Remarks on Confidence Intervals for Self-Similarity Parameter of a Subfractional Brownian Motion //// Hindawi Publishing Corporation Home Journals About Us About this Journal Submit a Manuscript Table of Contents Journal Menu Abstracting and Indexing Aims and Scope Annual Issues Article Processing Charges Articles in Press Author Guidelines Bibliographic Information Contact Information Editorial Board Editorial Workflow Free eTOC Alerts Reviewers Acknowledgment Subscription Information Open Special Issues Published Special Issues Special Issue Guidelines Abstract Full-Text PDF Full-Text HTML Full-Text ePUB Linked References How to Cite this Article Abstract and Applied Analysis Volume 2012 (2012), Article ID 804942, 14 pages doi:10.1155/2012/804942 Research Article <h2>Remarks on Confidence Intervals for Self-Similarity Parameter of a Subfractional Brownian Motion</h2> Junfeng Liu , 1 Litan Yan , 2 Zhihang Peng , 3 and Deqing Wang 4 1 Department of Mathematics, Nanjing Audit University, 86 West Yushan Road, Pukou, Nanjing 211815, China 2 Department of Mathematics, Donghua University, 2999 North Renmin Road, Songjiang, Shanghai 201620, China 3 Department of Epidemiology and Biostatistics, Nanjing Medical University, 140 Hanzhong Road, Gulou, Nanjing 210029, China 4 Department of Statistics, Xiamen University, 422 South Siming Road, Siming, Xiamen 361005, China Received 4 August 2011; Revised 11 October 2011; Accepted 28 October 2011 Academic Editor: Ljubisa Kocinac Copyright © 2012 Junfeng Liu et al. This is an open access article distributed under the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract We first present two convergence results about the second-order quadratic variations of the subfractional Brownian motion: the first is a deterministic asymptotic expansion; the second is a central limit theorem. Next we combine these results and concentration inequalities to build confidence intervals for the self-similarity parameter associated with one-dimensional subfractional Brownian motion. 1. Introduction A fundamental assumption in many statistical and stochastic models is that of independent observations. Moreover, many models that do not make the assumption have the convenient Markov property, according to which the future of the system is not affected by its previous states but only by the current one. The long-range dependence property has become an important aspect of stochastic models in various scientific areas including hydrology, telecommunication, turbulence, image processing, and finance. The best known and most widely used process that exhibits the long-range dependence property is fractional Brownian motion (fBm in short). The fBm is a suitable generalization of the standard Brownian motion. The reader is referred, for example, to Alòs et al. [ 1 ] and Nualart [ 2 ] for a comprehensive introduction to fractional Brownian motion. On the other hand, many authors have proposed to use more general self-similar Gaussian processes and random fields as stochastic models. Such applications have raised many interesting theoretical questions about self-similar Gaussian processes. As a generalization of Brownian motion, recently, Bojdecki et al. [ 3 , 4 ] introduced and studied a rather special class of self-similar Gaussian processes which preserves many properties of the fractional Brownian motion. This process arises from occupation time fluctuations of branching particle systems with Poisson initial condition, which is called the subfractional Brownian motion . The so-called subfractional Brownian motion (sub-fBm in short) with index 𝐻 ∈ ( 0 , 1 ) is a mean zero Gaussian process 𝑆 𝐻 = { 𝑆 𝐻 𝑡 , 𝑡 ≥ 0 } with 𝑆 𝐻 0 = 0 and the covariance  𝑆 𝑅 ( 𝑡 , 𝑠 ) ≡ 𝐸 𝐻 𝑡 𝑆 𝐻 𝑠  = 𝑠 2 𝐻 + 𝑡 2 𝐻 − 1 2  ( 𝑠 + 𝑡 ) 2 𝐻 + | 𝑡 − 𝑠 | 2 𝐻  , ( 1 . 