18th LSI Design Contests�Ein Okinawa  Design Specification - 4-5

4-5. Output example of the value

When finding the cosƒÆ and sinƒÆ using the CORDIC method, an example \(cos45�‹\) and \(sin45�‹\), to see how they approached.

First, in order to ensure that the value is approaching shows true value of \(cos45�‹\) and \(sin45�‹\)�D This is represented by a fixed-point and binary floating point decimal.

Also, a result of output in a C program for output of these values, the verification environment is Microsoft VisualStudio2010.

Table1�FTrue value of \(sin45�‹\) and \(cos45�‹\)

Graph 1

Next, Table 2 shows the values of the \(cos45�‹\) and \(sin45�‹\) as determined by the CORDIC method�D

Table2�F\(sin45�‹\) and \(cos45�‹\) by CORDIC method

Graph 1

From Table2, It is seen that approaches the true value of Table 1 by increasing the number of loops�D At this time, Are summarized in Table 3 which took the difference between the calculated value of the CORDIC method and the true value for each loop.

Table3�FThe difference between the calculated value of the CORDIC method and the true value

Graph 1

In Table 3, As you look at the things that were expressed in fixed-point, it shows the accuracy of the CORDIC method obtained by subtraction�D At the location where it was displayed in red. Some errors exist, but the accuracy of each loop is almost obtained�D However, the results obtained, since it depends on the execution environment, the result is only an example�D

I is shown in Table 4 that represents how an operation is whether advances in CORDIC.


Table4�FCalculation process by the CORDIC method of \(ƒÆ\) = 45 �‹

Graph 4

I can be seen that the calculation is performed repeatedly on the basis of the determination of \(z\) from Table 4, is approaching gradually to the angle \(Į\). The z determination at this time is determined relative to the z obtained in the calculation of the previous one.

<<Back                 Next>>