臺灣博碩士論文加值系統

本文所撰述為適用於通訊系統之訊號處理器核心完整研究。整體研究包含處理器核心之參數化架構、多重功能之數據處理路徑、內嵌式系統之輸出入介面、低功率設計、可供選擇之特殊功能區塊、模組化訊號處理器核心產生器。
參數化數位訊號處理器核心是設計使核心針對不同應用需求有最佳化之效能。歸功於參數化架構，此數位訊號處理器核心可由使用者自行選定參數以適用所預定之系統。本論文所設定之適用系統為通訊系統。
本論文成功的建立完整設計流程，使數位訊號處理器核心有更彈性的設計特質以及更好的效能表現。

This thesis is a comprehensive work of a parameterized DSP core for embedded system. The overall research includes the parameterized architecture of the DSP core, multi-function data-path, input/output modes for embedded applications, low-power techniques, optional special function blocks and module generator.
Parameterized DSP core is designed for optimal usage of different requirements in system applications. Owing to the parameterized structure, the DSP core can be custom made for dedicated system with parameters setting. The parameterized DSP core is especially suitable for an embedded system. The primary application is digital signal processing system which designed to achieve demodulation/synchronization with better performance and flexibility.
We set up a novel design flow of embedded parameterized DSP core. The design flow includes a newly designed flexible parameterized DSP core structure, a module generator and a methodology to build the DSP core with optimal performance. The features in this DSP core include parameterized data-path, dual MAC unit, sub-word MAC and optional function-specific blocks for accelerating communication system modulation operations. This DSP core also has a low power structure, which includes the gray code addressing mode, pipeline sharing and a novel buffered hardware looping. Users can select the parameters and special function blocks based on the specification of their applications and then a synthesizable DSP core is generated in a proposed module generator with graph user interface.
The proposed DSP core and the synthesizable RTL code have been verified with TSMC 0.35um SPQM, 0.25um 1P5M cell based and FPGA design flow.

Table of Contents
Abstract…………………………………………………….IV
Acknowledgements…………………………………….…V
Table of Contents…………………………………….….VII
List of Figures……………………………………………..X
List of Tables……………………………………………XIV
List of Abbreviations…………………………………….XVI
1. Introduction …………………………………………1
1.1. Motivation and related works …………………………1
1.2. Outline of Thesis ……………………………………5
2. DSP Architecture ……………………………………6
2.1. Overview of DSP architecture ……………………6
2.2. Program Address Generation Unit (PAGU) …………9
2.2.1. Hardware looping ………………………………10
2.2.2. Branch, call and return …………………………13
2.3. Data address generation unit(DAGU) ………………15
2.4. Data path ………………………………………………18
2.4.1. Status register ……………………………………18
2.4.2. Arithmetic logic unit(ALU) ……………………19
2.4.3. MAC unit ……………………………………22
2.4.4. Barrel shifter ……………………………………24
2.5. I/O ……………………………………………………25
2.5.1. Host port interface ………………………………25
2.5.2. Handshaking mode……………………………… 27
2.5.3. DMA mode ……………………………………28
2.5.4. Merge mode ……………………………………30
2.6. Pipeline Sharing ……………………………………30
2.7. Summary ………………………………………………32
3. Parameterized Design Flow …………………………33
3.1. Introduction …………………………………………33
3.2. Parameters …………………………………………33
3.2.1. Data_length, Dmem_size, Pmem_size, HPImem_size, PC_stack_size ……………………………………35
3.2.2. A_width G_width ………………………………36
3.2.3. R_num, …………………………………………36
3.2.4. IO_width …………………………………………36
3.2.5. AR_num …………………………………………36
3.2.6. Loop_num ……………………………………37
3.2.7. Buffer_mem ……………………………………37
3.3. Parameterized design ………………………..…….37
3.3.1. PAGU ……………………………………..….37
3.3.2. DAGU …………………………………………38
3.3.3. Memory architecture …………………………39
3.3.4. Data path …………………………………………40
3.3.5. I/O ………………………………………………41
3.4. Summary ………………………………………………41
4. Special Function Blocks ………………………………43
4.1. Introduction …………………………………………43
4.2. Optional special function blocks ……………………44
4.2.1. Slicer …………………………………………44
4.2.2. Hamming distance calculator ……………………45
4.2.3. Dedicated FIR block …………………………46
4.2.4. Buffered hardware nested looping ………………47
4.2.4.1. Introduction ……………………………………47
4.2.4.2. Implementation ………………………………48
4.2.4.3. Hardware looping ………………………………48
4.2.4.4. Buffered hardware looping ……………………51
4.2.5. Streamed I/O ……………………………………53
4.2.5.1. Introduction ……………………………………53