1 ) for all 𝑠 , 𝑡 ≥ 0 . For 𝐻 = 1 / 2 , 𝑆 𝐻 coincides with the standard Brownian motion. 𝑆 𝐻 is neither a semimartingale nor a Markov process unless 𝐻 = 1 / 2 , so many of the powerful techniques from stochastic analysis are not available when dealing with 𝑆 𝐻 . The sub-fBm has properties analogous to those of fBm (self-similarity, long-range dependence, Hölder paths) and, for 𝑠 ≤ 𝑡 , satisfies the following estimates:   2 − 2 2 𝐻 − 1   ∧ 1 ( 𝑡 − 𝑠 ) 2 𝐻   𝑆 ≤ 𝐸 𝐻 𝑡 − 𝑆 𝐻 𝑠  2  ≤   2 − 2 2 𝐻 − 1   ∨ 1 ( 𝑡 − 𝑠 ) 2 𝐻 . ( 1 . 2 ) Thus, Kolmogorov continuity criterion implies that the subfractional Brownian motion is Hölder continuous of order 𝜈 for any 𝜈 < 𝐻 . But its increments are not stationary. More works for sub-fBm can be found in Bojdecki et al. [ 3 , 4 ], Liu and Yan [ 5 , 6 ], Liu [ 7 ], Tudor [ 8 – 12 ], Yan and Shen [ 13 , 14 ], and others. The problem of the statistical estimation of the self-similarity parameter is of great importance. The self-similarity parameter characterizes all of the important properties of the self-similar processes and consequently describes the behavior of the underlying physical system. Therefore, properly estimating them is of the utmost importance. Several statistics have been introduced to this end, such as wavelets, 𝑘 -variations, variograms, maximum likelihood estimators, and spectral methods. This issue has generated a vast literature. See Chronopoulou et al. [ 15 , 16 ], Liu [ 7 ], Tudor and Viens [ 17 , 18 ], and references therein for more details. Recently, Breton et al. [ 19 ] firstly obtained the nonasymptotic construction of confidence intervals for the Hurst parameter 𝐻 of fractional Brownian motion. Observe that the knowledge of explicit nonasymptotic confidence intervals may be of great practical value, for instance in order to evaluate the accuracy of a given estimation of 𝐻 when only a fixed number of observations are available. Motivated by all these results, in the present note, we will construct the confidence intervals for the self-similarity parameter associated with the so-called subfractional Brownian motion. It is well known that, in contrast to the extensive studies on fractional Brownian motion, there has been little systematic investigation on other self-similar Gaussian processes. The main reasons are the complexity of dependence structures and the nonavailability of convenient stochastic integral representations for self-similar Gaussian processes which do not have stationary increments. As we know, in comparison with fractional Brownian motion, the subfractional Brownian motion has nonstationary increments, and the increments over nonoverlapping intervals are more weakly correlated and their covariance decays polynomially as a higher rate in comparison with fractional Brownian motion (for this reason in Bojdecki et al. [ 3 ] it is called subfractional Brownian motion). The above-mentioned properties make subfractional Brownian motion a possible candidate for models which involve long-range dependence, self-similarity, and nonstationary increments. Therefore, it seems interesting to construct the confidence intervals of self-similar parameter of subfractional Brownian motion. And we need more precise estimates to prove our results because of the nonstationary increments. The first aim of this note is to prove a deterministic asymptotic expansion and a central limit theorem of the so-called second-order quadratic variation 𝑉 𝑛 ( 𝑆 𝐻 ) which is defined by 𝑉 𝑛  𝑆 𝐻  = 𝑛 − 1  𝑘 = 0  𝑆 𝐻 ( 𝑘 + 2 ) / 𝑛 + 𝑆 𝐻 𝑘 / 𝑛 − 2 𝑆 𝐻 ( 𝑘 + 1 ) / 𝑛  2 , 𝑛 ≥ 1 , ( 1 . 