基于AMR标准的语音编码VLSI设计
|
摘要
随着现代无线多媒体通信的发展,以及当前通信系统有限的带宽资源,发展高质量和低码率的的语音编码器成为很迫切的课题。AMR(自适应多速率)语音编码器由于其低码率和高质量,使之在通信和多媒体领域得到广泛应用,同时由于集成电路设计方法学和工具软件的不断发展,低成本、低功耗、高性能的AMR语音编码器已经完全可以实现。
对AMR编码器的原理作了深入的研究,并对算法的C实现做了详细的模拟和分析。根据模拟得到的数据,针对AMR语音编码标准的算法特点,提出了一种“DSP软核+ASIC加速器”的体系结构。在设计好体系结构后,对编码器算法进行具体的硬件建模。首先,优化DSP软核,基于该DSP软核对AMR编码算法进行了改进;接着将AMR编码器用全汇编实现和优化,对汇编代码进行测试并给出了结果数据;然后根据汇编测试数据,详细地分析和设计了ASIC加速器;最后对整个设计进行了详细地验证和分析。采用了FPGA原型验证,在功能方面保证了设计的功能正确性。同时对设计中的ASIC加速器中的模块采用了复用设计,使得本设计方案对G.72x系列语音编码算法具有通用性和可移植性。
设计在HJTC0.18μm工艺库上实现,工作电压为1.8V,最高工作频率达110MHz。当工作在20MHz频率的时候,就可以满足语音编码实时性20ms的延时要求,此时设计功耗仅有20mW,设计面积只有15万门(包括片上存储器),最后设计进行了MPW投片。
With the development of wireless communication systems and the limit of communication systems bandwidth resource, research and design of speech coder with high quality and low bits rate imminently. AMR(adaptive mult rate)speech coder using widely in communication and mult-media field for high quality and low bits rate. With rapid and successful development in integrated circurits design method and software, design a AMR speech coder IP practically.
In this paper, research AMR speech coder principle deeply, then simulate and analyze the c code in detail. Then, with the data of simulation, presents a structure of“DSP soft core + ASIC accelerator”based on AMR speech coding algorithm. After that , hardware model should be implemented in detail. According the new structure, optimize the DSP core and improved AMR speech coding algorithm. After that, write assembler code and optimize the assembler code. Once the assembler design is finished, test and simulate the code, the record the data of testing and simulation. ASIC accelerator could be design using these result data. This design was implemented on FPGA device, and the experimental results indicate that the structure is feasible. Meanwhile , as the design in this structure can also be applied in many other speech coding algorithms like G.72X serial speech coder.
This design implementes on the HJTC0.18μm process, voltage is 1.8V. After synthesis, the design can work utmost at 110 MHz. While working at 20 MHz, the design can meet the 20ms limit of timing, and only have 20mW power and 150 K gates chip area.(include on chip memory).The design will be taped out in last.
对AMR编码器的原理作了深入的研究,并对算法的C实现做了详细的模拟和分析。根据模拟得到的数据,针对AMR语音编码标准的算法特点,提出了一种“DSP软核+ASIC加速器”的体系结构。在设计好体系结构后,对编码器算法进行具体的硬件建模。首先,优化DSP软核,基于该DSP软核对AMR编码算法进行了改进;接着将AMR编码器用全汇编实现和优化,对汇编代码进行测试并给出了结果数据;然后根据汇编测试数据,详细地分析和设计了ASIC加速器;最后对整个设计进行了详细地验证和分析。采用了FPGA原型验证,在功能方面保证了设计的功能正确性。同时对设计中的ASIC加速器中的模块采用了复用设计,使得本设计方案对G.72x系列语音编码算法具有通用性和可移植性。
设计在HJTC0.18μm工艺库上实现,工作电压为1.8V,最高工作频率达110MHz。当工作在20MHz频率的时候,就可以满足语音编码实时性20ms的延时要求,此时设计功耗仅有20mW,设计面积只有15万门(包括片上存储器),最后设计进行了MPW投片。
With the development of wireless communication systems and the limit of communication systems bandwidth resource, research and design of speech coder with high quality and low bits rate imminently. AMR(adaptive mult rate)speech coder using widely in communication and mult-media field for high quality and low bits rate. With rapid and successful development in integrated circurits design method and software, design a AMR speech coder IP practically.
