16 will transmit forward link communication signals to the mobile terminals 12 and will receive reverse link communication signals from the terminals. A Base Station (BS)20 or some other transceiver entity within the network 10 is capable of supporting such transmission and reception. Indeed, as a simplification of the foregoing, the RAN 16 may include multiple BSs 20 and the mobile terminal 12 may communicate with one or more BSs 20 or other network transceivers simultaneously. Regardless, for purposes of discussion, the mobile terminal 12 receives at least one forward link traffic (or control) channel signal carrying QAM symbol information. By way of non-limiting example, the mobile terminal 12 receives packet data signals from the network 10 using 16QAM (or some other level of QAM).

To support the reception of QAM data signals along with other traffic, control and broadcast channel signals, figure 2 shows exemplary details of the mobile terminal 12. The illustrated mobile terminal 12 includes a transmit/receive antenna 30, associated duplexer and/or switching circuitry 32, receiver front-end circuitry 34, transmitter circuitry 36, baseband processor circuitry 38 including soft value processing circuitry 40, or one or more memory/storage devices 42, a system controller 44, and a User Interface (UI)46, which may include, for example, a display screen, keypad, speaker, microphone, etc.

It should be noted that the illustrated mobile terminal 12 may comprise a radiotelephone for a wireless (cellular) communication network, although it is understood that the term as used herein is intended to have a broader meaning. Indeed, the term "mobile terminal" as used herein refers to any wireless communication device such as pagers, computer memory cards, Portable Digital Assistants (PDAs), laptop/notebook/palmtop computers with external or internal wireless connectivity and virtually any other "pervasive" computing device.

In any event, mobile terminal 12 receives and demodulates QAM signals transmitted by network 10. Fig. 3A shows a nominal modulation constellation for an embodiment of a 16QAM signal that involves gray coding, where the nominal modulation constellation includes 16 symmetrically spaced constellation points, each point representing a unique sequential combination of four bits. FIG. 3B illustrates an example Gray coded bit map for four bit positions of 16QAM, where the first (leftmost) bit position representsWatch r_xThe next bit position represents r_yNext bit position represents an inner/outer column and next (last) bit position represents an inner/outer row. The BS 20 uses a representation of this constellation to modulate the source information bits sent to the mobile terminals 12, and the mobile terminals 12 store or otherwise maintain a logical representation of the same constellation used to demodulate the symbols that they receive.

Fig. 4 broadly illustrates an exemplary demodulation process that results in the generation of "accurate" bit soft values in accordance with the present invention. In general, the process "begins" with mobile terminal 12 receiving a stream of gray coded modulated symbols contained in a communication signal transmitted by network 10 (step 100). Since the received symbol values are subject to channel degradations such as phase drift, fading, etc., they typically do not correspond very well to points in the nominal modulation constellation. (however, the received symbols maintain their relative constellation positions with respect to each other).

From a process simplification standpoint, the mobile terminal 12 "scales" the received symbols, i.e., compensates them for channel effects, gain, etc., so that the symbols are "normalized" with respect to the nominal modulation constellation (step 102). This process, when explained in more detail later in the exemplary embodiment, includes at least scaling the received symbol amplitude to normalize it relative to the amplitude scaling of the nominal modulation constellation. Optionally, the mobile terminal 12 further scales the symbols to compensate them for the measured noise in the received signal samples before, after, or simultaneously with any other scaling, such as: in an exemplary embodiment, the scaled symbol samples (or calculated bit soft values) are scaled or otherwise compensated for the signal-to-noise ratio of the received symbol samples. Alternatively, the scaling is based on the average energy of the received symbol samples.

Once the scaled symbol samples are generated, the mobile terminal 12 then calculates the bit soft value for each bit position of the modulation symbol value using an (accurate) unified equation that provides the same soft value calculation results as the applicable region-specific bit soft value equation for the nominal modulation constellation. Alternatively, these unified equations may include so-called "correction terms" that account for the effect of each scaled symbol sample on two or more "nearest" neighbors in the nominal modulation constellation.

In more detail, the receiver front end 34 may filter, amplify, and down-sample the received QAM signal to produce streams of digital samples, e.g., in-phase (I) and quadrature (Q) samples corresponding to real and imaginary signal components, for input to the baseband processor 38. Each received symbol in the sample stream may be referred to asAnd, after appropriate RAKE-type combining, the symbols are represented as,

here, theIs a transmitted symbol having a unit average symbol energy, h is a vector of the multipath channel response for the channel of interest (e.g., the HS-DSCH of a WCDMA network), h is_rIs the channel response of the reference channel (e.g., the transmitted pilot or other channel associated with the QAM channel of interest), g is the channel gain offset between the data channel and the reference channel (e.g., the reference traffic channel gain), w is the RAKE-type combining weights, andis an interference or noise term. (Note that exemplary receiver front end34 may include RAKE fingers (fingers) and a combining circuit or baseband processor 38 may provide all or a portion of such RAKE receiver functionality. ) These parameters are typically estimated from pilot channels or symbols in a CDMA system. Information regarding exemplary estimation METHODs is given in the Yi-Pin Eric Wang co-pending AND commonly assigned patent application entitled "METHOD AND APPARATUS FOR SIGNAL DEMODULATION". This application, filed on 26/9/2003, application No.10/672492, is hereby incorporated by reference.

The mobile terminal 12 is configured to scale the received symbols so as to normalize them with respect to a nominal modulation constellation and thereby simplify the calculation of bit soft values. More particularly, baseband processor 38 (e.g., processing circuitry 40) may be configured to perform such scaling. (note that in the nominal modulation constellation shown, the nearest point-to-point interval is two, and this determines the appropriate scaling, and also affects one or more coefficients appearing in the unified equation described later herein.) an exemplary scaling for 16QAM is shown in figure 3A,

where s is a converted transmission symbol where the inter-symbol distance is two and n is a symbol having a length of n

Scaled interference and noise for the given variance.

The log-likelihood of the next symbol can be expressed as

Using this, the mobile terminal 12 may be configured to perform the complete demodulation process for 16QAM as follows:

1. estimation of h from a reference channel (common pilot channel (CPICH) signal or pilot symbol)_r. The combining weights are calculated according to a suitable RAKE-type combiner principle.