3 ) because the standard quadratic variation does not satisfy a central limit theorem in general. The second aim is to exploit the concentration inequality proved by Nourdin and Viens [ 20 ] in order to derive an exact (i.e., nonasymptotic) confidence interval for the self-similar parameter of subfractional Brownian motion 𝑆 𝐻 . Our formula hinges on the class of statistics 𝑉 𝑛 ( 𝑆 𝐻 ) and 𝑍 𝑛 = 𝑛 2 𝐻 − 1 / 2 𝑉 𝑛 − √ 𝑛 ( 4 − 2 2 𝐻 ) . This note is organized as follows. In Section 2 we present some preliminaries for concentration inequality and two convergence results about the quadratic variations of some Gaussian processes. In Section 3 we prove the asymptotic expansion and central limit theorem for the second-order quadratic variations of subfractional Brownian motion with 𝐻 ∈ ( 0 , 1 ) . In Section 4 we state and prove the main result of this note. Notation . Most of the estimates of this paper contain unspecified constants. An unspecified positive and finite constant will be denoted by 𝐶 or 𝑐 , which may not be the same in each occurrence. Sometimes we will emphasize the dependence of these constants upon parameters. 2. Preliminaries Consider a finite centered Gaussian family 𝑋 = { 𝑋 𝑘 ∶ 𝑘 = 0 , … , 𝑀 } and write 𝑟 ( 𝑘 , 𝑙 ) = 𝐸 ( 𝑋 𝑘 𝑋 𝑙 ) . In what follows, we will consider two quadratic forms associated with 𝑋 and with some real coefficient 𝑐 . The first is obtained by summing up the squares of the elements of 𝑋 and by subtracting the corresponding variances 𝑄 1 ( 𝑐 , 𝑋 ) = 𝑐 𝑀  𝑘 = 0  𝑋 2 𝑘  ; − 𝑟 ( 𝑘 , 𝑘 ) ( 2 . 1 ) the second quadratic form is 𝑄 2 ( 𝑐 , 𝑋 ) = 2 𝑐 2 𝑀  𝑘 , 𝑙 = 0 𝑋 𝑘 𝑋 𝑙 𝑟 ( 𝑘 , 𝑙 ) . ( 2 . 2 ) The following result, whose proof relies on the Malliavin calculus techniques developed in Nourdin and Peccati [ 21 ], Nourdin and Viens [ 20 ],characterizes the tail behavior of 𝑄 1 ( 𝑐 , 𝑋 ) . Theorem 2.1 (Theorem 2.1 in Breton et al. [ 19 ]). If the above assumptions are satisfied, suppose that 𝑄 1 ( 𝑐 , 𝑋 ) is not a.s. zero and fix 𝛼 > 0 and 𝛽 > 0 . Assume that 𝑄 2 ( 𝑐 , 𝑋 ) ≤ 𝛼 𝑄 1 ( 𝑐 , 𝑋 ) + 𝛽 , a.s.- 𝑃 . Then, for all 𝑧 > 0 , one has 𝑃  𝑄 1   − 𝑧 ( 𝑐 , 𝑋 ) ≥ 𝑧 ≤ e x p 2   𝑄 2 𝛼 𝑧 + 2 𝛽 , 𝑃 1   − 𝑧 ( 𝑐 , 𝑋 ) ≤ − 𝑧 ≤ e x p 2  2 𝛽 , ( 2 . 3 ) in particular, 𝑃  | | 𝑄 1 | |   − 𝑧 ( 𝑐 , 𝑋 ) ≥ 𝑧 ≤ 2 e x p 2  2 𝛼 𝑧 + 2 𝛽 . ( 2 . 4 ) On the other hand, to be sure that the second-order quadratic variation 𝑉 𝑛 ( 𝑆 𝐻 ) converges almost surely to a deterministic limit, we need to normalize this quantity. A result of the form l i m 𝑛 → ∞ 𝑛 1 − 𝜃 𝑉 𝑛  𝑆 𝐻  =  1 0 𝑓 ( 𝑢 ) 𝑑 𝑢 , a . s . ( 2 . 5 ) is expected, where 𝜃 is related to the regularity of the paths of the subfractional Brownian motion 𝑆 𝐻 and 𝑓 is related to the nondifferentiability of 𝑟 on the diagonal { 𝑠 = 𝑡 } and is called the singularity function of the process. Begyn [ 22 ] considered a class of processes for which a more general normalization is needed. Moreover, he presented a better result about the asymptotic expansion of the left hand of ( 2.5 ) and proved a central limit theorem. Because Theorems 1 and 2 in Begyn [ 22 ] are crucial in the proofs of Theorems 3.1 and 3.2 , it is useful to recall the results. We define the second-order increments of the covariance function 𝑅 of a Gaussian process 𝑋 as follows: 𝛿 ℎ 1 𝛿 𝑅 ( 𝑠 , 𝑡 ) = 𝑅 ( 𝑠 + ℎ , 𝑡 ) + 𝑅 ( 𝑠 − ℎ , 𝑡 ) − 2 𝑅 ( 𝑠 , 𝑡 ) , ℎ 2 𝑅 ( 𝑠 , 𝑡 ) = 𝑅 ( 𝑠 , 𝑡 + ℎ ) + 𝑅 ( 𝑠 , 𝑡 − ℎ ) − 2 𝑅 ( 𝑠 , 𝑡 ) . ( 2 . 