In this paper, research AMR speech coder principle deeply, then simulate and analyze the c code in detail. Then, with the data of simulation, presents a structure of“DSP soft core + ASIC accelerator”based on AMR speech coding algorithm. After that , hardware model should be implemented in detail. According the new structure, optimize the DSP core and improved AMR speech coding algorithm. After that, write assembler code and optimize the assembler code. Once the assembler design is finished, test and simulate the code, the record the data of testing and simulation. ASIC accelerator could be design using these result data. This design was implemented on FPGA device, and the experimental results indicate that the structure is feasible. Meanwhile , as the design in this structure can also be applied in many other speech coding algorithms like G.72X serial speech coder.
This design implementes on the HJTC0.18μm process, voltage is 1.8V. After synthesis, the design can work utmost at 110 MHz. While working at 20 MHz, the design can meet the 20ms limit of timing, and only have 20mW power and 150 K gates chip area.(include on chip memory).The design will be taped out in last.
引文
1王炳锡.语音编码.西安电子科技大学出版社. 2002:118~120
2 Atal B, Remde J. A New Model of LPC Excitation for Producing Natural -sounding Speech at Low Bit Rates. Acoustics, Speech and Signal Processing, IEEE International Conference on ICASSP’82, Paris,France, 1982: 614~617
3黄桃,李小文,袁李林. AMR语音编解码算法分析及优化.通信技术. 2007,40(12):391~393
4 Tao Huang, Xiaowen Li, Shitao Li. AMR Codec Mode Selection in TD-SCDMA. IEEE International Workshop on Cross Layer Design, Jinan,China, 2007: 157~160
5夏娟,申敏. TD-SCDMA中AMR语音编码器的实现. 2004,16(1):51~53
6 Lundberg T, de Bruin P, Bruhn S, Hakansson S, Craig S. Adaptive Thresholds for AMR Codec Mode Selection. IEEE 61st Vehicular Technology Conference, Stockholm,Sweden, 2005:2325~2329
7郑雪帆,刘春,陆诚等. AMR语音编码算法研究及复杂度剖析.电讯技术. 2003,(1):92~96
8余志刚,胡波. AMR与G.729A的参数直接转换算法.信息与电子工程.2005,3(4):253~256.
9孙健,吴镇扬,杜涛.从AMR到G.729的参数层编码转换方案.信息与电子工程. 2004,2(1):17~20..
10 Feize Song, Wan Lei, Zhao Shenghui, Kuang Jingming. Investigation of Adaptive Multi-Rate Link Adaptation of 3G System. Journal of Beijing Institute of Technology. 2002, 11( 4):369~374
11 Imre Varga. Audio Codec for Mobile Multimedia Applications. IEEE 6th Workshop on Multimedia Signal Processing, Siena,Italie, 2004:450~453
12 Gan X D, Chen T, Sii.S.M. Implementation of Silence Compression Scheme for G.723.1 Speech Coder Using TI TMS320C51 DSP Chip. Proceedings of 1997 International Conference on Information, Communications and Signal Processing, Singapore, 1997: 1284~1287
13 Jaewon Kim,Hyungjung Kim,Songin Choi ,Younggap You. The Implementation of G.729 Speech Coder on A 16-bit DSP Chip for The CDMA IMT-2000 System. IEEE Trans. on Consumer Electronics. 1999,45(2):443~448
14 Langi, A Z R. Rapid Development of A Real-time Speech Coder on A TMS320C54x DSP. IEEE 2002 Canadian Conference on Electrical and Computer Engineering, Winnipeg,Canada, 2002: 1045~1048.
15 3GPP TS 26.071 V6.0.0. AMR Speech CODEC:General Description. 2004.
