Next: 2 Target
architectures Up: 3
Implementation on embedded Previous: 3
Implementation on embedded
The whole algorithm has been implemented in C language, using double precision floating point. The profiling analysis of the algorithm shows that the 71.4% of the execution time is spent in float point operations. The 21.23% is used for integers operations and the final 7.37% for remainder operations, including trigonometric ones. The profile suggests the use of a float point unit (FPU) for improving the execution time.