2. A signal-to-noise ratio (SNR) is estimated. Using this estimate, the scaling can be expressed as

Using this estimate, the variance of the scaled noise is

Optionally, the SNR is approximated by the average energy of the symbol samples.

3. Scaling the combined symbol values relative to the nominal modulation constellation is used to normalize and optionally scale the noise as described above.

4. Soft values at the bit level are calculated from the received symbols.

Using the appropriate scaling given in (2), a transmitted 16QAM symbol for a forward link channel signal received by the mobile terminal 12, such as an HS-DSCH signal in a WCDMA network, may be considered one symbol of the gray mapped 16QAM constellation shown in fig. 3A. Optimal LogMAP (maximum priority) bit soft value according to

Is calculated out on the basis of the data of the received signal,

here { *: B_i(*) ═ b } is the set of constellation points * with the ith bit b and the COM operation is defined as

Although there is a way to approximate COM operation, performing equation (7) can be computationally expensive. By "max" operation, one such approximation replaces COM operation, leading to a LogMax demodulation algorithm:

in equation (9), a soft value of the ith bit may be obtained by subtracting the maximum symbol log likelihood in the tuple with the ith bit of 0 from the maximum symbol log likelihood in the tuple with the ith bit of 1. Equation (9) is referred to herein as a "direct LogMax" demodulator. One drawback of direct LogMax is its computational overhead. For the exemplary 16QAM shown in fig. 3A, approximately twenty additions and eight multiplications are required per bit soft value. In comparison, using the unified equation provided by the present invention, only about two additions and one multiplication are required for each soft value.

In the exemplary 16QAM constellation shown in fig. 3A, the first bit of the gray mapping determines the sign of the real part, i.e. whether the symbol is to the left or to the right of the y-axis. Similarly, the second bit determines whether the symbol is above or below the x-axis (r for purposes of this discussion)_xAnd r_yRepresenting the real and imaginary components, respectively, of the symbol value r). Further, as previously described, the third bit determines whether the symbol belongs to the inner two columns or the outer two columns of modulation constellation points. Similarly, the fourth bit determines whether the symbol belongs to the inner two rows or the outer two rows of modulation constellation points.

These attributes may be employed to simplify the calculation of LogMax soft values. For example, assume that the real part of the scaled combined symbol value r is located at (-2, 2). Then the LogMax soft value of the first bit is

Thus, the soft value of the first bit does not depend on the imaginary part of r. In fact, the following equation can be created to calculate the (LogMax or LogMAP) bit soft values for gray coded 16 QAM:

soft value ^ (b)₁) And ^ (b)₃) Dependent only on r_x。

Soft value ^ (b)₂) And ^ (b)₄) Dependent only on r_y。

These attributes can be noted in U.S. patent No.6078626 to Ramesh, where they are employed to develop "non-exact" equations for generating approximate bit soft values. As used herein, the term "inaccurate" means that the bit soft value calculation cannot provide the same bit soft values as those derived from the applicable "area" equation corresponding to the constellation modulation region in which the symbol of interest is located. These details are given in the more complete description below.

Using the sign/column/row properties of gray coded bit positions, it can be shown that the bit soft values for various bit positions can be determined according to an applicable one of the aforementioned region equations. The constellation for fig. 3A may be shown:

may be ^ (b) with respect to the imaginary symbol component₂) And ^ (b)₄) Similar results were derived. These results provide the basis for the calculation of the soft values of the bits according to the appropriate region equations, i.e. the particular equation used to calculate the soft value for a given bit position varies according to the region of the modulation constellation in which the received (scaled) symbol value is located. Thus, as part of decoding a series of received symbols, the computational logic must (1) identify the region of each symbol, and (2) select and execute the applicable regionA domain equation. Such operations are not conducive to fast bit soft value calculations, and the overhead associated with the required region determination logic adds undesirable complexity to performing bit soft value calculations on a logic basis.

The present invention provides a unified equation to alter this approach, i.e., to use the same unified equation for a given bit position regardless of the constellation region involved. Such equations remove the need for the overhead of a sequential "look-up" operation with respect to the region equation method and still provide as accurate a solution as that obtained from the region equation. As used herein, the exemplary unified equation method of the present invention is generally referred to as the fast and accurate LogMax method. The area equations given above can be expressed as a unified equation according to the fast and accurate LogMax method,

here, the

It can be seen in equation (11) that the power σ is based on the noise_n ²Optional scaling of the reciprocal.

Optionally, the unified equation described above may be compensated to account for "nearest" neighbor effects. That is, the soft bit values calculated by the unified equation above may be compensated to account for two or more nearest constellation points within the nominal modulation constellation. Such methods add one or more correction terms to the unified equation and use these compensated unified equations for bit soft value calculations are referred to herein as "fast Login" methods because the compensated unified equation approaches the optimal LogMax solution in terms of performance. Thus, the correction term can be viewed as a "multi-region" compensation term that incorporates the effects of the constellation points of two or more regions into a unified equation.

An exemplary unified equation according to the fast LogLin method is given as follows

Wherein

And

efficient execution of the expression x + f using a pipeline_c(x) It is advantageous wherein the processing circuitry 38 and, in particular, the soft value processor 40 of the mobile terminal 12 comprise one or more ASICs, microprocessors or other logic processing devices.

Of course, the present invention is applicable to applications other than 16 QAM. Corresponding to the exemplary nominal 64QAM constellation shown in fig. 5, the fast and accurate LogMax method according to the present invention is given by

Wherein

Those skilled in the art will appreciate that additional versions of the unified equation may be given for QAM of essentially any order, and that the particular coefficients appearing in the above equations, such as "2", "4" and "8", may vary or change depending on, for example, the nominal modulation constellation mapped to and the scaling factor. Further, referring again to fig. 3A, the four-bit position definition of 16QAM can be changed and this bit reconfiguration will change which of the four equations is used for each four bit position. That is, the form of the equation does not change with the reordering of the QAM bits, but the particular unified equation used to calculate the bit soft values for a particular QAM bit position changes.