16鲍长春.数字语音编码原理.西安电子科技大学出版社. 2007: 171~172
17 Jerry D. Gibson. Speech Coding Methods, Standards,and Applications. IEEE Circutis and Systems Magzine. 2005,(4): 33~36
18 M.Sadek, S EI-Ramly, A Tawfik. A New CELP Speech Coder for Wireless Communication. IEEE 2002 14th International Conference on Digital Signal Processing, island of Santorini,Greece, 2002: 997~1000
19 S M. Tsai, J F Yang. Efficient Algebraic Code-excited Linear Predictive. IEE Proc.-Vis. Image Signal Process. 2006,1(6):761~768
20 R. Salami, C. Laflamme, J-P. Adoul, and D. Massaloux. A toll quality 8 kbis speech codec for the personal communications system (PCS). IEEE Transactions on Vehicular Technology. 1994,43(3):808~816
21梅寒.第三代移动通信中的语音编(AMR).北京邮电大学硕士论文. 2007:17~25
22李宋琛.第三代移动通信中的AMR声码器研究.电子科技大学硕士论文. 2002:7~19
23 E Ekudden, R Hagen, I Johansson, J Svedberg. The Adaptive Multi-rate Speech Coder. IEEE Workshop on Speech Coding proccedings, Porvoo,Finland, 1999:117~119
24朱敏. AMR-WB编码算法的研究以及AMR-NB基于ZSP500的优化实现.南京邮电大学硕士学位论文. 2006:29~45
25杨猛. AMR算法的DSP实现与优化.北京邮电大学硕士学位论文. 2005:29~35
26 Chen Xinfu, Zhang Zhengyang, Yi Kechu. AMR Vocoder and Iits Multi-channel Implementation Based on A Single DSP Chip. ICII 2001 International Conferences, Beijing,China, 2001: 650~655.
27 Kyung Jin Byun, Ik Soo Eo, Hee Bum Jeong. Real-time Implementation of AMR and AMR-WB Using the Fixed-point DSP for WCDMA Systems. IEEE Tenth International Symposium on Consumer Electronics, StPetersburg,Russia 2006:1~6.
28 Kyung Jin Byun, Hee Bum Jung, Minsoo Hahn, Kyung Soo Kim. Computationally Efficient Implementation of AMR Speech Coder. Proceedings of the 3rd International Symposium on Image and Signal Processing and Analysis, Rome Italy, 2003: 528~531
29 Jie Yang, Sheng sheng Yu, Mian Zhao. The Implementation and Optimization of AMR Speech Codec on DSP. International Symposium on Intelligent Signal Processing and Communication Systems, HuaqiaoUniversty,Xiamen,China, 2007:365 ~367
30鲍长春,窦庚欣,范睿,刘泽新. ACELP语音编码中代数码书的快速搜索算法.北京工业大学学报. 2007,33(8):823~827.
31 Erdal Paksoy, Juan Carlos De Martinf, Alan McCree, Christian G. Gerlach. An Adaptive Multi-rate Speech Coder for Digital Cellular Telephony. IEEE Proceedings International Conference on Acoustics, Speech,and Signal Processing, Phoenix, Arizona,USA, 1999:193~196
32 Kyung Jin Byun, Hee Bum Jung, Minsoo Hahn. A Fast ACELP Codebook Search Method. IEEE 2002 6th International Conference on Signal Processing, Beijing,China, 2002: 422~425
33 Jan M Rabaey, Anantha Chandrakasan, Borivoje Nikolic.数字集成电路-电路、系统与设计.电子工业出版社. 2006:277
34林奕琳,李巧玲,李江源等. AMR-WB语音编码算法及仿真.计算机工程与应用. 2003,(29):69
35 M. R. Smith. How RISCy is DSP. IEEE Micro. 1992, 12(6):10~23
36 Kyoung Ho Bang, Nam Hun Jeong, Joon Seok Kim. Design and VLSI Implmentation of a Digital Audio-specific DSP Core for MP3/AAC. IEEE Transactions on Consumer Electronics. 2002,48(3):790~794
37严伟,龚幼民.高性能数字信号处理器的设计.微处理机. 2004, (4):10~14
38 Jan Weil. Optimierung Eines AMR-Codecs Für Eingebettete RISC Prozessoren. Technique University Berlin ,Sommersemester 2004:23~33
39陈文. AMR声码器自适应码本探索技术及其DSP实现.电子科技大学硕士学位论文. 2000:58~60
40刘春余,王晓霞. AMR语音编解码在DSP上的实现与优化.无线电工程. 2007,37(2):49~51
41 Rachana Nagal, Manish Kumar, Rahul Jain. Optimal Real Time DSP Implemen-tation of Extended Adaptive Multirate Wide Band(AMR-WB+) Speech Codec. TENCON 2008 IEEE Region 10 Conference, Hyderabad,India, 2008:1~6
42 Koczela, L.J, Wang, G.Y. The Design of a Highly Parallel Computer Organization. IEEE Transactions on Computers. 1969,C-18(6): 520~529
43 David A Patterson, John L Hennessy. Computer Organization and Design:The Hardware/Software Interface,Third Edition. Elsevier Inc. 2007:238~240.
