Efficient circular clipping implementation using simplified CORDIC rotators

This application is related to U.S. application Ser. No. 10/608,489, filed Jun. 26, 2003, now U.S. Pat. No. 7,313,196, issued Dec. 25, 2007, which is incorporated by reference herein in its entirety.

1. Field of the Invention

The present invention relates to CORDIC rotator systems and, more particularly, to simplified CORDIC rotators in CDMA communication systems and other hardware implementations.

2. Prior Art

Many algorithms designed for communications, navigation, robotics, and digital signal processing generally do not translate efficiently to a hardware implementation and must rely upon software to execute.

In one example, many digital communications systems utilize amplitude and phase modulation, producing complex baseband signals, which are pulse shaped before modulation to the carrier frequency. The complex, baseband signals are typically represented as an in-phase signal (I) and a quadrature signal (Q). For many modulation formats and multiple access schemes, these complex baseband signals have large peak values relative to the average signal value. When the peak-to-average power ratio of the baseband signal is large, the signal requires reduction in the power amplifier to avoid saturation. The peak-to-average power ratio limits the average power that can be transmitted due to the finite maximum output power of the power amplifier in the RF front-end (RFFE). Alternatively, for a fixed average transmit power, the peak-to-average power ratio will determine the required maximum output power of the RF power amplifier. In both cases, a decreased peak to average power ratio is desirable.

To decrease the peak-to-average power ratio of the signal, many digital communications systems employ clipping, also termed amplitude limiting, on the baseband complex signal. Clipping suppresses the signal peaks, reducing the peak-to-average power ratio of the signal with minimal signal distortion. The most common and effective form of clipping is circular clipping, which preserves the angle of the complex signal while limiting the maximum magnitude. Circular clipping follows the input-output relation,

I

out

=

In hardware implementations, circular clipping is often implemented using a look-up table (LUT). A LUT implementation avoids the need to compute the constellation magnitude, implement the square root, and division functions as are done in U.S. Pat. No. 6,266,320 to Hedberg et al.; these functions, as well as others, are complex operations for digital hardware. With a LUT-implementation, the LUT holds the output I and Q (I out ,Q out ) values for all possible combinations of input values (I in , Q in ). When the number of possible input I and Q values is small, the LUT approach is attractive. However, as the number of possible I and Q input values increases the size of the memory used to store the LUT must grow proportionately. With 10-bit input (I,Q) values and 8-bit (I,Q) output values, the LUT would require a memory of size 2,097,152 bytes. If the input constellation is symmetric in the four quadrants, the LUT size may be reduced by a factor of four, but still requires a large memory of size 524,288 bytes. In many applications a LUT-based circular clipper may be impractical when the number of possible input I and Q values is large.

In still other hardware implementations to solve complex algorithms is the use of CORDIC rotators. Yet, even CORDIC rotators require the use of LUTs.

Therefore, it is desirable to provide a hardware efficient method and system to solve complex algorithms.

The foregoing and other problems are overcome, and other advantages are realized, in accordance with the presently preferred embodiments of these teachings.

In accordance with one embodiment of the present invention a system for the efficient use of CORDIC (Coordinate Rotation DIgital Computer) rotators is provided. The system includes at least one first rotator operating in vector mode and includes a d i storage buffer for the angle direction vector. The system also includes at least one bit inverter coupled between the first rotator and a second rotator, the second rotator operating in rotation mode.

In accordance with another embodiment of the present invention a system for baseband amplitude limiting is provided. The system includes an improved CORDIC vector mode rotator device having a d i storage buffer in addition to a gain device and limiter. The system also includes a second CORDIC rotator operating in rotation mode, and is coupled to the limiter.

In accordance with another embodiment a method for limiting the amplitude of complex signals is provided. The method includes the steps of rotating and limiting a voltage vector magnitude. The next steps generate a signal based upon the limited and rotated voltage vector magnitude. The signal includes at least x-component and at least one y-component.