6 ) First, we recall the result of asymptotic expansion of 𝑉 𝑛 ( 𝑋 ) under some certain conditions on the covariance function. Theorem 2.2 (Theorem 1 in Begyn [ 22 ]). Assume that the Gaussian process 𝑋 satisfies the following statements. (1) 𝑡 → 𝑀 𝑡 = 𝔼 𝑋 𝑡 has a bounded first derivative in [ 0 , 1 ] . (2) The covariance function 𝑅 of 𝑋 has the following properties: (a) 𝑅 is continuous in [ 0 , 1 ] 2 . (b) The derivative 𝜕 4 𝑅 / 𝜕 𝑠 2 𝜕 𝑡 2 exists and is continuous in ( 0 , 1 ] 2 / { 𝑠 = 𝑡 } . There exists a constant 𝐶 > 0 , a real 𝛾 ∈ ( 0 , 2 ) and a positive slowly varying function 𝐿 ∶ ( 0 , 1 ) → ( 0 , + ∞ ) such that ] ∀ 𝑡 , 𝑠 ∈ ( 0 , 1 2 , | | | | 𝜕 { 𝑠 = 𝑡 } 4 𝑅 𝜕 𝑠 2 𝜕 𝑡 2 | | | | ( 𝑠 , 𝑡 ) ≤ 𝐶 𝐿 ( | 𝑠 − 𝑡 | ) | 𝑠 − 𝑡 | 2 + 𝛾 . ( 2 . 7 ) (c) There exist 𝑞 + 1 functions ( 𝑞 ∈ ℕ ) 𝑔 0 , 𝑔 1 , … , 𝑔 𝑞 from ( 0 , 1 ) to ℝ , 𝑞 real numbers 0 < 𝜈 1 < ⋯ < 𝜈 𝑞 and a function 𝜙 ∶ ( 0 , 1 ) → ( 0 , ∞ ) such that (i) if 𝑞 ≥ 1 , then ∀ 0 ≤ 𝑖 ≤ 𝑞 − 1 , 𝑔 𝑖 is Lipschitz on ( 0 , 1 ) ; (ii) 𝑔 𝑞 is bounded on ( 0 , 1 ) ; (iii) one has s u p ℎ ≤ 𝑡 ≤ 1 − ℎ | | | | |  𝛿 ℎ 1 ∘ 𝛿 ℎ 2 𝑅  ( 𝑡 , 𝑡 ) ℎ 2 − 𝛾 𝐿 ( ℎ ) − 𝑔 0 ( 𝑡 ) − 𝑞  𝑖 = 1 𝑔 𝑖 ( 𝑡 ) 𝜙 ( ℎ ) 𝜈 𝑖 | | | | |  = 𝑜 𝜙 ( ℎ ) 𝜈 𝑞  , 𝑎 𝑠 ℎ ⟶ 0 + , ( 2 . 8 ) where the symbol “ ∘ ” denotes the composition of functions and if 𝑞 = 0 , then ∑ 𝑞 𝑖 = 1 𝑔 𝑖 ( 𝑡 ) 𝜙 ( ℎ ) 𝜈 𝑖 = 0 and 𝜙 ( ℎ ) 𝜈 𝑞 = 1 ; else if 𝑞 ≠ 0 , then l i m ℎ → 0 + 𝜙 ( ℎ ) = 0 . (3) If 𝑞 ≠ 0 , we assume that l i m 𝑛 → + ∞ l o g 𝑛 𝑛 𝜙 ( 1 / 𝑛 ) 𝜈 𝑞 = 0 . ( 2 . 9 ) (4) If 𝑋 is not centered, we make the additional assumption l i m 𝑛 → + ∞ 1 𝑛 𝛾 𝐿 ( 1 / 𝑛 ) 𝜙 ( 1 / 𝑛 ) 𝜈 𝑞 = 0 , ( 2 . 1 0 ) where if 𝑞 = 0 , then 𝜙 ( 1 / 𝑛 ) 𝜈 𝑞 = 1 . Then, for all 𝑡 ∈ [ 0 , 1 ] , one has almost surely l i m 𝑛 → + ∞ 𝑛 1 − 𝛾 𝑉 𝐿 ( 1 / 𝑛 ) 𝑛  ( 𝑋 ) = 1 0 𝑔 0 ( 𝑥 ) 𝑑 𝑥 + 𝑞  𝑖 = 1   1 0 𝑔 𝑖  𝜙  1 ( 𝑥 ) 𝑑 𝑥 𝑛  𝜈 𝑖  𝜙  1 + 𝑜 𝑛  𝜈 𝑞  . ( 2 . 1 1 ) Second, let us recall the result of central limit theorem. Theorem 2.3 (Theorem 2 in Begyn [ 22 ]). Assume that the Gaussian process 𝑋 is centered and satisfies the following statements. (1) The covariance function 𝑅 of 𝑋 is continuous in [ 0 , 1 ] 2 . (2) Let 𝑇 = { 0 ≤ 𝑡 ≤ 𝑠 ≤ 1 } . We assume that the derivative 𝜕 4 𝑅 / 𝜕 𝑠 2 𝜕 𝑡 2 exists in ( 0 , 1 ] 2 / { 𝑠 = 𝑡 } and that there exists a continuous function 𝐶 ∶ 𝑇 → ℝ , a real 𝛾 ∈ ( 0 , 2 ) and a positive slowly varying function 𝐿 ∶ ( 0 , 1 ) → ℝ such that ∀ 𝑡 , 𝑠 ∈ 𝑇 , ( 𝑠 − 𝑡 ) 2 + 𝛾 𝐿 𝜕 ( 𝑠 − 𝑡 ) 4 𝑅 𝜕 𝑠 2 𝜕 𝑡 2 ( 𝑠 , 𝑡 ) = 𝐶 ( 𝑠 , 𝑡 ) , ( 2 . 1 2 ) where 𝑇 denotes the interior of 𝑇 (i.e., 𝑇 = { 0 < 𝑠 < 𝑡 < 1 } ). (3) We assume that there exist 𝑞 + 1 functions ( 𝑞 ∈ ℕ ) 𝑔 0 , 𝑔 1 , … , 𝑔 𝑞 from ( 0 , 1 ) to ℝ , 𝑞 real numbers 0 < 𝜈 1 < ⋯ < 𝜈 𝑞 and a function 𝜙 ∶ ( 0 , 1 ) → ( 0 , ∞ ) such that (a) if 𝑞 ≥ 1 , then for all 0 ≤ 𝑖 ≤ 𝑞 − 1 , 𝑔 𝑖 is Lipschitz on ( 0 , 1 ) ; (b) 𝑔 𝑞 is ( 1 / 2 + 𝛼 𝑞 ) -Hölderian on ( 0 , 1 ) with 0 < 𝛼 𝑞 ≤ 1 / 2 ; (c) there exists 𝑡 ∈ ( 0 , 1 ) such that 𝑔 0 ( 𝑡 ) ≠ 0 ; (d) one has l i m ℎ → 0 + 1 √ ℎ ⎛ ⎜ ⎜ ⎝ s u p ℎ ≤ 𝑡 ≤ 1 − ℎ | | | | |  𝛿 ℎ 1 ∘ 𝛿 ℎ 2 𝑅  ( 𝑡 , 𝑡 ) ℎ 2 − 𝛾 𝐿 ( ℎ ) − 𝑔 0 ( 𝑡 ) − 𝑞  𝑖 = 1 𝑔 𝑖 ( 𝑡 ) 𝜙 ( ℎ ) 𝜈 𝑖 | | | | | ⎞ ⎟ ⎟ ⎠ = 0 , ( 2 . 1 3 ) where if 𝑞 = 0 , then ∑ 𝑞 𝑖 = 1 𝑔 𝑖 ( 𝑡 ) 𝜙 ( ℎ ) 𝜈 𝑖 = 0 and where if 𝑞 ≠ 0 , then l i m ℎ → 0 + 𝜙 ( ℎ ) = 0 . (e) there exists a bounded function ̃ 𝑔 ∶ ( 0 , 1 ) → ℝ such that l i m ℎ → 0 + s u p ℎ ≤ 𝑡 ≤ 1 − 2 ℎ | | | |  𝛿 ℎ 1 ∘ 𝛿 ℎ 2 𝑅  ( 𝑡 + ℎ , 𝑡 ) ℎ 2 − 𝛾 | | | | 𝐿 ( ℎ ) − ̃ 𝑔 ( 𝑡 ) = 0 . ( 2 . 1 4 ) Then one has √ 𝑛  𝑛 1 − 𝛾 𝑉 𝐿 ( 1 / 𝑛 ) 𝑛  ( 𝑋 ) − 1 0 𝑔 0 ( 𝑥 ) 𝑑 𝑥 − 𝑞  𝑖 = 1   1 0 𝑔 𝑖  𝜙  1 ( 𝑥 ) 𝑑 𝑥 𝑛  𝜈 𝑖   ⟶ 𝒩 0 , 𝜎 2  ( 2 . 1 5 ) in distribution as 𝑛 tends to infinity where 𝜎 2  = 2 1 0 𝑔 0 ( 𝑥 ) 2  𝑑 𝑥 + 4 1 0 ̃ 𝑔 ( 𝑥 ) 2 ‖ ‖ 𝜌 𝑑 𝑥 + 4 𝛾 ‖ ‖ 2  1 0 𝐶 ( 𝑥 , 𝑥 ) 2 𝑑 𝑥 ( 2 . 