44崔晓平,王成华.基于方块超前进位的快速进位跳跃加法器.南京航空航天大学学报. 2006,38(6):786~790
45 Hussin, R, Yeon, A, Shakaff, N, Idris, N, Ismail, R.C, Kamarudin, A. Improved Booth Encoding for Reduced Area Multiplier. IEEE International Conference on Semiconductor Electronics, Shanghai,China, 2006: 773~775
2 Atal B, Remde J. A New Model of LPC Excitation for Producing Natural -sounding Speech at Low Bit Rates. Acoustics, Speech and Signal Processing, IEEE International Conference on ICASSP’82, Paris,France, 1982: 614~617
3黄桃,李小文,袁李林. AMR语音编解码算法分析及优化.通信技术. 2007,40(12):391~393
4 Tao Huang, Xiaowen Li, Shitao Li. AMR Codec Mode Selection in TD-SCDMA. IEEE International Workshop on Cross Layer Design, Jinan,China, 2007: 157~160
5夏娟,申敏. TD-SCDMA中AMR语音编码器的实现. 2004,16(1):51~53
6 Lundberg T, de Bruin P, Bruhn S, Hakansson S, Craig S. Adaptive Thresholds for AMR Codec Mode Selection. IEEE 61st Vehicular Technology Conference, Stockholm,Sweden, 2005:2325~2329
7郑雪帆,刘春,陆诚等. AMR语音编码算法研究及复杂度剖析.电讯技术. 2003,(1):92~96
8余志刚,胡波. AMR与G.729A的参数直接转换算法.信息与电子工程.2005,3(4):253~256.
9孙健,吴镇扬,杜涛.从AMR到G.729的参数层编码转换方案.信息与电子工程. 2004,2(1):17~20..
10 Feize Song, Wan Lei, Zhao Shenghui, Kuang Jingming. Investigation of Adaptive Multi-Rate Link Adaptation of 3G System. Journal of Beijing Institute of Technology. 2002, 11( 4):369~374
11 Imre Varga. Audio Codec for Mobile Multimedia Applications. IEEE 6th Workshop on Multimedia Signal Processing, Siena,Italie, 2004:450~453
12 Gan X D, Chen T, Sii.S.M. Implementation of Silence Compression Scheme for G.723.1 Speech Coder Using TI TMS320C51 DSP Chip. Proceedings of 1997 International Conference on Information, Communications and Signal Processing, Singapore, 1997: 1284~1287
13 Jaewon Kim,Hyungjung Kim,Songin Choi ,Younggap You. The Implementation of G.729 Speech Coder on A 16-bit DSP Chip for The CDMA IMT-2000 System. IEEE Trans. on Consumer Electronics. 1999,45(2):443~448
14 Langi, A Z R. Rapid Development of A Real-time Speech Coder on A TMS320C54x DSP. IEEE 2002 Canadian Conference on Electrical and Computer Engineering, Winnipeg,Canada, 2002: 1045~1048.
15 3GPP TS 26.071 V6.0.0. AMR Speech CODEC:General Description. 2004.
16鲍长春.数字语音编码原理.西安电子科技大学出版社. 2007: 171~172
17 Jerry D. Gibson. Speech Coding Methods, Standards,and Applications. IEEE Circutis and Systems Magzine. 2005,(4): 33~36
18 M.Sadek, S EI-Ramly, A Tawfik. A New CELP Speech Coder for Wireless Communication. IEEE 2002 14th International Conference on Digital Signal Processing, island of Santorini,Greece, 2002: 997~1000
19 S M. Tsai, J F Yang. Efficient Algebraic Code-excited Linear Predictive. IEE Proc.-Vis. Image Signal Process. 2006,1(6):761~768
20 R. Salami, C. Laflamme, J-P. Adoul, and D. Massaloux. A toll quality 8 kbis speech codec for the personal communications system (PCS). IEEE Transactions on Vehicular Technology. 1994,43(3):808~816
21梅寒.第三代移动通信中的语音编(AMR).北京邮电大学硕士论文. 2007:17~25
22李宋琛.第三代移动通信中的AMR声码器研究.电子科技大学硕士论文. 2002:7~19
23 E Ekudden, R Hagen, I Johansson, J Svedberg. The Adaptive Multi-rate Speech Coder. IEEE Workshop on Speech Coding proccedings, Porvoo,Finland, 1999:117~119
24朱敏. AMR-WB编码算法的研究以及AMR-NB基于ZSP500的优化实现.南京邮电大学硕士学位论文. 2006:29~45
25杨猛. AMR算法的DSP实现与优化.北京邮电大学硕士学位论文. 2005:29~35
26 Chen Xinfu, Zhang Zhengyang, Yi Kechu. AMR Vocoder and Iits Multi-channel Implementation Based on A Single DSP Chip. ICII 2001 International Conferences, Beijing,China, 2001: 650~655.
