Digital pipelined heterodyne circuit.

BACKGROUND OF THE INVENTION

0001The present invention relates to a digital pipelined heterodyne circuit for heterodyning a data input signal by a signal of a predetermined frequency.

0002Heterodyning is a technique that has applica­tion in many areas of communiations and signal process­ing. For example, in spectrum analysis, it is often necessary to shift the center frequency of a broad-band signal into a region where a narrow-band filter may be located which is used to sample input data. In order to accomplish this function, digital heterodyne cir­cuits are available which digitize a data input signal and multiply that signal by the sine and cosine com­ponents of a user-chosen reference signal having a pre­determined frequency. The problem with such circuits is that each data byte must be digitally multiplied by digital sine and cosine coefficients one at a time. Digital multiplication is a two-step process involving the combining of one bit of the sine and cosine coef­ficient with the data byte in an AND-gate array and then adding the product to the previous product formed by the previous gate array. It will take as many steps to completely multiply the data byte as there are bits in the sine and cosine coefficients. For example, if there are m significant bits, each data input byte must be processed, one bit at a time, m times, before the next byte can be processed. This leads to unacceptably slow processing time, especially if high resolution is required.

SUMMARY OF THE INVENTION

0003The present invention provides a digital heterodyne circuit which includes a data input line for providing a sequence of digitized data signals and digital sine and cosine function generators for generating m-bit digital coefficients as a function of time at a predetermined frequency. An m-stage digital multiplier is provided for multiplying the digitized data signals by the coefficients generated by the sine and cosine function generators. Time delay circuits are connected between the function generators and the m-stage multiplier for providing each bit of each digi­tal coefficient to each respective stage of the multi­plier means simultaneously with the arrival at each such stage of a predetermined digitized data signal.

0004This circuit provides a pipelined heterodyne circuit in which data input bytes are sequentially pro­cessed such that each time a data byte enters a par­ticular stage for partial multiplication, the previous data byte is being processed downstream in the next stage. As each data byte proceeds sequentially from one stage to the next, a unique shift register makes available an additional bit of digital sine and cosine coefficients that were generated when the byte first entered the data input. This is made possible by a unique triangular shift register in which the number of stages of the shift register connecting each bit output of the sine and cosine function generators with each respective stage of the multiplier includes the same number of stages as the rank in significance of the particular bit. Thus, the most significant bit output of the sine and cosine function generators include a one-stage shift register. The second most significant bit includes a two-stage shift register, and so on, up to an m-stage shift register connecting the mth signi­ficant bit with the m^th multiplication stage. This circuit provides an increase in speed over conventional digital heterodyne circuits by a factor of m where m is the number of significant bits produced by the sine and cosine function generators. The output is filtered in a low-pass decimation filter.

0005It is a principal object of this invention to provide a digital pipelined heterodyne circuit for increasing the speed of digitally heterodyning an input data signal with a signal of a given frequency.

BRIEF DESCRIPTION OF THE DRAWINGS

0006<ul id="ul0001" list-style="none"><li>FIG. l is a block schematic diagram of the digital heterodyne circuit of the present invention.</li><li>FIG. 2 is a block schematic diagram of multiplier stage number l shown in FIG. l.</li><li>FIG. 3 is a block schematic diagram of a circuit utilized in multiplier stages 2 through m in FIG. l.</li></ul>

DETAILED DESCRIPTION OF THE INVENTION

0007A digital heterodyne circuit l0 includes a cosine coefficient generator l2 and a sine coefficient generator l4. Cosine coefficient generator l2 is con­nected to a triangular shift register l6, and sine coefficient generator l4 is connected to a similar triangular shift register l8. Each of the triangular shift registers l6 and l8 comprise a number of individ­ual single-stage or multi-stage shift registers for each bit output of respective cosine and sine coef­ficient generators l2 and l4. For example, the most significant bit output of cosine coefficient generator l2 is connected to a single-stage shift register 20, while the next two significant bits of cosine coef­ficient generator l2 are connected to two-stage shift register 22 and three-stage shift register 23. The least significant bit of the cosine coefficient genera­tor l2 is connected to an m stage shift register 24. The connections between the sine coefficient generator l4 and triangular shift register l8 are the same as those described with respect to cosine coefficient generator l2. The most significant bit of sine coef­ficient generator l4 is connected to single-stage shift register 26. The next most significant bit of sine coefficient generator l4 is connected to two-stage shift register 28, as is the next most significant bit connected to three-stage shift register 30. The least significant bit of sine coefficient generator l4 is connected to m-stage shift register 32.