For example, referring to the bit mapping for the nominal modulation constellation given in fig. 3B, the first (leftmost) bit indicates whether the symbol is to the left or right of the imaginary (quadrature) axis, and the second bit position indicates whether the symbol is above or below the real (in-phase) axis. For that particular bitmap, the values shown for λ by equation (12) are used₁Of is dependent on r_xThe unified equation of (a) obtains the bit soft value for the first bit position. Likewise, the use shown for λ₂' is dependent on r_yThe unified equation of (a) obtains the bit soft value for the second bit position. If one wants to swap the bit definitions such that the first bit position represents up/down and the second bit position represents left/right, then the unified equations can be swapped accordingly. Note also that with such a swap, the bit soft value of the first bit position will depend on r_yAnd the soft bit value of the second bit position will depend on r_x。

Illustratively, if one wants to swap the first and third bits in the bit maps depicted in fig. 3A and 3B, then λ is in equation (12)₁' is renamed to λ₃' and λ₃Is heavily hitNamed as lambda₁'. Continuing with the proposed bit swapping of the second and third bits of the 16QAM constellation of fig. 3A, the soft values may be calculated as follows:

wherein

Thus, one skilled in the art will appreciate that the form of the unified equation does not change with changing bit position definitions, but the particular unified equation for each bit position and whether or not the bit position is changedIs dependent on r_xAnd r_yMay vary.

Then, generally, for the exemplary modulation constellation in fig. 3A, the modulation constellation is represented by λ_i′＝-4r_x+|r_x+2|-|r_x-2| giving the bit soft value of the bit position of the symbol representing the real part of the symbol sample, by λ_i′＝-4r_y+|r_y+2|-|r_y-2| giving the soft bit value of the bit position of the symbol representing the imaginary part of the symbol sample, by λ_i′＝-4+2|r_xI gives the number used to indicate whether the symbol sample is located in the inner or outer two columns of the constellation, and is represented by_i′＝-4+2|r_yAnd | gives a value indicating whether the symbol samples are located in the inner two rows or the outer two rows. A similar generalization is apparent for the exemplary 64QAM constellation shown in fig. 5.

While the above exemplary equations normalize the symbol samples using one or more scaling values with respect to a nominal modulation constellation, one or more exemplary embodiments of the present invention may employ an alternative approach in which the nominal modulation constellation is scaled as a function of the amplitude of the received signal. That is, instead of normalizing the amplitude metric (amplitude scale) of the received symbol to match the amplitude metric of the nominal modulation constellation, the nominal modulation constellation is scaled to match the amplitude metric of the received symbol. Note that the soft value processor 40 and/or the baseband processor 38 of the mobile terminal 12 may be configured as scaling circuitry or used to scale the symbol samples to generate scaled symbol samples to scale the nominal modulation constellation to generate a scaled nominal modulation constellation.

Using the constellation scaling method, 2a is the interval of the scaled nominal modulation constellation, i.e. the distance of the nearest two symbols in the scaled nominal modulation constellation. The equation for the 16QAM constellation in fig. 3A according to this method is given as follows:

a similar constellation scaling applied to the 64QAM case is shown as such,

of course, in the unified equations above, the equation pairs differ only by using the real or imaginary part of the symbol. Thus, rather than storing six unified equations as in the 64QAM example given above, only three equations are stored. Then, to calculate the bit soft values for a given bit position, processing logic selects the appropriate unified equation and the appropriate one of the real and imaginary parts of the symbol samples based on the particular gray code mapping used in a given application.

Regardless of the exemplary mobile terminal 12 may be configured to store or otherwise implement logical representations of the unified equations, such as storing them in the storage device 42 as encoded program instructions, query values, and the like. In this regard, the present invention may be implemented in hardware, software, or a combination thereof. It will be appreciated that the present invention may be embodied in whole or in part as stored program instructions, microcode, software, or any other stored logic program representation that is executed by a logic processing circuit that, in an exemplary embodiment, includes a soft value processor 40 implemented in whole or in part by a DSP, microcontroller, microprocessor, ASIC, FPGA, Programmable Logic Device (PLD), or other type of processor.

It is also understood that the soft value processor may be implemented independently of the baseband processing functions and that the physical circuitry of the mobile terminal 12 may be implemented differently than shown without departing from the scope of this disclosure. Indeed, the invention is not limited to the foregoing detailed description. Rather, the present invention is limited only by the following claims and their reasonable equivalents.

Claims (89)

1. A method of calculating bit soft values for gray coded modulation symbols in a received communication signal, comprising:

providing a set of unified equations, each unified equation corresponding to a bit position of a bit soft value to be computed and producing a solution as accurate as an available region equation selected from a set of region equations associated with a nominal modulation constellation for the bit position;

generating symbol samples of gray coded modulation symbols in the received communication signal, each symbol sample comprising a real part and an imaginary part;

scaling the symbol samples or the nominal modulation constellation to compensate for the amplitude of the received modulation symbols;

the bit soft values are determined for each symbol sample for each bit position to be determined based on calculating the bit soft values by solving the corresponding unified equations using the real or imaginary parts of the symbol samples depending on the bit position.

2. The method of claim 1, wherein scaling the symbol samples or the nominal modulation constellation to compensate for the amplitude of the received modulation symbols comprises scaling the symbol samples to normalize their amplitude relative to a nominal modulation constellation position.

3. The method of claim 2, wherein scaling the symbol samples to normalize their amplitudes relative to a nominal modulation constellation comprises compensating the symbol samples for channel and filter effects and gains associated with the received communication symbol.

4. The method of claim 3, wherein the received communication signal comprises a traffic channel signal transmitted with a gain relative to a reference signal transmitted in relation to the traffic channel signal, and wherein compensating the symbol samples for effects and gains of the channel and filter comprises compensating for the reference traffic channel gain.

5. The method of claim 2, wherein scaling the symbol samples to normalize their amplitudes relative to the nominal modulation constellation comprises forming a scaling value as a ratio of a square root of an average nominal energy of the nominal modulation constellation to an average signal-to-noise ratio of the symbol samples, and scaling the symbol samples based on the scaling value.