1 6 ) and ‖ 𝜌 𝛾 ‖ 2 = ∑ + ∞ 𝑙 = 2 𝜌 𝛾 ( 𝑙 ) 2 with, if 𝛾 ≠ 1 , 𝜌 𝛾 | | | | ( 𝑙 ) = 𝑙 − 2 2 − 𝛾 | | | | − 4 𝑙 − 1 2 − 𝛾 | | 𝑙 | | + 6 2 − 𝛾 | | | | − 4 𝑙 + 1 2 − 𝛾 + | | | | 𝑙 + 2 2 − 𝛾 ; ( 𝛾 − 2 ) ( 𝛾 − 1 ) 𝛾 ( 𝛾 + 1 ) ( 2 . 1 7 ) if 𝛾 = 1 , 𝜌 1 1 ( 𝑙 ) = 2  | | | | | | | | | | | | | | | | | | 𝑙 | | | | 𝑙 | | | | | | | | | | | | | | | | | |  . 𝑙 − 2 l o g 𝑙 − 2 − 4 𝑙 − 1 l o g 𝑙 − 1 + 6 l o g − 4 𝑙 + 1 l o g 𝑙 + 1 + 2 𝑙 + 2 l o g 𝑙 + 2 ( 2 . 1 8 ) 3. Asymptotic Expansion and Central Limit Theorem In the following theorem the almost sure convergence of the second-order quadratic variations 𝑉 𝑛 ( 𝑆 𝐻 ) is proved. Theorem 3.1. For all 𝑡 ∈ [ 0 , 1 ] , one has almost surely l i m 𝑛 → ∞ 𝑛 2 𝐻 − 1 𝑉 𝑛  𝑆 𝐻  = 4 − 2 2 𝐻 . ( 3 . 1 ) Proof. It is clear that the derivative ( 𝜕 4 / 𝜕 𝑠 2 𝜕 𝑡 2 ) 𝑅 ( 𝑠 , 𝑡 ) exists on ( 0 , 1 ] 2 / { 𝑠 = 𝑡 } . Moreover we can check that, for all 𝑠 , 𝑡 ∈ ( 0 , 1 ] 2 / { 𝑠 = 𝑡 } , 𝜕 4 𝜕 𝑠 2 𝜕 𝑡 2  𝑅 ( 𝑠 , 𝑡 ) = 𝐻 ( 2 𝐻 − 1 ) ( 2 𝐻 − 2 ) ( 2 𝐻 − 3 ) − | 𝑠 − 𝑡 | 2 𝐻 − 4 − | 𝑠 + 𝑡 | 2 𝐻 − 4  . ( 3 . 2 ) Therefore the assumption 2(b) in Theorem 2.2 is satisfied with 𝐿 ( 𝐻 ) = 1 and 𝛾 = 2 − 2 𝐻 . For the assumption 2(b) in Theorem 2.2 , standard computations yield  𝛿 ℎ 1 ∘ 𝛿 ℎ 2 𝑅  ( 𝑡 , 𝑡 ) ℎ 𝐻 = 4 − 2 2 𝐻 + 𝜆 𝑡 ( ℎ ) ℎ 2 𝐻 , ( 3 . 3 ) with 𝜆 𝑡 1 ( ℎ ) = ℎ 2 𝐻  − 2 2 𝐻 − 1 ( 𝑡 + ℎ ) 2 𝐻 − 3 ⋅ 2 2 𝐻 𝑡 2 𝐻 − 2 2 𝐻 − 1 ( 𝑡 − ℎ ) 2 𝐻 + 2 ( 2 𝑡 + ℎ ) 2 𝐻 + 2 ( 2 𝑡 − ℎ ) 2 𝐻  , ( 3 . 4 ) and we can check 𝜆 𝑡 ( 0 ) = 𝜆  𝑡 ( 0 ) = 𝜆 𝑡   ( 0 ) = 𝜆 𝑡 ( 3 ) ( 0 ) = 0 . So that Taylor formula yields 𝜆 𝑡  ( ℎ ) = ℎ 0 ( ℎ − 𝑥 ) 3 𝜆 3 ! 𝑡 ( 4 ) ( 𝑥 ) 𝑑 𝑥 , ∀ ℎ ≤ 𝑡 ≤ 1 − ℎ . ( 3 . 5 ) Therefore, we have s u p ℎ ≤ 𝑡 ≤ 1 − ℎ s u p 0 ≤ 𝑥 ≤ ℎ | | 𝜆 𝑡 ( 4 ) | | ( 𝑥 ) = 𝒪 ( 1 ) , a s ℎ ⟶ 0 + , ( 3 . 6 ) which yields s u p ℎ ≤ 𝑡 ≤ 1 − ℎ | | | |  𝛿 ℎ 1 ∘ 𝛿 ℎ 2 𝑅  ( 𝑡 , 𝑡 ) ℎ 𝐻 −  4 − 2 2 𝐻  | | | |  ℎ = 𝒪 4 − 2 𝐻  , a s ℎ ⟶ 0 + . ( 3 . 7 ) Therefore, the assumption 2(c) in Theorem 2.2 is fulfilled with 𝑔 0 ( 𝑡 ) = 4 − 2 2 𝐻 . ( 3 . 8 ) Consequently, we can apply Theorem 2.2 to 𝑉 𝑛 ( 𝑆 𝐻 ) and obtain the desired result. Next we study the weak convergence. Theorem 3.2. One has the following weak convergence √ 𝑛  𝑛 2 𝐻 − 1 𝑉 𝑛  𝑆 𝐻  −  4 − 2 2 𝐻 ⟶   ( ℒ ) 𝑁  0 , 𝜎 2 𝐻  , a s 𝑛 ⟶ ∞ , ( 3 . 9 ) where 𝜎 2 𝐻  = 2 4 − 2 2 𝐻  2 +  2 2 𝐻 + 2 − 7 − 3 2 𝐻  2 + [ ] 2 𝐻 ( 2 𝐻 − 1 ) ( 2 𝐻 − 2 ) ( 2 𝐻 − 3 ) 2 ‖ ‖ 𝜌 2 − 2 𝐻 ‖ ‖ 2 , ‖ ‖ 𝜌 2 − 2 𝐻 ‖ ‖ 2 = ∞  𝑙 = 2  − | | | | 𝑙 − 2 2 𝐻 | | | | + 4 𝑙 − 1 2 𝐻 | | 𝑙 | | − 6 2 𝐻 | | | | + 4 𝑙 + 1 2 𝐻 − | | | | 𝑙 + 2 2 𝐻  2 𝐻 ( 2 𝐻 − 1 ) ( 2 𝐻 − 2 ) ( 2 𝐻 − 3 ) 2 . ( 3 . 1 0 ) Proof. We apply Theorem 2.3 to 𝑉 𝑛 ( 𝑆 𝐻 ) . As in the proof of Theorem 3.1 , we need only to show that the assumptions 2 and 3 in Theorem 2.3 are satisfied. For assumption 2, the previous computation showed that, for all 𝑠 , 𝑡 ∈ ( 0 , 1 ] 2 / { 𝑠 = 𝑡 } , 𝜕 4 𝜕 𝑠 2 𝜕 𝑡 2  𝑅 ( 𝑠 , 𝑡 ) = 𝐻 ( 2 𝐻 − 1 ) ( 2 𝐻 − 2 ) ( 2 𝐻 − 3 ) − | 𝑠 − 𝑡 | 2 𝐻 − 4 − | 𝑠 + 𝑡 | 2 𝐻 − 4  . ( 3 . 1 1 ) Therefore ( 𝑠 − 𝑡 ) 4 − 2 𝐻 𝜕 4 𝜕 𝑠 2 𝜕 𝑡 2  𝑅 ( 𝑠 , 𝑡 ) = − 𝐻 ( 2 𝐻 − 1 ) ( 2 𝐻 − 2 ) ⋅ ( 2 𝐻 − 3 ) 1 + ( 𝑠 − 𝑡 ) 4 − 2 𝐻 ( 𝑠 + 𝑡 ) 2 𝐻 − 4  ∶ ≡ 𝐶 ( 𝑠 , 𝑡 ) . ( 3 . 1 2 ) This means that the assumption 2 in Theorem 2.3 is satisfied with 𝐿 ( 𝐻 ) = 1 , 𝛾 = 2 − 2 𝐻 and 𝐶 ( 𝑠 , 𝑡 ) defined by ( 3.12 ). For assumption 3 in Theorem 2.3 , the expression ( 3.7 ) shows that the assumption 3 in Theorem 2.3 is fulfilled with 𝑞 = 0 , 𝑔 0 ( 𝑡 ) = 4 − 2 2 𝐻 and 𝛼 0 = 1 / 2 . Moreover, one can check that  𝛿 ℎ 1 ∘ 𝛿 ℎ 2 𝑅  ( 𝑡 , 𝑡 + ℎ ) ℎ 𝐻 = 1 2  2 2 𝐻 + 2 − 3 2 𝐻  + 𝜉 − 7 𝑡 ( ℎ ) ℎ 2 𝐻 . ( 3 . 1 3 ) Using the same arguments as those used for 𝜆 𝑡 ( ℎ ) in the previous proof, we obtain s u p ℎ ≤ 𝑡 ≤ 1 − ℎ | | 𝜉 𝑡 | |  ℎ ( ℎ ) = 𝒪 4  , a s ℎ ⟶ 0 + . ( 3 . 