In this lab you are to implement and optimize the a pseudo-noise (PN) sequence generator, IIR filter, and autocorrelation routines that are part of the previous lab's PSD estimator. For the lab grade, you will be judged on the execution time of your system (memory usage need not be minimized).

Reference Implementation After taking a look at the source code of the PSD estimator reference implementation, you will likely discover inefficiencies. This implementation is provided as the "reference implementation" of the optimization process and to define the expected input and output of the application. The computational efficiency of your code will be judged against this implementation. While the given code might serve as a starting point, you should do whatever you need to do to make your code as efficient as possible, while operating in an equivalent manner as the given code. The exact portion of the code to be optimized is defined below. You may write in C, assembly, or any combination of the two; choose whatever will allow you to write the fastest code. The optimization process will be smoother if you plan for optimization before you begin any programming.

Optimization Since a primary purpose of this lab is to learn optimization and efficient code techniques, your lab grade will be based primarily on the total execution time of your system. You are not required to optimize memory use. Note that by execution time we mean cycle count, not the number of instructions in your program. Remember that some of the TMS320C55xx instructions take more than one cycle. However, unlike the TMS320C54xx instructions, most operations take only one cycle and can be placed in parallel with other operations. Branch and repeat statements are the most common instructions that require several cycles to execute. Most C instructions take more than one cycle. The debugger can be used to determine the exact number of cycles used by your code. The instructions on how to do this can be found in Cycle Counts. We will grade you based on the number of cycles used between the rand_fillbuffer(); and cfft((DATA *)fft_data,N, SCALE); statements. Thus, you can optimize rand_fillbuffer function but optimizing the fft function will not help. Note that some instructions, like RPT, are non-repeatable instructions; their use may cause unnecessary glitches in I/O. For grading simplicity, your final code should not have modifications except between these two instructions, and M should be set to 31. If the number of cycles between the two points is variable, the maximum possible number of cycles will be counted. You must use the dma.c and swi_process.h files in v:\ece420\55x\lab4 as provided by the TAs; these files may not be modified. We reserve the right to test your code by modifying the inputs.

Routine-Specific Optimization Tips If you are programming the PN generator in assembly, you may wish to refer to the description of assembly instructions for logical operations in the C55x Mnemonic Instruction Set reference. Initialize the shift register to one. You can debug the PN output by comparing it to the output of the MATLAB code. Be prepared to prove to a TA that your PN generator works properly as part of your quiz. Your IIR filtering routine can debugged by writing an impulse followed by zeros in autocorr_in instead of randsample. Your autocorrelation routine can be debugged by commenting out the IIR-filtering routine and writing the maximum DC value into autocorr_in in a similar manner as described the IIR-debugging step. Note that each of these tips is the most helpful if the output is inspected in memory.

Grading Grading for this lab will be a bit different from past labs: 2 points: Working code, implemented from scratch in assembly language or C. 5 points: Optimization. These points will be assigned based on your cycle counts and the optimizations you have made. 3 points: Oral quiz.