6. The method of claim 2, wherein scaling the symbol samples to normalize their amplitudes relative to the nominal modulation constellation comprises forming a scaling value as a ratio of a square root of an average nominal energy of the nominal modulation constellation and an average energy of the symbol samples, and scaling the symbol samples based on the scaling value.

7. The method of claim 1, wherein scaling the symbol samples or the nominal modulation constellation to compensate for the amplitude of the received modulation symbols comprises scaling the nominal modulation constellation based on the amplitude of the received modulation symbols.

8. The method of claim 1, wherein scaling the symbol samples or the nominal modulation constellation to compensate for the amplitude of the received modulation symbols comprises scaling the nominal modulation constellation.

9. The method of claim 8, wherein scaling the nominal modulation constellation comprises scaling a nominal symbol interval of the nominal modulation constellation based on an average signal-to-noise ratio of symbol samples.

10. The method of claim 8, wherein scaling the nominal modulation constellation comprises scaling a nominal symbol interval of the nominal modulation constellation based on an average energy of symbol samples.

11. The method of claim 1, wherein providing a set of unified equations, each unified equation corresponding to a bit position of the bit soft values to be computed and producing a solution as accurate as an available region equation selected from a set of region equations associated with a nominal modulation constellation for the bit position comprises providing stored program instructions in a digital memory that implement the set of unified equations.

12. The method of claim 1, wherein providing a set of uniform equations, each uniform equation corresponding to a bit position of the bit soft values to be calculated and producing a solution as accurate as an available region equation selected from a set of region equations associated with a nominal modulation constellation for the bit position comprises providing a uniform equation for each bit position that includes all of the region equation terms associated with that bit position.

13. The method of claim 1, wherein providing a set of uniform equations for 16QAM comprises providing a first uniform equation for calculating bit soft values corresponding to sign bits of the symbol samples, wherein the first uniform equation generates a first term based on a magnitude of a sum of real or imaginary parts of the symbol samples and a nominal symbol interval, generates a second term based on a magnitude of a difference of the real or imaginary parts of the symbol samples and the nominal symbol interval, generates a third term based on four times the real or imaginary parts of the symbol samples, and subtracts the second and third terms from the first term.

14. The method of claim 13, wherein providing a set of uniform equations for 16QAM comprises providing a second uniform equation for calculating soft bit values for bit positions used to represent symbol samples located in inner or outer rows of a nominal modulation constellation or symbol samples located in inner or outer columns of the nominal modulation constellation, and wherein the second uniform equation generates a first term based on twice a real or imaginary amplitude of a symbol sample, generates a second term based on twice a nominal symbol interval, and subtracts the second term from the first term.

15. The method of claim 1, wherein providing a set of uniform equations for 16QAM comprises providing a first uniform equation for calculating soft bit values for bit positions, the bit positions corresponding to signs of real and imaginary parts, and providing a second uniform equation for calculating soft bit values for bit positions, the bit positions indicating that the symbol samples are located in an inner or outer column of a nominal modulation constellation and the bit positions indicating that the symbol samples are located in an inner or outer row of the nominal modulation constellation.

16. The method of claim 1, wherein providing a set of unified equations comprises providing a unified equation of the form,

wherein λ_1...4' corresponds to four bit positions, r, associated with 16QAM_xAnd r_yRespectively the real and imaginary parts of the symbol samples, and 2a is the interval of the nominal modulation constellation.

17. The method of claim 16, further comprising using for λ₁' to calculate bit soft values for bit positions of the symbol used to represent the real part of the symbol sample, using the equation for λ₂The unified equation of' calculates the soft bit values for the bit positions of the symbols representing the imaginary part of the symbol samples, using the values for λ₃' the unified equation is used to represent the symbol samplesCalculating soft bit values using bit positions located in inner or outer columns of a nominal modulation constellation, and using the soft bit values for λ₄The unified equation of' is to compute bit soft values for bit positions that represent symbol samples located in inner or outer rows of a nominal modulation constellation.

18. The method of claim 1, wherein providing a set of uniform equations for 64QAM comprises providing uniform equations of the form:

wherein λ_1...6' corresponds to six bit positions, r, associated with 64QAM_xAnd r_yRespectively the real and imaginary parts of the symbol samples, and 2a is the interval of the nominal modulation constellation.

19. The method of claim 1, further comprising scaling the computed bit soft values by a signal-to-noise ratio of the symbol samples.

20. The method of claim 1, further comprising calculating bit soft values by average energy scaling of symbol samples.

21. The method of claim 1, further comprising scaling the computed bit soft values as a function of noise in the symbol samples.

22. The method of claim 1, further comprising scaling the calculated soft bit values based on a reciprocal of a noise power in the symbol samples.

23. The method of claim 1, further comprising providing, for each symbol sample, one or more correction terms for each unified equation that compensate the bit soft values for one or more additional nearest neighbors of the symbol sample in the nominal modulation constellation.

24. The method of claim 1, further comprising compensating the unified equation using one or more multi-region compensation terms that compensate the soft-values of the bits calculated from the unified equation for the effects of constellation points located in two or more regions of the nominal modulation constellation.

25. The method of claim 1, wherein scaling the symbol samples or the nominal modulation constellation to compensate for the amplitude of the received modulation samples comprises scaling the nominal modulation constellation to obtain a scaled nominal modulation constellation.

26. The method of claim 25, wherein, for 16QAM, providing a set of unified equations comprises providing the following first and second unified equations, respectively, to calculate the bit soft values for the ith bit position:

λ_i′＝|ar_{x or y}+2a²|-|ar_{x or y}-2a²|-4ar_{x or y}and

λ_i′＝2|ar_{x or y}|-4a²

where 2a comprises the symbol interval of the scaled nominal modulation constellation, r_xIncluding the real part of the symbol sample, and r_yIncluding the imaginary part of the symbol samples.