1 4 ) This shows that the assumption 3(e) in Theorem 2.3 is satisfied with 1 ̃ 𝑔 ( 𝑡 ) = 2  2 2 𝐻 + 2 − 3 2 𝐻  . − 7 ( 3 . 1 5 ) Consequently, we can apply Theorem 2.3 to 𝑉 𝑛 ( 𝑆 𝐻 ) to obtain the desired result. 4. Confidence Intervals Let 𝑆 𝐻 is a subfractional Brownian motion with unknown Hurst parameter 𝐻 ∈ ( 0 , 𝐻 ∗ ] , with 𝐻 ∗ < 1 / 2 known. The following result is the main finding of the present note. Theorem 4.1. For 𝑉 𝑛 ( 𝑆 𝐻 ) defined in ( 1.3 ), fix 𝑛 ≥ 0 and a real number 𝑎 such that 0 < 𝑎 < ( 4 − 2 2 𝐻 ∗ ) √ 𝑛 . For 𝑥 ∈ ( 0 , 1 ) , set 𝑔 𝑛 ( 𝑥 ) = 𝑥 − l o g ( 4 − 2 2 𝑥 ) / 2 l o g 𝑛 . Then, with probability at least ⎡ ⎢ ⎢ ⎣ ⎛ ⎜ ⎜ ⎝ − 𝑎 𝜑 ( 𝑎 ) = 1 − 2 e x p 2 4 𝐶 𝐻 ∗  √ 𝑎 / 𝑛 + 3 + 𝐶 𝐻 ∗  ⎞ ⎟ ⎟ ⎠ ⎤ ⎥ ⎥ ⎦ / 𝑛 + , ( 4 . 1 ) where 𝐶 𝐻 ∗ is a positive constant depending only on 𝐻 ∗ and [ ⋅ ] + stands for the positive part function; the unknown quantity 𝑔 𝑛 ( 𝐻 ) belongs to the following confidence intervals 𝐼 𝑛 =  𝐼 𝑙 ( 𝑛 ) , 𝐼 𝑟  = ⎡ ⎢ ⎢ ⎣ 1 ( 𝑛 ) 2 − l o g 𝑉 𝑛 +  √ 2 l o g 𝑛 l o g 1 − 𝑎 / 𝑛  4 − 2 2 𝐻 ∗   , 1 2 l o g 𝑛 2 − l o g 𝑉 𝑛 +  √ 2 l o g 𝑛 l o g 1 + 𝑎 / 𝑛  4 − 2 2 𝐻 ∗   ⎤ ⎥ ⎥ ⎦ . 2 l o g 𝑛 ( 4 . 2 ) Proof. The idea used here is essentially due to Breton et al. [ 19 ]. Define 𝑋 𝑛 = { 𝑋 𝑛 , 𝑘 ; 𝑘 = 0 , 1 , … , 𝑛 − 1 } , where 𝑋 𝑛 , 𝑘 = 𝑆 𝐻 ( 𝑘 + 2 ) / 𝑛 + 𝑆 𝐻 𝑘 / 𝑛 − 2 𝑆 𝐻 ( 𝑘 + 1 ) / 𝑛 . ( 4 . 3 ) One can prove by standard computations that the covariance structure of Gaussian family 𝑋 𝑛 , 𝑘 is described by the relation 𝐸  𝑋 𝑛 , 𝑘 𝑋 𝑛 , 𝑙  = 1 𝑛 2 𝐻 𝜌 𝐻 ( 𝑘 , 𝑙 ) , ( 4 . 4 ) where 𝜌 𝐻 1 ( 𝑘 , 𝑙 ) = − 2 ( 𝑘 + 𝑙 + 4 ) 2 𝐻 + 2 ( 𝑘 + 𝑙 + 3 ) 2 𝐻 − 3 ( 𝑘 + 𝑙 + 2 ) 2 𝐻 + 2 ( 𝑘 + 𝑙 + 1 ) 2 𝐻 − 1 2 ( 𝑘 + 𝑙 ) 2 𝐻 − 3 ( 𝑘 − 𝑙 ) 2 𝐻 + 2 ( 𝑘 − 𝑙 + 1 ) 2 𝐻 − 1 2 ( 𝑘 − 𝑙 + 2 ) 2 𝐻 + 2 ( 𝑘 − 𝑙 − 1 ) 2 𝐻 − 1 2 ( 𝑘 − 𝑙 − 2 ) 2 𝐻 . ( 4 . 5 ) Now let 𝑍 𝑛 = 𝑛 2 𝐻 − 1 / 2 𝑉 𝑛 − √ 𝑛 ( 4 − 2 2 𝐻 ) , where 𝑉 𝑛 is defined in ( 1.3 ). It is easy to see that 𝑍 𝑛 = 𝑄 1  𝑛 2 𝐻 − 1 / 2 , 𝑋 𝑛  + 1 √ 𝑛 𝑛 − 1  𝑘 = 0 𝑎 𝑘 , 𝐻 , ( 4 . 6 ) where 𝑎 𝑘 , 𝐻 1 = − 2 ( 2 𝑘 + 4 ) 2 𝐻 + 2 ( 2 𝑘 + 3 ) 2 𝐻 − 3 ( 2 𝑘 + 2 ) 2 𝐻 + 2 ( 2 𝑘 + 1 ) 2 𝐻 − 1 2 ( 2 𝑘 ) 2 𝐻 . ( 4 . 7 ) On the other hand 𝑄 2  𝑛 2 𝐻 − 1 / 2 , 𝑋 𝑛  = 2 𝑛 4 𝐻 − 1 𝑛 − 1  𝑘 , 𝑙 = 0 𝑋 𝑛 , 𝑘 𝑋 𝑛 , 𝑙 𝜌 𝐻 ( 𝑘 , 𝑙 ) ≤ 2 𝑛 2 𝐻 − 1 𝑛 − 1  𝑘 , 𝑙 = 0 | | 𝑋 𝑛 , 𝑘 ‖ ‖ 𝑋 𝑛 , 𝑙 ‖ ‖ 𝜌 𝐻 | | ( 𝑘 , 𝑙 ) ≤ 𝑛 2 𝐻 − 1 𝑛 − 1  𝑘 , 𝑙 = 0  | | 𝑋 𝑛 , 𝑘 | | 2 + | | 𝑋 𝑛 , 𝑙 | | 2  | | 𝜌 𝐻 | | ( 𝑘 , 𝑙 ) = 2 𝑛 2 𝐻 − 1 𝑛 − 1  𝑘 , 𝑙 = 0 | | 𝑋 𝑛 , 𝑘 | | 2 | | 𝜌 𝐻 | | ( 𝑘 , 𝑙 ) ≤ 2 𝑛 2 𝐻 − 1 𝑛 − 1  𝑘 = 0 | | 𝑋 𝑛 , 𝑘 | | 2   𝑖 , 𝑗 ∈ ℤ | | 𝜌 𝐻 | |  ≤ 2 ( 𝑖 , 𝑗 ) √ 𝑛   𝑖 , 𝑗 ∈ ℤ | | 𝜌 𝐻 ( | |   𝑍 𝑖 , 𝑗 ) 𝑛 +  4 − 2 2 𝐻  √ 𝑛  ≤ 2 √ 𝑛   𝑖 , 𝑗 ∈ ℤ | | 𝜌 𝐻 | |   𝑍 ( 𝑖 , 𝑗 ) 𝑛 √ + 3 𝑛  = 2 √ 𝑛   𝑖 , 𝑗 ∈ ℤ | | 𝜌 𝐻 | | 𝑄 ( 𝑖 , 𝑗 )   1  𝑛 2 𝐻 − 1 / 2 , 𝑋 𝑛  + 1 √ 𝑛 𝑛 − 1  𝑘 = 0 𝑎 𝑘 , 𝐻 √ + 3 𝑛  ≤ 2 √ 𝑛   𝑖 , 𝑗 ∈ ℤ | | 𝜌 𝐻 | | 𝑄 ( 𝑖 , 𝑗 )   1  𝑛 2 𝐻 − 1 / 2 , 𝑋 𝑛  + 1 √ 𝑛 𝑛 − 1  𝑘 = 0 | | 𝑎 𝑘 , 𝐻 | | √ + 3 𝑛  = 𝛼 𝑛 𝑄 1  𝑛 2 𝐻 − 1 / 2 , 𝑋 𝑛  + 𝛽 𝑛 , ( 4 . 8 ) with 𝛼 𝑛 = 2 √ 𝑛  𝑖 , 𝑗 ∈ ℤ | | 𝜌 𝐻 | |  ( 𝑖 , 𝑗 ) , 𝛽 = 2 𝑖 , 𝑗 ∈ ℤ | | 𝜌 𝐻 | |  1 ( 𝑖 , 𝑗 ) 3 + 𝑛 𝑛 − 1  𝑘 = 0 | | 𝑎 𝑘 , 𝐻 | |  . ( 4 . 9 ) Since 𝑍 𝑛 ≠ 0 , Theorem 2.1 yields 𝑃  | | 𝑍 𝑛 | |  ⎛ ⎜ ⎜ ⎝ − 𝑎 ≥ 𝑎 ≤ 2 e x p 2 4 ∑ 𝑖 , 𝑗 ∈ ℤ | | 𝜌 𝐻 | |  √ ( 𝑖 , 𝑗 ) 𝑎 / ∑ 𝑛 + 3 + ( 1 / 𝑛 ) 𝑛 − 1 𝑘 = 0 | | 𝑎 𝑘 , 𝐻 | |  ⎞ ⎟ ⎟ ⎠ . ( 4 . 1 0 ) Now let us find bounds on ∑ 𝑖 , 𝑗 ∈ ℤ | 𝜌 𝐻 ( 𝑖 , 𝑗 ) | . Using ( 1 + 𝑥 ) 𝛼 = 1 + ∞  𝑘 = 1 𝛼 ( 𝛼 − 1 ) ⋯ ( 𝛼 − 𝑘 + 1 ) 𝑥 𝑘 ! 𝑘 , f o r − 1 < 𝑥 < 1 . ( 4 . 1 1 ) We denote by 𝜌 𝐻 ( 𝑖 , 𝑗 ) = 𝜌 𝐻 , 1 ( 𝑖 , 𝑗 ) + 𝜌 𝐻 , 2 ( 𝑖 , 𝑗 ) , ( 4 . 1 2 ) where 𝜌 𝐻 , 1 1 ( 𝑖 , 𝑗 ) = − 2 ( 𝑘 + 𝑙 + 4 ) 2 𝐻 + 2 ( 𝑘 + 𝑙 + 3 ) 2 𝐻 − 3 ( 𝑘 + 𝑙 + 2 ) 2 𝐻 + 2 ( 𝑘 + 𝑙 + 1 ) 2 𝐻 − 1 2 ( 𝑘 + 𝑙 ) 2 𝐻 , 𝜌 𝐻 , 2 ( 𝑖 , 𝑗 ) = − 3 ( 𝑘 − 𝑙 ) 2 𝐻 + 2 ( 𝑘 − 𝑙 + 1 ) 2 𝐻 − 1 2 ( 𝑘 − 𝑙 + 2 ) 2 𝐻 + 2 ( 𝑘 − 𝑙 − 1 ) 2 𝐻 − 1 2 ( 𝑘 − 𝑙 − 2 ) 2 𝐻 . ( 4 . 