| Project/Area Number |
01550268
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
電子通信系統工学
|
|---|
| Research Institution | Kyushu University |
|---|
Principal Investigator |
MOTOISHI Kohji Kyushu University, Faculty of Eng., Associate Professor, 工学部, 助教授 (00038118)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
SUHARA Yosiro Kyushu University, Faculty of Eng., Research Assistant, 工学部, 助手 (80187799)
KOHDA Tohru Kyushu University, Faculty of Eng., Associate Professor, 工学部, 助教授 (20038102)
NISHI Tetsuo Kyushu University, Faculty of Eng., Professor, 工学部, 教授 (40037908)
KOGA Tosiro Kyushu University, Faculty of Eng., Professor, 工学部, 教授 (00037706)
|
|---|
| Project Period (FY) |
1989 – 1990
|
| Project Status |
Completed (Fiscal Year 1990)
|
| Budget Amount *help |
¥1,800,000 (Direct Cost: ¥1,800,000)
Fiscal Year 1990: ¥800,000 (Direct Cost: ¥800,000)
Fiscal Year 1989: ¥1,000,000 (Direct Cost: ¥1,000,000)
|
|---|
| Keywords | Data Compression / Error Correction / Discrete Cosine Transform / Impulse Rejection / Analog DCT code / インパルス除法 / アナログ符号 |
|---|
| Research Abstract |
Generally, data compression is performed in the analog domain and error correction is performed in the digital domain. So far it was difficult to realize data compression and error correction as a whole, although both techniques are related to redundancy control. In this research, we studied a unified technique of redundancy control.
1. Discrete Cosine Transform (DCT) is well known as an excellent data compression technique for the voice and image data. We investigated DCT from a new angle on the analog codes.
2. By modifying appropriately, DCT is able to have error correcting ability. We call this as DCT analog code.
3. The error correcting capability of the DCT analog code is [L/4]. By expanding receiving sequence in a mirror-symmetric form, we can improve the capability to [L/2]. L is the number of additional redundancy points and [.] is denoted as the gaussian symbol.
4. The Berlekamp-Massy algorithm or the Euclidian algorithm can be principally adopted as a decoding algorithm. However, these algorithms are not robust against the stationary gaussian noise on the communication channel.
5. In order to reduce the influence of the stationary gaussian noise, we studied the co-variance method and a kind of the majority voting method. The majority voting method with orthogonal vectors very well performed in case of short code length.
A demerit of majority voting method is that the number of orthogonal vectors rapidly increases as the code length is longer. The reduction of the number of orthogonal vectors is a subject for a future study.
|
