EPSILON-CAPACITY OF CLASSES OF UNKNOWN CHANNELS

被引:9
作者
FORYS, LJ
VARAIYA, PP
机构
[1] Department of Electrical Engineering and Computer Sciences, University of California, Berkeley
[2] Bell Telephone Laboratories, Whippany
[3] Department of Electrical Engineering and Computer Sciences, University of California, Berkeley
来源
INFORMATION AND CONTROL | 1969年 / 14卷 / 04期
基金
美国国家航空航天局; 美国国家科学基金会;
关键词
D O I
10.1016/S0019-9958(69)90101-6
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
In the Shannon theory of communication, nearly all of the results pertaining to the rate of transmission of information have depended upon accurate characterizations of the channel operator and the statistics of the noise. The concept of the ε-capacity of classes of unknown channels can be applied to estimate the capacity of channels having additive noise when only a rudimentary knowledge of both the channel operator and the noise statistics are available. We discuss three types of channel models in which the channel operator is known only to belong to a class of such operators. Bounds are established for the ε-capacity of these classes of operators. These bounds are determined through an application of known results on the channel capacity of channel models having additive Gaussian noise. The resulting bounds when specialized to the case where the class consists of only a single operator are often better than the results obtained by other investigators who consider only single operators. © 1969 Academic Press, Inc.
引用
收藏
页码:376 / &
相关论文
共 8 条
[1]   THE CAPACITY OF A CLASS OF CHANNELS [J].
BLACKWELL, D ;
BREIMAN, L ;
THOMASIAN, AJ .
ANNALS OF MATHEMATICAL STATISTICS, 1959, 30 (04) :1229-1241
[2]  
Bochner S., 1949, FOURIER TRANSFORMS
[3]  
FORYS LJ, 1968, ERLM242 U CAL MEM
[4]  
GALLAGER RG, 1968, INFORMATION THEORY R
[5]  
Kolmogorov A., 1959, USP MAT NAUK, V14, P3
[6]  
PROSSER RT, 1966, J MATH ANALYSIS APPL, V16
[7]   CAPACITY OF CLASSES OF GAUSSIAN CHANNELS [J].
ROOT, WL ;
VARAIYA, PP .
SIAM JOURNAL ON APPLIED MATHEMATICS, 1968, 16 (06) :1350-&
[8]  
ROOT WL, 1968, IEEE T INFORM THEORY, VIT14, P361