1 3 ) The second term 𝜌 𝐻 , 2 ( 𝑖 , 𝑗 ) has been bounded by Breton et al. [ 19 ]. They proved that  𝑖 , 𝑗 ∈ ℤ 𝜌 𝐻 , 2 ( 𝑖 , 𝑗 ) ≤ 7 1 4 . ( 4 . 1 4 ) Now let us bound the first term 𝜌 𝐻 , 1 ( 𝑖 , 𝑗 ) . We denote by 𝜌 𝐻 , 1 ( 𝑖 , 𝑗 ) = 𝜌 𝐻 , 1 ( 𝑟 ) , 𝑟 = 𝑖 + 𝑗 . ( 4 . 1 5 ) We can write for any 𝑟 ≥ 5 , 𝜌 𝐻 , 1 𝑟 ( 𝑟 ) = 2 𝐻 2  −  4 1 + 𝑟  2 𝐻  3 + 4 1 + 𝑟  2 𝐻  2 − 6 1 + 𝑟  2 𝐻  1 + 4 1 + 𝑟  2 𝐻  = 𝑟 − 1 2 𝐻 2  ∞  𝑘 = 1 2 𝐻 ( 2 𝐻 − 1 ) ⋯ ( 2 𝐻 − 𝑘 + 1 )  𝑘 ! − 4 𝑘 + 4 ⋅ 3 𝑘 − 6 ⋅ 2 𝑘  𝑟 + 4 − 𝑘  . ( 4 . 1 6 ) Note that the sign of 2 𝐻 ( 2 𝐻 − 1 ) ⋯ ( 2 𝐻 − 𝑘 + 1 ) is the same as that of ( − 1 ) 𝑘 − 1 , and | | | | | | | | ⋯ | | | | 2 𝐻 ( 2 𝐻 − 1 ) ⋯ ( 2 𝐻 − 𝑘 + 1 ) = 2 𝐻 ( 2 𝐻 − 1 ) 2 𝐻 − 𝑘 + 1 < 2 ⋅ 1 ⋅ 2 ⋯ ( 𝑘 − 1 ) = 2 ( 𝑘 − 1 ) ! . ( 4 . 1 7 ) Hence we can write, for any 𝑟 ≥ 5 , | | 𝜌 𝐻 , 1 ( | | ≤ 𝑟 𝑟 ) 2 𝐻 2 ∞  𝑘 = 1 | | | 1 𝑘  − 4 𝑘 + 4 ⋅ 3 𝑘 − 6 ⋅ 2 𝑘  𝑟 + 4 − 𝑘 | | | = 𝑟 2 𝐻 2 ∞  𝑘 = 1 1 𝑘 | | | | −  4 𝑟  𝑘  3 + 4 ⋅ 𝑟  𝑘  2 − 6 ⋅ 𝑟  𝑘  1 + 4 𝑟  𝑘 | | | | = 𝑟 2 𝐻 2 | | |  4 l o g 1 − 𝑟   3 − 4 l o g 1 − 𝑟   2 + 6 l o g 1 − 𝑟   1 − 4 l o g 1 − 𝑟  | | | ≤ 𝑟 2 𝐻 2  | | |  4 l o g 1 − 𝑟   1 − 4 l o g 1 − 𝑟  | | | + | | |  3 − 4 l o g 1 − 𝑟   2 + 6 l o g 1 − 𝑟  | | |  . ( 4 . 1 8 ) One can easily check that | l o g ( 1 − 4 𝑥 ) − 4 l o g ( 1 − 𝑥 ) | ≤ ( 2 4 3 / 2 0 ) 𝑥 2 , if 0 ≤ 𝑥 ≤ 1 / 5 . And moreover, | | | | ≤ | | | | − 4 l o g ( 1 − 3 𝑥 ) + 6 l o g ( 1 − 2 𝑥 ) l o g ( 1 − 4 𝑥 ) − 4 l o g ( 1 − 𝑥 ) . ( 4 . 1 9 ) Then we have, for any 𝑟 ≥ 5 , | | 𝜌 𝐻 , 1 | | ≤ ( 𝑟 ) 2 4 3 𝑟 2 0 − 3 / 2 + 𝐻 ∗ . ( 4 . 2 0 ) Consequently, we get  𝑟 ∈ ℤ | | 𝜌 𝐻 , 1 ( | | ≤ | | 𝜌 𝑟 ) 𝐻 , 1 ( | | + | | 𝜌 0 ) 𝐻 , 1 ( | | + | | 𝜌 1 ) 𝐻 , 1 ( | | + | | 𝜌 2 ) 𝐻 , 1 ( | | + | | 𝜌 3 ) 𝐻 , 1 ( | | +  4 ) 𝑟 ≥ 5 | | 𝜌 𝐻 , 1 ( | | = 1 𝑟 ) 2 | | − 4 2 𝐻 + 4 ⋅ 3 2 𝐻 − 6 ⋅ 2 2 𝐻 | | + 1 + 4 2 | | − 5 2 𝐻 + 4 ⋅ 4 2 𝐻 − 6 ⋅ 3 2 𝐻 + 4 ⋅ 2 2 𝐻 | | + 1 − 1 2 | | − 6 2 𝐻 + 4 ⋅ 5 2 𝐻 − 6 ⋅ 4 2 𝐻 + 4 ⋅ 3 2 𝐻 − 2 2 𝐻 | | + 1 2 | | − 7 2 𝐻 + 4 ⋅ 6 2 𝐻 − 6 ⋅ 5 2 𝐻 + 4 ⋅ 4 2 𝐻 − 3 2 𝐻 | | + 1 2 | | − 8 2 𝐻 + 4 ⋅ 7 2 𝐻 − 6 ⋅ 6 2 𝐻 + 4 ⋅ 5 2 𝐻 − 4 2 𝐻 | | +  𝑟 ≥ 5 | | 𝜌 𝐻 , 1 | | ( 𝑟 ) ≤ 𝐶 1 + 𝐶 𝐻 ∗ ∶ ≡ 𝐶 𝐻 ∗ < ∞ , ( 4 . 2 1 ) and the positive constant 𝐶 𝐻 ∗ does not depend on the unknown parameter 𝐻 . Putting this bound in ( 4.10 ) yields 𝑃  | | 𝑍 𝑛 | |  ⎛ ⎜ ⎜ ⎝ − 𝑎 ≥ 𝑎 ≤ 2 e x p 2 4 𝐶 𝐻 ∗  √ 𝑎 / 𝑛 + 3 + 𝐶 𝐻 ∗  ⎞ ⎟ ⎟ ⎠ . / 𝑛 ( 4 . 2 2 ) Now we can construct the confidence interval for 𝑔 𝑛 ( 𝐻 ) = 𝐻 − l o g ( 4 − 2 2 𝐻 ) / 2 l o g 𝑛 . First observe that 𝑍 𝑛 = 𝑛 2 𝐻 − 1 / 2 𝑉 𝑛 − √ 𝑛 ( 4 − 2 2 𝐻 ) . Using the assumption 𝐻 ≤ 𝐻 ∗ < 1 / 2 on the one hand and ( 4.22 ) on the other hand, we get 𝑃 ⎛ ⎜ ⎜ ⎝ 1 2 − l o g 𝑉 𝑛 +   2 l o g 𝑛 l o g 1 − 𝑎 / 4 − 2 2 𝐻 ∗  √ 𝑛  2 l o g 𝑛 ≤ 𝑔 𝑛 1 ( 𝐻 ) ≤ 2 − l o g 𝑉 𝑛 +   2 l o g 𝑛 l o g 1 + 𝑎 / 4 − 2 2 𝐻 ∗  √ 𝑛  ⎞ ⎟ ⎟ ⎠ ⎛ ⎜ ⎜ ⎝ 1 2 l o g 𝑛 ≥ 𝑃 2 − l o g 𝑉 𝑛 +   2 l o g 𝑛 l o g 1 − 𝑎 / 4 − 2 2 𝐻  √ 𝑛   2 l o g 𝑛 ≤ 𝐻 − l o g 4 − 2 2 𝐻  ≤ 1 2 l o g 𝑛 2 − l o g 𝑉 𝑛 +   2 l o g 𝑛 l o g 1 + 𝑎 / 4 − 2 2 𝐻  √ 𝑛  ⎞ ⎟ ⎟ ⎠ ⎛ ⎜ ⎜ ⎝ 1 2 l o g 𝑛 = 𝑃 4 − l o g 𝑉 𝑛 +   2 l o g 𝑛 l o g 4 − 2 2 𝐻  √  𝑛 − 𝑎 1 2 l o g 𝑛 ≤ 𝐻 ≤ 4 − l o g 𝑉 𝑛 +   2 l o g 𝑛 l o g 4 − 2 2 𝐻  √  𝑛 + 𝑎 ⎞ ⎟ ⎟ ⎠  | | 𝑍 2 l o g 𝑛 = 𝑃 𝑛 | |  ⎛ ⎜ ⎜ ⎝ − 𝑎 ≤ 𝑎 ≥ 1 − 2 e x p 2 4 𝐶 𝐻 ∗  √ 𝑎 / 𝑛 + 3 + 𝐶 𝐻 ∗  ⎞ ⎟ ⎟ ⎠ , / 𝑛 ( 4 . 