27 Kyung Jin Byun, Ik Soo Eo, Hee Bum Jeong. Real-time Implementation of AMR and AMR-WB Using the Fixed-point DSP for WCDMA Systems. IEEE Tenth International Symposium on Consumer Electronics, StPetersburg,Russia 2006:1~6.
28 Kyung Jin Byun, Hee Bum Jung, Minsoo Hahn, Kyung Soo Kim. Computationally Efficient Implementation of AMR Speech Coder. Proceedings of the 3rd International Symposium on Image and Signal Processing and Analysis, Rome Italy, 2003: 528~531
29 Jie Yang, Sheng sheng Yu, Mian Zhao. The Implementation and Optimization of AMR Speech Codec on DSP. International Symposium on Intelligent Signal Processing and Communication Systems, HuaqiaoUniversty,Xiamen,China, 2007:365 ~367
30鲍长春,窦庚欣,范睿,刘泽新. ACELP语音编码中代数码书的快速搜索算法.北京工业大学学报. 2007,33(8):823~827.
31 Erdal Paksoy, Juan Carlos De Martinf, Alan McCree, Christian G. Gerlach. An Adaptive Multi-rate Speech Coder for Digital Cellular Telephony. IEEE Proceedings International Conference on Acoustics, Speech,and Signal Processing, Phoenix, Arizona,USA, 1999:193~196
32 Kyung Jin Byun, Hee Bum Jung, Minsoo Hahn. A Fast ACELP Codebook Search Method. IEEE 2002 6th International Conference on Signal Processing, Beijing,China, 2002: 422~425
33 Jan M Rabaey, Anantha Chandrakasan, Borivoje Nikolic.数字集成电路-电路、系统与设计.电子工业出版社. 2006:277
34林奕琳,李巧玲,李江源等. AMR-WB语音编码算法及仿真.计算机工程与应用. 2003,(29):69
35 M. R. Smith. How RISCy is DSP. IEEE Micro. 1992, 12(6):10~23
36 Kyoung Ho Bang, Nam Hun Jeong, Joon Seok Kim. Design and VLSI Implmentation of a Digital Audio-specific DSP Core for MP3/AAC. IEEE Transactions on Consumer Electronics. 2002,48(3):790~794
37严伟,龚幼民.高性能数字信号处理器的设计.微处理机. 2004, (4):10~14
38 Jan Weil. Optimierung Eines AMR-Codecs Für Eingebettete RISC Prozessoren. Technique University Berlin ,Sommersemester 2004:23~33
39陈文. AMR声码器自适应码本探索技术及其DSP实现.电子科技大学硕士学位论文. 2000:58~60
40刘春余,王晓霞. AMR语音编解码在DSP上的实现与优化.无线电工程. 2007,37(2):49~51
41 Rachana Nagal, Manish Kumar, Rahul Jain. Optimal Real Time DSP Implemen-tation of Extended Adaptive Multirate Wide Band(AMR-WB+) Speech Codec. TENCON 2008 IEEE Region 10 Conference, Hyderabad,India, 2008:1~6
42 Koczela, L.J, Wang, G.Y. The Design of a Highly Parallel Computer Organization. IEEE Transactions on Computers. 1969,C-18(6): 520~529
43 David A Patterson, John L Hennessy. Computer Organization and Design:The Hardware/Software Interface,Third Edition. Elsevier Inc. 2007:238~240.
44崔晓平,王成华.基于方块超前进位的快速进位跳跃加法器.南京航空航天大学学报. 2006,38(6):786~790
45 Hussin, R, Yeon, A, Shakaff, N, Idris, N, Ismail, R.C, Kamarudin, A. Improved Booth Encoding for Reduced Area Multiplier. IEEE International Conference on Semiconductor Electronics, Shanghai,China, 2006: 773~775