The invention is also directed towards a method for limiting the peak-to-average power ratio of a plurality of complex telecommunications signals. The method includes the steps of summing I 0 . . . I n and Q 0 . . . Q n signals to produce an

I i ⁢ ⁢ n = ∑ k = 0 n ⁢ I k ⁢ ⁢ and ⁢ ⁢ Q i ⁢ ⁢ n = ∑ k = 0 n ⁢ Q k

composite signal, respectively, and where n is predetermined, e.g., number of channels. The next steps determine a peak power vector and an average power level. The peak power vector is compared to the average power level and adjusted according to the comparison.

In accordance with another embodiment of the invention a program storage device readable by a machine is provided. The program storage device tangibly embodies a program of instructions executable by the machine to perform method steps for limiting the amplitude of complex signals. The method includes the steps of rotating and limiting a voltage vector magnitude and generating a signal based upon the limited first voltage vector magnitude. The signal includes at least one x-component (in-phase signal I) and at least one y-component (quadrature phase signal Q).

In accordance with another embodiment the invention is also directed towards a method for efficiently limiting a vector magnitude. The method includes the step of providing a first vector having magnitude and direction. The next step iteratively rotates the vector through diminishing angles to a reference angle such that the vector angle is substantially zero and each angle direction is stored in a direction vector. Once the reference axis is reached the next step limits the vector magnitude to a predetermined magnitude to form a second vector. The last step rotates the second vector through a second angle according to each angle direction stored in the first direction vector.

The foregoing aspects and other features of the present invention are explained in the following description, taken in connection with the accompanying drawings, wherein:

FIG. 1 is a block diagram of a communications system implementing baseband amplitude limiting features of the present invention;

FIG. 2 is a block diagram of a preferred embodiment of a baseband amplitude-limiting feature shown in FIG. 1 ;

FIGS. 2A-2C are axis diagrams showing phase and magnitude of a vector relative to an average value according to FIG. 2 ; and

FIG. 3 is a flow chart of one method for implementing the baseband limiting features of the present invention shown in FIG. 2 .

Referring to FIG. 1 , there is shown a pictorial diagram of a multi-user communications system incorporating features of the present invention. Although the present invention will be described with reference to the embodiment shown in the drawings, i.e., a method and system to decrease a peak-to-average power ratio of a communications signal before the signal is amplified by a power-amplifier, it should be understood that the present invention could be embodied in many alternate forms of embodiments. For example, the invention may be embodied in any suitable device requiring the solution of complex algorithms such as navigation systems, radar systems, robotics, computers, or handheld calculators.

Still referring to FIG. 1 there is shown a Fixed Wireless System (FWS) 10 that is suitable for practicing this invention. Specifically, the FWS 10 employs direct sequence spread spectrum based CDMA techniques over an air link to provide local access to subscribers, and offers very high quality, highly reliable service. The FWS 10 is a synchronous CDMA (S-CDMA) communications system wherein forward link (FL) transmissions from a base station, referred to also as a radio base unit (RBU) 12 , for a plurality of transceiver units, referred to herein as user or subscriber units (SUs) 14 , are symbol and chip aligned in time, and wherein the SUs 14 operate to receive the FL transmissions and to synchronize to one of the transmissions. Each SU 14 also transmits a signal on a reverse link (RL) to RBU 12 in order to synchronize the timing of its transmissions to the RBU 12 , and to generally perform bi-directional communications. The FWS 10 is suitable for use in implementing a telecommunications system that conveys multirate voice and/or data between the RBU 12 and the SUs 14 .

The RBU 12 includes circuitry for generating a plurality of user signals (USER 1 to USER n ), which are not shown in FIG. 1 , and a synchronous side channel (SIDE chan ) signal that is continuously transmitted. Each of these signals is assigned a respective PN spreading code and is modulated therewith before being applied to a transmitter 12 a having an antenna 12 b . When transmitted on the FL the transmissions are modulated in phase quadrature, and the SUs 14 are assumed to include suitable phase demodulators for deriving in-phase (I) and quadrature (Q) components there from. The RBU 12 is capable of transmitting a plurality of frequency channels. By example, each frequency channel includes up to 128 code channels, and has a center frequency in the range of 2 GHz to 4 GHz.