27. The method of claim 26, further combining the real parts of the symbol samples r using a first uniform equation_xTo calculate soft bit values for bit positions representing real symbols of symbol samples, the imaginary part r of the symbol samples is combined using a first uniform equation_yTo calculate soft bit values for bit positions representing imaginary symbols of the symbol samples, the real part r of the symbol samples is combined using a second uniform equation_xTo calculate soft bit values for bit positions representing symbol samples located in inner or outer columns of a nominal modulation constellation and to use a second uniform equation in combination with the imaginary part r of the symbol samples_yTo compute bit soft values for bit positions representing symbol samples located in inner or outer rows of a nominal modulation constellation.

28. The method of claim 26, further comprising selecting a particular one of the two unified equations and r_xAnd r_yOne of the components calculates a bit soft value for the ith bit position based on the particular gray coded mapping employed by the nominal modulation constellation.

29. The method of claim 1, wherein, for 64QAM, providing a unified set of equations comprises providing the following three equations for calculating the ith bit position:

λ_i′＝-8ar_{x or y}+|ar_{x or y}+2a²|-|ar_{x or y}-2a²|+|ar_{x or y}+4a²|-|ar_{x or y}-4a²|+|ar_{x or y}+6a²|-|ar_{x or y}-6a²|，

λ_i′＝-16a²+4|ar_{x or y}|-||ar_{x or y}|-2a²|+||ar_{x or y}|-6a²|，and

λ_i′＝-4a²+|2|ar_{x or y}|-8a²|，

where 2a comprises the interval of the scaled nominal modulation constellation, r_xIncluding the real part of the symbol sample, and r_yIncluding the imaginary part of the symbol samples.

30. The method of claim 29, further comprising selecting a particular one of three unified equations and r_xAnd r_yOne of the components calculates a bit soft value for the ith bit position based on the particular gray coded mapping employed by the nominal modulation constellation.

31. An Application Specific Integrated Circuit (ASIC) for computing bit soft values from gray coded modulation symbols in a received communication signal, the ASIC comprising:

scaling circuitry for scaling symbol samples corresponding to a modulation symbol of a gray coded modulation in the received communication signal or scaling a nominal modulation constellation with respect to the symbol samples, each symbol sample comprising a real part and an imaginary part; and

a calculation circuit that performs bit soft value determination based on a set of unified equations, each unified equation corresponding to a bit position of a bit soft value to be calculated and producing a solution as accurate as a usable region equation selected from a set of region equations associated with a nominal modulation constellation for the bit position;

the calculation circuit is configured to determine a bit soft value for each symbol sample based on calculating the bit soft value for each bit position by solving a corresponding uniform equation using a real or imaginary part of the bit position dependent symbol sample.

32. The ASIC of claim 31, wherein the scaling circuit is configured to scale the symbol samples to normalize their amplitudes relative to a nominal modulation constellation.

33. The ASIC of claim 32, wherein the scaling circuit is configured to normalize symbol sample amplitudes relative to a nominal modulation constellation by compensating the symbol samples for channel and filter effects and gains associated with the received communication symbols.

34. The ASIC of claim 33, wherein the received communication signal comprises a traffic channel signal transmitted with a gain relative to a reference signal transmitted in relation to the traffic channel signal, and wherein the scaling circuit is configured to compensate the symbol samples for an effect of the reference traffic channel gain.

35. The ASIC of claim 32, wherein the scaling circuit is configured to normalize the amplitude of the symbol samples by forming a scaled value as a ratio of a square root of an average nominal energy of the nominal modulation constellation to an average signal-to-noise ratio of the symbol samples and scaling the symbol samples based on the scaled value.

36. The ASIC of claim 32, wherein the scaling circuit is configured to normalize the amplitude of the symbol samples by forming a scaled value as a ratio of a square root of an average nominal energy of the nominal modulation constellation to an average energy of the symbol samples and scaling the symbol samples based on the scaled value.

37. The ASIC of claim 31, wherein the scaling circuit is configured to scale the nominal modulation constellation based on an amplitude of the received modulation symbol.

38. The ASIC of claim 31, wherein the scaling circuit is configured to scale the nominal modulation constellation based on an average signal-to-noise ratio of the symbol samples.

39. The ASIC of claim 31, wherein the scaling circuit is configured to scale the nominal modulation constellation by scaling a nominal symbol interval of the nominal modulation constellation based on an average energy of the symbol samples.

40. The ASIC of claim 31, wherein the unified equation set comprises program instructions stored in digital memory that implement the unified equation set.

41. The ASIC of claim 31, wherein the unified equation set includes one unified equation for each bit location, the unified equation including all of the region equation terms associated with that bit location.

42. The ASIC of claim 31, wherein the unified set of equations for 16QAM includes a first unified equation for computing bit soft values corresponding to the sign bits of the symbol samples, wherein the first unified equation generates a first term based on the magnitude of the sum of the real or imaginary part of the symbol samples and the nominal symbol interval, generates a second term based on the magnitude of the difference of the real or imaginary part of the symbol samples and the nominal symbol interval, generates a third term based on four times the real or imaginary part of the symbol samples and subtracts the second and third terms from the first term.

43. The ASIC of claim 42, wherein the unified set of equations for 16QAM comprises a second unified equation for calculating soft bit values for bit positions used to represent symbol samples located in inner or outer rows of the nominal modulation constellation or symbol samples located in inner or outer columns of the nominal modulation constellation, and wherein the second unified equation produces a first term based on twice a real or imaginary amplitude of the symbol samples, produces a second term based on twice a nominal symbol interval, and subtracts the second term from the first term.

44. The ASIC of claim 31, wherein the unified set of equations for 16QAM comprises a first unified equation for calculating soft bit values for bit positions corresponding to the signs of the real and imaginary parts, and a second unified equation providing for calculating soft bit values for bit positions indicating that the symbol samples are located in the inner or outer columns of the nominal modulation constellation and that the bit positions indicate that the symbol samples are located in the inner or outer rows of the nominal modulation constellation.

45. The ASIC of claim 31, wherein the unified set of equations comprises a unified equation of the form,

wherein λ_1...4' corresponds to four bit positions, r, associated with 16QAM_xAnd r_yAre the real and imaginary parts of the symbol samples, respectively, and 2a is of the nominal modulation constellationAnd (4) spacing.