0008A digitized data signal Di enters on data input line 34 where it is multiplied in multiplier l by the output of the most significant bit shift registers 20 and 26 respectively. If the data signal at multi­plier l is arbitrarily designated as the current data input, it can be seen that the current digital cosine and sine coefficients, which are labelled with the letter "c", are loaded into the triangular shift regis­ters l6 and l8, respectively. This happens each time a new data input signal Di is provided to multiplier l. A system clock (not shown) clocks a data byte into multiplier l and simultaneously enables the cosine and sine function generators l2 and l4 to generate a new m-bit coefficient to be loaded into shift registers l6 and l8, respectively. Thus, at predetermined time intervals, the cosine and sine coefficient generators l2 and l4 load a digital coefficient representing the sine function and cosine functions respectively of an angle ωt which varies with time where ω is a predeter­mined frequency chosen by the user.

0009Downstream from multiplier l, in multiplier 2, the previous data input signal is simultaneously processed by the previous cosine and sine coefficients labelled c-l, and in the m^th multiplier stage, multi­plier m, the m^th previous data input signal is being processed by the c-m cosine and sine coefficients, respectively. Thus, as a data input signal enters the "pipeline" at multiplier l, it will be partially multiplied in each of the multipliers by a bit gener­ated by cosine coefficient generator l2 and sine coef­ficient generator l4 of sine and cosine coefficients respectively that were generated at the particular time that the data signal was clocked into multiplier l. Simultaneously, other data signals are multiplied by different bits of different sine and cosine coeffi­cients that were generated at earlier times. In this way, as a data input signal progresses through each of the multiplier stages, being partially multiplied each time by one bit of the same sine and cosine coeffi­cients, other data signals are being simultaneously processed in other multiplier stages by partial multiplication using the appropriately generated bits of the sine and cosine coefficients. This speeds up the multiplication process by a factor of m.

0010Referring to FIG. 2, multiplier l is a digi­tal multipler includes an input register 36 which is connected to data input line 34. The input register 36 is connected to a 2's complement generator 38 which simultaneously provides the 2's complement of each data byte stored in the input register 36 to AND gate arrays 40 and 42, respectively. The n bits of the output of 2's complement generator 38 are provided to a set of parallel AND gates which are all combined with the most significant bit of the sine and cosine coefficients, respectively. The 2's complement of each data input signal is used for this purpose in order to properly maintain the word length of the data input signals which may change in polarity from plus to minus. The outputs of AND gate arrays 40 and 42, respectively, are provided to adders 44 and 46, respectively, which add the input data signal multiplied by the most signifi­cant bit of the cosine and sine coefficients, respec­tively to an n-bit ground signal. The sum of these two numbers is stored in storage registers 48 and 50, respectively. These signals, along with the data input signal, are provided to the next multiplier stage.

0011The following multiplier stages are shown in FIG. 3. For each such stage, an input line from the stored output of the previous stage is provided to adders 52 and 54, respectively. The k^th most signifi­cant bits of the cosine and sine coefficient generators l2 and l4 are provided to AND gate arrays 56 and 58, respectively. This bit is combined with the input data signal held in input register 60. The output of AND gate arrays 56 and 58 are added to the stored output of the previous stage in adders 53 and 54, and the results are stored in storage registers 62 and 64, respectively.

0012Referring again to FIG. l, the last multi­plier stage, labelled multiplier M in FIG. l, passes the outputs of its storage registers, which are now the product of digitized input signal D_i and a sine and cosine coefficient labelled S_i and C_i, respectively, to a decimation low-pass filter 66 for the purpose of removing transient noise.

0013The terms and expressions which have been employed in the foregoing specification are used therein as terms of descrip tion and not of limitation, and there is no intention, in the use of such terms and expressions, of excluding equivalents of the features shown and described or portions thereof, it being recog­nized that the scope of the invention is defined and limited only by the claims which follow.