4.2.5.2. Accumulator I/O ………………………………55
4.2.5.3. Memory connection I/O …………………………55
4.3. Optional multi-function blocks ……………………57
4.3.1. MAC options ……………………………………57
4.3.2. Indirection addressing modes ……………………65
4.4. Design results of optional special function blocks ……65
4.4.1. Hardware overhead of special function blocks 65
4.4.2. Buffered hardware looping ……………………66
4.4.3. Streamed I/O ……………………………………67
5. Low Power Designs ………………………………72
5.1. Introduction …………………………………………72
5.2. Reducing execution cycle …………………………72
5.2.1. Data forwarding ………………………………73
5.2.2. Dual MAC and sub-word MAC…………………..73
5.2.3. Hamming distance calculator ……………………73
5.2.4. Multi-level slicer ………………………………73
5.3. Reducing switching activities …………………………74
5.3.1. Gray code addressing …………………………74
5.3.2. Buffered hardware looping ……………………75
5.3.3. Pipeline sharing ………………………………76
6. Module Generator ……………………………………78
6.1. Introduction …………………………………………78
6.2. Structure consideration ………………………………79
6.3. Parameters consideration …………………………80
6.4. Graph user interface ………………………………80
7. Implementation and Design Results ………………86
7.1. Introduction …………………………………………86
7.2. Chip implementation ………………………………86
7.3. Gray code addressing ………………………………88
7.4. Parameterized design ………………………………88
7.5. Parameterized design evaluation ……………………90
7.5.1. Boundary estimator…………………………… 91
7.5.2. Interpolator ……………………………………93
7.5.3. Timing error detector …………………………96
7.5.4. Loop filter & timing controller ………………98
8. Conclusions …………………………………………102
8.1. Conclusions …………………………………………102
8.2. Future works …………………………………………103
9. References …………………………………………104
Appendix: Instruction Set ………………………………110
List of Publications………………………………………….156

[1]D. Clark, “Innovation and Obstacles: the Future of Computing,” IEEE Computer, vol.31, no.1, pp.29-38, Jan. 1998.
[2]G. Frantz, ”Digital Signal Processor Trends,” IEEE Micro, vol.20, no.6, pp.52 -59, Nov./Dec. 2000.
[3]W. Strauss,”DSP STRATEGIES -The Embedded Chip Imperative”, Report no.4010, Forward Concepts Co. 2002.
[4]J. Turley and H. Hakkarainen, “TI’s New ’C6x DSP Screams at 1600 MIPS,” Microprocessor Report, vol.11, pp.14–17, 1997.
[5]I. Verbauwhede, M. Touriguian, "A Low Lower DSP Engine for Wireless Communications," Journal of VLSI Signal Processing Systems, vol.18, no.2, pp.177-186, Feb. 1998.
[6]DSP1618 digital signal processor, AT&T Data Sheet, Feb. 1994.
[7]TMS320C54X DSP Reference Set: Volume 1: CPU and Peripherals, Texas Instruments, 1997.
[8]M. Alidina, G. Burns, C. Holmqvist, E. Morgan, and D. Rhodes, "DSP16000: A High performance, Low Power Dual-MAC DSP Core for Communication Applications," Proceedings of IEEE Custom Integrated Circuits Conference, 1998, pp.119-122.
[9]B. W. Kim, et. Al., "MDSP-II: A 16-Bit DSP with Mobile Communication Accelerator," IEEE Journal of Solid-State Circuits, vol.34, no.3, pp.397-404, Mar. 1999.
[10]Y. Tsao, S. Jou, H. Lee, Y. Chen, M. Tan “An Embedded DSP Core for Wireless Communication”, Proceedings of International Symposium on Circuit and System, vol.4, May 2002, pp.524-527.
[11]R. Mehra, L. M. Guerra, J. M. Rabaey, “A Partitioning Scheme for Optimizing Interconnect Power,” IEEE Journal of Solid-State Circuits, vol. 32 no.3, pp.433-443, Mar. 1997.
[12]M. Kuulusa, J. Nurmi, J. Jakala, P. Ojala, H. Herranen, “A Flexible DSP Core for Embedded Systems,” IEEE Design & Test of Computers, vol.14, no.4, pp.60-68, Oct.-Dec. 1997.
[13]A. Gierlinger, R. Forsyth, E. Ofner, “GEPARD: A Parameterisable DSP Core for ASICS”, Proceedings on International Conference on Signal Processing Applications & Technology, 1997, pp.203-207.
[14]F. Culloch, “Speeding development of multi-DSP of applications,” Embedded Edge, pp.22-29, June 2001.
[15]C. K. Chen, P. C. Tseng, Y. C. Chang, L. G. Chen, “A digital signal processor with programmable correlator array architecture for third generation wireless communication system,” IEEE Transactions on Circuits and Systems II: Analog and Digital Signal Processing, vol.48, Issue:12, pp.1110-1120, Dec. 2001.
[16]J. L. Hennessy, D. A. Patterson, “Computer Organization & Design: The Hardware/Software Interface,” 2nd edition, Morgan Kaufmann Publishers, 1998.