46. The ASIC of claim 45 wherein the computing circuit is configured to use the information for λ₁' to calculate bit soft values for bit positions of the symbol used to represent the real part of the symbol sample, using the equation for λ₂The unified equation of' calculates the soft bit values for the bit positions of the symbols representing the imaginary part of the symbol samples, using the values for λ₃The unified equation of' calculates bit soft values for bit positions representing symbol samples located in inner or outer columns of a nominal modulation constellation, and uses the values for λ₄The unified equation of' is to compute bit soft values for bit positions that represent symbol samples located in inner or outer rows of a nominal modulation constellation.

47. The ASIC of claim 31, wherein the unified set of equations for 64QAM comprises unified equations of the form,

wherein λ_1...6' corresponds to six bit positions, r, associated with 64QAM_xAnd r_yRespectively the real and imaginary parts of the symbol samples, and 2a is the interval of the nominal modulation constellation.

48. The ASIC of claim 31, wherein the ASIC is configured to further scale the calculated bit soft values by a signal-to-noise ratio of the symbol samples.

49. The ASIC of claim 31, wherein the ASIC is configured to further scale the calculated bit soft values by an average energy of the symbol samples.

50. The ASIC of claim 31, wherein the ASIC is configured to further scale the calculated bit soft values as a function of noise of the symbol samples.

51. The ASIC of claim 31, wherein the ASIC is configured to further scale the calculated bit soft values based on a reciprocal of a noise power in the symbol samples.

52. The ASIC of claim 31, wherein the ASIC is configured to apply, for each symbol sample, one or more correction terms for bit soft values to compensate one or more additional nearest neighbors of the symbol sample in the nominal modulation constellation for each unified equation.

53. The ASIC of claim 31, wherein the ASIC is configured to compensate the unified equation using one or more multi-region compensation terms that compensate the bit soft values calculated from the unified equation for the effects of constellation points located in two or more regions of the nominal modulation constellation.

54. The ASIC of claim 31, wherein the scaling circuit is configured to scale the nominal modulation constellation to obtain a scaled nominal modulation constellation, and wherein the unified equation set comprises first and second unified equations, respectively, for calculating the bit soft value for the ith bit position:

λ_i′＝|ar_{x or y}+2a²|-|ar_{x or y}-2a²|-4ar_{x or y}and

λ_i′＝2|ar_{x or y}|-4a²

where 2a comprises the symbol interval of the scaled nominal modulation constellation, r_xIncluding the real part of the symbol sample, and r_yIncluding the imaginary part of the symbol samples.

55. The ASIC of claim 54 wherein the ASIC is configured to combine the real parts of the symbol samples r using a first uniform equation_xTo calculate soft bit values for bit positions representing real symbols of symbol samples, the imaginary part r of the symbol samples is combined using a first uniform equation_yTo calculate soft bit values for bit positions representing imaginary symbols of the symbol samples, the real part r of the symbol samples is combined using a second uniform equation_xTo calculate soft bit values for bit positions representing symbol samples located in inner or outer columns of a nominal modulation constellation and to use a second uniform equation in combination with the imaginary part r of the symbol samples_yTo compute bit soft values for bit positions representing symbol samples located in inner or outer rows of a nominal modulation constellation.

56. The ASIC of claim 54, further comprising selecting a particular one of two unified equations and r_xAnd r_yOne of the components calculates a bit soft value for the ith bit position based on the particular gray coded mapping employed by the nominal modulation constellation.

57. The ASIC of claim 31, wherein, for 64QAM, the unified equation set includes the following three equations for calculating the ith bit position:

λ_i′＝-8ar_{x or y}+|ar_{x or y}+2a²|-|ar_{x or y}-2a²|+|ar_{x or y}+4a²|-|ar_{x or y}-4a²|+|ar_{x or y}+6a²|-|ar_{x or y}-6a²|，

λ_i′＝-16a²+4|ar_{x or y}|-||ar_{x or y}|-2a²|+||ar_{x or y}|-6a²|，

λ_i′＝-4a²+|2|ar_{x or y}|-8a²|，

where 2a comprises the interval of the scaled nominal modulation constellation, r_xIncluding the real part of the symbol sample, and r_yIncluding the imaginary part of the symbol samples.

58. The ASIC of claim 57, wherein, for each symbol sample, the ASIC is configured to select a particular one of three unified equations and r_xAnd r_yOne of the components calculates a bit soft value for the ith bit position based on the particular gray coded mapping employed by the nominal modulation constellation.

59. A computer readable medium storing a computer program for computing bit soft values from gray coded modulation symbols in a received communication signal, the computer program comprising:

program instructions to scale symbol samples corresponding to gray coded modulation symbols in a received communication signal or to scale a nominal modulation constellation relative to symbol samples, each symbol sample comprising a real part and an imaginary part; and

program instructions to apply a set of uniform equations, each uniform equation corresponding to a bit position of a bit soft value to be computed and producing a solution as accurate as a usable region equation selected from a set of region equations associated with a nominal modulation constellation for the bit position;

program instructions determine a bit soft value for each symbol sample based on calculating the bit soft value for each bit position by solving a corresponding unified equation using real or imaginary parts of the bit position dependent symbol sample.

60. The computer readable medium storing a computer program of claim 59, wherein the unified set of equations comprises one unified equation for each bit position, the unified equation including all of the region equation terms associated with that bit position.

61. The computer readable medium storing a computer program of claim 59, wherein the set of uniform equations for 16QAM includes a first uniform equation for computing bit soft values corresponding to sign bits of the symbol samples, wherein the first uniform equation generates a first term based on a magnitude of a sum of real or imaginary parts of the symbol samples and a nominal symbol interval, generates a second term based on a magnitude of a difference of the real or imaginary parts of the symbol samples and the nominal symbol interval, generates a third term based on four times the real or imaginary parts of the symbol samples, and subtracts the second and third terms from the first term.

