- [Home]
- [Research achievement]
- [Research achievement detail]
Title | Variable-Length Intrinsic Randomness Problem Allowing Non-Vanishing Underflow Probability (in Japanese) |
---|---|
Authors | Jun Yoshizawa 、Shota Saito 、Toshiyasu Matsushima |
Released Year | 2017 |
Format | Conference |
Category | Source coding |
Jounal Name | |
Jounal Page | vol. 117, no. 120, IT2017-29, pp. 73-78 |
Published Year | 2017 |
Published Month | 7 |
Abstract (English) |
This paper deals with the variable-length intrinsic randomness problem for general sources. Previous study adopts the mean length as performance criterion in variable-length intrinsic randomness problem. On the other hand, this study focuses on the underflow probability instead of the mean length. Under the constraint that the underflow probability is smaller than or equal to some constant, we characterize the supremum of underflow thresholds by using quantity defined by entropy spectrum. |
Note (English) |
|
Manuscript | |
Presentation |