[17]M. Sami, D. Sciuto, C. Silvano,; V. Zaccaria, R. Zafalom, “Exploiting data forwarding to reduce the power budget of VLIW embedded processors,” Proceedings of Conference and Exhibition on Design, Automation and Test in Europe, 2001, pp.252 -257.
[18]P. Lapsley, J. Bier, A. Shoham, E. A. Lee, "DSP Processor Fundamentals," IEEE Press, 1997.
[19]J.Y. Chen, W.B. Jone, J.S. Wang, H.-I. Lu, T.F. Chen, “Segmented bus design for low-power systems,” IEEE Transactions on Very Large Scale Integration (VLSI) Systems, vol.7, Issue:1, pp.25-29, Mar. 1999.
[20]H. Yagi, R. E. Owen, “Architectural considerations in a configurable DSP core for consumer electronics,” Workshop on IEEE Signal Processing, Sep. 1995, pp.70-81.
[21]M. C. Lin, C. L. Chen, D. Y. Shin, C. H. Lin and S. J. Jou, “Low-Power Multiplierless FIR Filter Synthesizer Based on CSD Code,” IEEE International Symposium on Circuits and Systems, May 2001, pp.IV-666-IV-669.
[22]C. Wu, T. T. Hwang, “Instruction buffering for nested loops in low power design,” IEEE International Symposium on Circuits and Systems, vol.4, 2002, pp.IV-81 -IV-84.
[23]M.L.L. Brackenbury, “An instruction buffer for a low-power DSP,” Proceedings. of Sixth International Symposium on Advanced Research in Asynchronous Circuits and Systems, 2000, pp.176 -186.
[24]R.S. Bajwa, M. Hiraki, H. Kojima, D.J. Gorny, K. Nitta, A. Shridhar, K. Seki, K. Sasaki, “Instruction buffering to reduce power in processors for signal processing,” IEEE Transactions on Very Large Scale Integration (VLSI) Systems, vol.5, Issue:4, pp.417-424, Dec. 1997.
[25]L. H. Lee, B. Moyer and J. Arends, “Instruction fetch energy reduction using loop caches for embedded applications with small tight loops,” International Symposium on Low Power Electronics and Design, Aug. 1999, pp.267 – 269.
[26]C. Panis, H. Grunbacher and J. Nurmi, “A scalable instruction buffer and align unit for xDSPcore,” IEEE Journal of Solid-State Circuits, vol.39, Issue 7, pp.1094 – 1100, July 2004.
[27]I. Verbauwhede, C. Nicol, “Low power DSP's for wireless communications,” Proceedings of the 2000 International Symposium on Power Electronics and Design, 26-27 July 2000, pp.303-310.
[28]J. Warden, “Sub-Word Parallelism in Digital Signal Processing,” IEEE Signal Processing Magazine, vol.17, no.2, pp.27-35, Mar. 2000.
[29]Y. Huang, T. D. Chiueh, “A Sub-Word Parallel Digital Signal Processor for Wireless Communication Systems,” Proceedings on IEEE Asia-Pacific Conference on ASIC, Aug. 2002, pp.287-290.
[30]J. Terry and J. Heiskala, “OFDM Wireless LANs: A Theoretical and Practical Guide,” Sams Publishing, 2002.
[31]M. H. Hsieh and C. H. Wei, ”A Low-Complexity Frame Synchronization and Frequency Offset Compensation Scheme for OFDM Systems over Fading Channels,” IEEE Transactions on Vehicular Technology ,vol.48, no.5, pp. 1596-1609,Sep. 1999.
[32]C. L. Su, C. Y. Tsui, A. M. Despain, “Low Power Architecture Design and Compilation Techniques for High-Performance Processors,” IEEE International Computer Conference, Spring 1994, pp.489-498.
[33]H. Mehta, R.M. Owens, M.J. Irwin, “Some issues in gray code addressing,” Proceedings of Sixth Great Lakes Symposium on VLSI, 22-23 Mar. 1996, pp.178-181.
[34]C. C. Chang, M. S. Wang and T. D. Chiueh, “Design of a DMT-based baseband transceiver for very-righ-speed digital subscriber lines,” Proceeding on IEEE Asia-Pacific Advanced System Integrated Circuits, Aug. 2002, pp.367-370.
[35]F. M. Gardner, “Interpolation in digital modems – part I: fundamentals,” IEEE Transactions on Communications, vol.41, pp.501-507, Mar. 1993.
[36]L. Erup, F. M. Gardner, R. A. Harris, “Interpolation in digital modems – part II: implementation and performance,” IEEE Transactions on Communications, vol.41, pp.998-1008, June 1993.
[37]E. Martos-Naya, J. Lopez-Fernandez, L. D. del Rio, M. C. Aguayo-Torres, J. T. E. Munoz, “Optimized interpolator filters for timing error correction in DMT systems for xDSL applications,” IEEE Journal on Communications, vol.19, pp.2477-2485, Dec. 2001.