62. The computer readable medium storing a computer program of claim 61, the set of uniform equations for 16QAM includes a second uniform equation for calculating soft values of bits for bit positions used to represent symbol samples located in inner or outer rows of a nominal modulation constellation or symbol samples located in inner or outer columns of the nominal modulation constellation, and wherein the second uniform equation generates a first term based on twice a real or imaginary amplitude of a symbol sample, generates a second term based on twice a nominal symbol interval, and subtracts the second term from the first term.

63. The computer readable medium storing a computer program of claim 59, the unified set of equations for 16QAM includes a first unified equation for computing soft bit values for bit positions corresponding to signs of the real and imaginary components, and a second unified equation for computing soft bit values for bit positions indicating that the symbol sample is located in an inner or outer column of the nominal modulation constellation and that the bit position indicates that the symbol sample is located in an inner or outer row of the nominal modulation constellation.

64. The computer readable medium storing a computer program of claim 59, wherein the set of unified equations comprises unified equations of the form,

wherein λ_1...4' corresponds to four bit positions, r, associated with 16QAM_xAnd r_yRespectively the real and imaginary parts of the symbol samples, and 2a is the interval of the nominal modulation constellation.

65. The computer readable medium storing a computer program of claim 64, wherein the program instructions to determine the bit soft values comprise using a value for λ₁' to calculate bit soft values for bit positions of the symbol used to represent the real part of the symbol sample, using the equation for λ₂The unified equation of' calculates the soft bit values for the bit positions of the symbols representing the imaginary part of the symbol samples, using the values for λ₃The unified equation of' calculates bit soft values for bit positions representing symbol samples located in inner or outer columns of a nominal modulation constellation, and uses the values for λ₄' to calculate the bit soft values for bit positions representing symbol samples located in inner or outer rows of a nominal modulation constellation.

66. The computer readable medium storing a computer program of claim 59, wherein the set of unified equations for 64QAM includes unified equations of the form,

wherein λ_1...6' corresponds to six bit positions, r, associated with 64QAM_xAnd r_yRespectively the real and imaginary parts of the symbol samples, and 2a is the interval of the nominal modulation constellation.

67. The computer readable medium storing a computer program of claim 59, wherein the computer program further comprises program instructions to, for each symbol sample, apply one or more correction terms for each unified equation that compensate for bit soft values for one or more additional nearest neighbors of the symbol sample in the nominal modulation constellation.

68. The computer readable medium storing a computer program of claim 59, wherein the computer program further comprises program instructions to compensate the unified equation using one or more multi-region compensation terms that compensate the bit soft values calculated from the unified equation for the effects of constellation points located in two or more regions of the nominal modulation constellation.

69. The computer readable medium storing a computer program of claim 59, wherein the program instructions to scale the symbol samples or the nominal modulation constellation comprise program instructions to scale the nominal modulation constellation to obtain a scaled nominal modulation constellation.

70. The computer readable medium storing a computer program of claim 69, wherein, for 16QAM, a unified equation set is provided comprising the following first and second unified equations used to calculate the bit soft values for the ith bit position, respectively:

λ_i′＝|ar_{x or y}+2a²|-|ar_{x or y} -2a²|-4ar_{x or y}and

λ_i′＝2|ar_{x or y}|-4a²

where 2a comprises the symbol interval of the scaled nominal modulation constellation, r_xIncluding the real part of the symbol sample, and r_yIncluding the imaginary part of the symbol samples.

71. The computer readable medium storing a computer program of claim 70, wherein the program instructions for determining the soft bit values comprise using a first uniform equation in conjunction with the real part r of the symbol samples_xTo calculate soft bit values for bit positions representing real symbols of symbol samples, the imaginary part r of the symbol samples is combined using a first uniform equation_yTo calculate soft bit values for bit positions representing imaginary symbols of the symbol samples, the real part r of the symbol samples is combined using a second uniform equation_xTo calculate soft bit values for bit positions representing symbol samples located in inner or outer columns of a nominal modulation constellation and to use a second uniform equation in combination with the imaginary part r of the symbol samples_yTo compute bit soft values for bit positions representing symbol samples located in rows inside or outside of a nominal modulation constellation.

72. The computer readable medium storing a computer program of claim 71, wherein the program instructions for determining the soft bit values comprise selecting a particular one of two uniform equations and r_xAnd r_yProgram instructions for calculating a bit soft value for the ith bit position based on the particular gray code mapping employed by the nominal modulation constellation for one of the components.

73. The computer readable medium storing a computer program of claim 59, wherein, for 64QAM, the unified set of equations includes the following three equations used to calculate the ith bit position:

λ_i′＝-8ar_{x or y}+|ar_{x or y}+2a²|-|ar_{x or y}-2a²|+|ar_{x or y}+4a²|-|ar_{x or y}-4a²|+|ar_{x or y}+6a²|-|ar_{x or y}-6a²|，

λ_i′＝-16a²+4|ar_{x or y}|-||ar_{x or y}|-2a²|+||ar_{x or y}|-6a²|，

λ_i′＝-4a²+|2|ar_{x or y}|-8a²|，

where 2a comprises the interval of the scaled nominal modulation constellation, r_xIncluding the real part of the symbol sample, and r_yIncluding the imaginary part of the symbol samples.

74. The computer readable medium storing a computer program of claim 73, wherein the program instructions for determining the soft bit values comprise selecting a particular one of three uniform equations and r_xAnd r_yProgram instructions for calculating a bit soft value for the ith bit position based on the particular gray code mapping employed by the nominal modulation constellation for one of the components.

75. A mobile terminal for a wireless communication network comprising:

a transmitter circuit configured to transmit a signal to a wireless communication network;

a receiver circuit configured to receive a signal from a wireless communication network, the signal comprising a received communication signal comprising a gray-coded modulation signal;

processing circuitry to process symbol samples corresponding to gray coded modulation symbols in a received communication signal, wherein the processing circuitry is configured to:

scaling symbol samples corresponding to the modulated symbols of the gray coded modulation in the received communication signal to produce symbol samples normalized with respect to a nominal modulation, each symbol sample comprising a real part and an imaginary part, or scaling a nominal modulation constellation with respect to the received symbol samples; and

applying a set of uniform equations, each uniform equation corresponding to a bit position of a bit soft value to be calculated and producing a solution as accurate as an available region equation selected from a set of region equations associated with a nominal modulation constellation for the bit position;

determining a bit soft value for each symbol sample based on the set of unified equations based on calculating the bit soft value for each bit position by solving the corresponding unified equation using the real or imaginary part of the bit position dependent symbol sample.