The RBU 12 also includes a receiver 12 c having an output coupled to a side channel receiver 12 d . The side channel receiver 12 d receives as inputs the spread signal from the receiver 12 c , a scale factor signal, and a side channel despread PN code. These latter two signals are sourced from a RBU processor or controller 12 e . The scale factor signal can be fixed, or can be made adaptive as a function of the number of SUs 14 that are transmitting on the reverse channel. The side channel receiver 12 d outputs a detect/not detect signal to the RBU controller 12 e for indicating a detection of a transmission from one of the SUs 14 , and also outputs a power estimate value. A read/write memory (MEM) 12 f is bi-directionally coupled to the RBU controller 12 e for storing system parameters and other information, such as SU timing phase information and power estimate values.

A Network Interface Unit (NIU) 13 connects the RBU 12 to the public network, such as the public switched telephone network (PSTN) or the Internet 13 a , through analog or digital trunks that are suitable for use with the local public network. The RBU 12 connects to the NIU 13 using E1 trunks and to its master antenna 12 b using a coaxial cable. The SU 14 communicates with the RBU 12 via the radio interface, as described above.

In the illustrated embodiment the SU-RBU air link provides a separate 2.72 MHz (3.5 MHz including guard bands) channel in each direction separated by 100 MHz, 119 MHz, or 175 MHz. In alternate embodiments any suitable separation may be employed. The nominal spectrum of operation is 2.1-2.3 GHz or 2.5-2.7 GHz. However, the system is adaptable such that the frequency can be varied as required. In addition, before the I and Q signals are applied to transmitter 12 a the Peak-to-Average power ratio of the signals are clipped to a desirable level by clipper 20 in accordance with the teachings of this invention. In general, circular clipping operates on the input I in and Q in values to compute output I out and Q out values such that the peak signal magnitude of the output is Pp out ≦Pp in . Circular clipping preserves the angle of the input I and Q pair but limits peak signal magnitude.

Referring to FIG. 2 there is shown a block diagram of a preferred embodiment of the present invention. The COordinate Rotation DIgital Computer (CORDIC) rotator 21 is known in the art and need not be discussed here other than to note that CORDIC rotators are normally operated in one of two possible modes. The rotation mode rotates the input vector by a specified angle. The other mode is the vectoring mode, which rotates the input vector to an arbitrary axis, nominally the x-axis while recording the angle required to make that rotation. Rotation mode essentially performs a change of coordinates from Cartesian coordinates (I and Q) to polar coordinates (magnitude and phase).

Referring now to FIG. 2 there is shown a block diagram of a circuit incorporating features of the present invention. In will be appreciated that in a preferred embodiment at least two CORDIC rotators 21 , 22 are used. In an alternate embodiment any suitable rotator or CORDIC-like device may be used. Also, in a preferred embodiment CORDIC rotator 21 is operated in vector mode while CODIC rotator 22 is operated in rotation mode.

Referring still to FIG. 2 , CORDIC 21 operates in a preferred vectoring mode with inputs x 0 =I in , and y 0 =Q in . In a preferred embodiment the I in and Q in signals are 8-bits and the I out and Q out signals are 10-bits. In alternate embodiments any suitable bit length could be used. The x and y values are recursively updated using the following set of equations where the CORDIC rotator 21 computes

x

i

+

1

=

=

=

where i is the iteration number.

For implementation, x, y and d i 1 storage buffers are required. In a preferred embodiment the number of iterations is ten. However, it will be recognized that any number of suitable number of iterations may be used to achieve the desired accuracy.

In vectoring mode with I iterations, the final outputs of CORDIC 21 are

x I =1.647·√{square root over ( x 0 2 y 0 2 )}= A

y I =0

Here, A is the magnitude of the complex input signal (I in +jQ in ) while θ is the angle in radians. As can be seen, CORDIC 21 operating in vectoring mode converts the input signal from Cartesian to a vector magnitude with a small scaling on the magnitude.

In the clipping circuit 20 , CORDIC 22 operates in rotation mode using similar recursive update equations and initialization 2x 0 =A′ and 2y 0 =0. CORDIC 22 in the clipper block 20 restore the angle to the scaled and clipped amplitude. As shown in FIGS. 2A-2C , CORDIC 22 operating in rotation mode converts the clipped signal from polar back to Cartesian coordinates.

In clipping circuit 20 , d i 1 =−d i 2 , where d i 1 are values used in CORDIC rotator 21 and where d i 2 are the values used in CORDIC rotator 22 . Since the d i 1 values computed in CORDIC rotator 21 are the negative of the d i 2 values required in CORDIC rotator 22 , the d i 1 values are stored for subsequent use in CORDIC rotator 22 . Sign inverter 28 reverses the sign of d i 1 and couples the output to CORDIC rotator 22 . It will be appreciated that knowledge of the d i 2 values eliminates the need for z accumulator update hardware and the corresponding arctan(2 −i ) look-up-table (ATAN-LUT) in both CORDICs 21 , 22 ; advantageously reducing hardware complexity by approximately 25 percent per CORDIC. CORDIC rotator 22 then iterates according to:

x i+1 =x i −y i d i 2 2 −i

y i+1 =y i −x i d i 2 2 −i

d i 2 =−1· d i 1

Notice that in the circular clipping circuit 20 , neither CORDIC requires a z accumulator or a corresponding ATAN-LUT. It should further be noted that in FIG. 2 , the d variable is a direction vector with d=[d 0 1 , d 1 1 , . . . , d I 1 ].

Referring also to FIG. 3 , there is shown a method flow chart illustrating the steps of one method for implementing the invention shown in FIG. 2 . First, a vector quantity is derived from I in and Q in signal values and an iteration counter is set to 1, step 41 . In alternate embodiments the I in and Q in may be summed composites of multiple I 0 . . . I n and Q 0 . . . Q n signals, respectively. In alternate embodiments summers 26 , 27 may be any suitable device for combining multiple I and Q signals. A sign indicator is derived from the I in and Q in signals and stored, step 42 , in vector d (not shown), within CORDIC rotator 21 . Step 43 increments the iteration counter and a new sign (d i ) is again calculated and stored, step 42 , Note, that upon exiting step 45 , an appropriate number of iterations result in the vector magnitude laying substantially in the x-plane. The vector magnitude A is then scaled, step 46 , by scaler G and hard-limited if necessary, step 47 , to a predetermined maximum 8-bit amplitude by limiter L as shown in FIGS. 2A-2C . Scaler 23 allows input signals that do not exceed the rails of the hard-limiter L to be scaled and may be varied as number of users varies to maintain constant average power output. The scaled and clipped signal polar coordinate signal, along with direction vector d, is coupled to CORDIC 22 . CORDIC 22 , operating in rotation mode effectively rotates the clipped vector through an angle according to the inverse of the direction vector d (e.g. −1d) and produces 8-bit I out and Q out components, steps 48 - 51 . Knowledge of the input peak-to-average power ratio (PAR in ) and component specifications allows the scaler 23 and hard-limiter L value to be chosen to such that a desired output peak-to-average power ratio (PAR out ≦PAR in ) is achieved.

It is appreciated therefore that an efficient baseband circular clipping method and system to decrease a peak-to-average power ratio of a telecommunications signal before the signal is amplified by a power amplifier has been provided. Advantageously, the method and system provided avoids the need for an accumulator and ATAN look-up-tables. The method and system advantageously also offers significant reductions in the number of gates needed to implement the circular clipping relative to the LUT approaches of the prior art when the number of possible input combinations is large. The method and system also advantageously provides efficient baseband circular clipping without mathematically intensive square root, and division functions.

It should be understood that the foregoing description is only illustrative of the invention. Various alternatives and modifications can be devised by those skilled in the art without departing from the invention. Exemplary alternatives include, but are not limited to, employing the improved CORDIC devices in trigonometric calculators, fast Fourier transformers (FFTs), and Digital Fourier Transforms (DFTs). Accordingly, the present invention is intended to embrace all such alternatives, modifications and variances that fall within the scope of the appended claims.