76. The mobile terminal of claim 75, wherein the set of uniform equations for 16QAM includes a first uniform equation for computing bit soft values corresponding to the sign bits of the symbol samples, wherein the first uniform equation generates a first term based on the magnitude of the sum of the real or imaginary part of the symbol samples and the nominal symbol interval, generates a second term based on the magnitude of the difference of the real or imaginary part of the symbol samples and the nominal symbol interval, generates a third term based on four times the real or imaginary part of the symbol samples and subtracts the second and third terms from the first term.

77. The mobile terminal of claim 76, wherein the set of uniform equations for 16QAM includes a second uniform equation for calculating soft bit values for bit positions used to represent symbol samples located in inner or outer rows of the nominal modulation constellation or symbol samples located in inner or outer columns of the nominal modulation constellation, and wherein the second uniform equation produces a first term based on twice a real or imaginary amplitude of the symbol samples, produces a second term based on twice a nominal symbol interval, and subtracts the second term from the first term.

78. The mobile terminal of claim 75, wherein the set of uniform equations for 16QAM includes a first uniform equation for calculating soft bit values for bit positions corresponding to signs of a real part and an imaginary part, and a second uniform equation for calculating soft bit values for bit positions indicating that the symbol samples are located in an inner or outer column of a nominal modulation constellation and that the bit positions indicate that the symbol samples are located in an inner or outer row of the nominal modulation constellation.

79. The mobile terminal of claim 75, wherein the unified equation comprises a unified equation of the form,

wherein λ_1...4' corresponds to four bit positions, r, associated with 16QAM_xAnd r_yRespectively the real and imaginary parts of the symbol samples, and 2a is the interval of the nominal modulation constellation.

80. The mobile terminal of claim 79, wherein the processing circuit is configured to use for λ₁' to a unified equation for representing symbol samplesCalculating bit soft values using bit positions of the symbols of the real part for lambda₂The unified equation of' calculates the soft bit values for the bit positions of the symbols representing the imaginary part of the symbol samples, using the values for λ₃The unified equation of' calculates bit soft values for bit positions representing symbol samples located in inner or outer columns of a nominal modulation constellation, and uses the values for λ₄The unified equation of' is to compute bit soft values for bit positions that represent symbol samples located in inner or outer rows of a nominal modulation constellation.

81. The mobile terminal of claim 75, wherein the set of unified equations for 64QAM includes a unified equation of the form,

wherein λ_1...6' corresponds to six bit positions, r, associated with 64QAM_xAnd r_yAre respectivelyThe real and imaginary parts of the symbol samples, and 2a is the interval of the nominal modulation constellation.

82. The mobile terminal of claim 75, wherein the processing circuitry is configured to apply, for each symbol sample, one or more correction terms for each unified equation that compensate the bit soft values for one or more additional nearest neighbors of the symbol sample in the nominal modulation constellation.

83. The mobile terminal of claim 75, wherein the processing circuit is configured to compensate the unified equation using one or more multi-region compensation terms that compensate the bit soft values calculated from the unified equation for effects of constellation points located in two or more regions of the nominal modulation constellation.

84. The mobile terminal of claim 75, wherein the processing circuit is configured to scale the nominal modulation constellation to obtain a scaled nominal modulation constellation.

85. The mobile terminal of claim 84, wherein, for 16QAM, the unified set of equations respectively includes the following first and second unified equations for calculating the bit soft values for the ith bit position:

λ_i′＝|ar_{x or y}+2a²|-|ar_{x or y}-2a²|-4ar_{x or y}and

λ_i′＝2|ar_{x or y}|-4a²

where 2a comprises the symbol interval of the scaled nominal modulation constellation, r_xIncluding the real part of the symbol sample, and r_yIncluding the imaginary part of the symbol samples.

86. The mobile terminal of claim 85 wherein the processing circuit is configured to combine the real parts of the symbol samples r using a first uniform equation_xTo calculate soft bit values for bit positions representing the real part symbols of a symbol sampleCombining the imaginary part r of the symbol samples using a first uniform equation_yTo calculate soft bit values for bit positions representing imaginary symbols of the symbol samples, the real part r of the symbol samples is combined using a second uniform equation_xTo calculate soft bit values for bit positions representing symbol samples located in inner or outer columns of a nominal modulation constellation and to use a second uniform equation in combination with the imaginary part r of the symbol samples_yTo compute bit soft values for bit positions representing symbol samples located in inner or outer rows of a nominal modulation constellation.

87. The mobile terminal of claim 85 wherein the processing circuit is configured to select a particular one of the two unified equations and r_xAnd r_yOne of the components calculates a bit soft value for the ith bit position based on the particular gray coded mapping employed by the nominal modulation constellation.

88. The mobile terminal of claim 75, wherein the processing circuitry is configured to scale the nominal modulation constellation to obtain a scaled nominal modulation constellation, and for 64QAM, the unified equation set comprises the following three equations used to calculate the ith bit position:

λ_i′＝-8ar_{x or y}+|ar_{x or y}+2a²|-|ar_{x or y}-2a²|+|ar_{x or y}+4a²|-|ar_{x or y}-4a²|+|ar_{x or y}+6a²|-|ar_{x or y}-6a²|，

λ_i′＝-16a²+4|ar_{x or y}|-||ar_{x or y}|-2a²|+||ar_{x or y}|-6a²|，

λ_i′＝-4a²+|2|ar_{x or y}|-8a²|，

where 2a comprises the interval of the scaled nominal modulation constellation, r_xIncluding the real part of the symbol sample, and r_yIncluding the imaginary part of the symbol samples.

89. The mobile terminal of claim 88 wherein the processing circuit is configured to select a particular one of the three unified equations and r_xAnd r_yOne of the components calculates a bit soft value for the ith bit position based on the particular gray coded mapping employed by the nominal modulation constellation.