2 3 ) where ∑ 𝑖 , 𝑗 ∈ ℤ | 𝜌 ( 𝑖 , 𝑗 ) | ≤ 𝐶 𝐻 ∗ and the positive constant 𝐶 𝐻 ∗ does not depend on the unknown parameter 𝐻 . This is the desired result. Acknowledgments The authors want to thank the academic editor and anonymous referee whose remarks and suggestions greatly improved the presentation of this paper. The project is sponsored by NSFC (10871041), NSFC (81001288), NSRC (10023). Innovation Program of Shanghai Municipal Education Commission (12ZZ063) and NSF of Jiangsu Educational Committee (11KJD11002). <h4>References</h4> E. Alòs, O. Mazet, and D. Nualart, “ Stochastic calculus with respect to Gaussian processes ,” The Annals of Probability , vol. 29, no. 2, pp. 766–801, 2001. D. Nualart, The Malliavin Calculus and Related Topics , Probability and its Applications, Springer, New York, NY, USA, 2nd edition, 2006. T. Bojdecki, L. G. Gorostiza, and A. Talarczyk, “ Sub-fractional Brownian motion and its relation to occupation times ,” Statistics & Probability Letters , vol. 69, no. 4, pp. 405–419, 2004. T. Bojdecki, L. G. Gorostiza, and A. Talarczyk, “Some extensions of fractional Brownian motion and sub-fractional Brownian motion related to particle systems,” Electronic Communications in Probability , vol. 12, pp. 161–172, 2007. J. Liu and L. Yan, “ Remarks on asymptotic behavior of weighted quadratic variation of subfractional Brownian motion ,” Journal of the Korean Statistical Society . In press. J. Liu and L. Yan, “A note on some approximation schemes associated to stochastic differential equations driven by sub-fractional Brownian motion,” Preprint. J. Liu, “Variations and estimators for self-similarity parameter of sub-fractional Brownian motion via Malliavin calculus,” Preprint. C. Tudor, “ Some properties of the sub-fractional Brownian motion ,” Stochastics , vol. 79, no. 5, pp. 431–448, 2007. C. Tudor, “ Multiple sub-fractional integrals and some approximations ,” Applicable Analysis , vol. 87, no. 3, pp. 311–323, 2008. C. Tudor, “ Inner product spaces of integrands associated to subfractional Brownian motion ,” Statistics & Probability Letters , vol. 78, no. 14, pp. 2201–2209, 2008. C. Tudor, “Some aspects of stochastic calculus for the sub-fractional Brownian motion,” Analele Universit\u a\c tii Bucureşti. Matematic\u a , vol. 57, no. 2, pp. 199–230, 2008. C. Tudor, “ On the Wiener integral with respect to a sub-fractional Brownian motion on an interval ,” Journal of Mathematical Analysis and Applications , vol. 351, no. 1, pp. 456–468, 2009. L. Yan and G. Shen, “ On the collision local time of sub-fractional Brownian motions ,” Statistics & Probability Letters , vol. 80, no. 5-6, pp. 296–308, 2010. L. Yan, G. Shen, and K. He, “Itô's formula for a sub-fractional Brownian motion,” Communications on Stochastic Analysis , vol. 5, no. 1, pp. 135–159, 2011. A. Chronopoulou, C. A. Tudor, and F. G. Viens, “ Application of Malliavin calculus to long-memory parameter estimation for non-Gaussian processes ,” Comptes Rendus Mathématique , vol. 347, no. 11-12, pp. 663–666, 2009. A. Chronopoulou and F. Viens, “Hurst estimation for self-similar processes with long memory,” Interdisciplinary Mathematical Sciences , vol. 8, pp. 91–117, 2009. C. A. Tudor and F. Viens, “Variations of the fractional Brownian motion via Malliavin calculus,” submitted to Australian Journal of Mathematics . C. A. Tudor and F. G. Viens, “ Variations and estimators for self-similarity parameters via Malliavin calculus ,” The Annals of Probability , vol. 37, no. 6, pp. 2093–2134, 2009. J. C. Breton, I. Nourdin, and G. Peccati, “ Exact confidence intervals for the Hurst parameter of a fractional Brownian motion ,” Electronic Journal of Statistics , vol. 3, pp. 416–425, 2009. I. Nourdin and F. G. Viens, “Density formula and concentration inequalities with Malliavin calculus,” Electronic Journal of Probability , vol. 14, pp. 2287–2309, 2009. I. Nourdin and G. Peccati, “ Stein's method on Wiener chaos ,” Probability Theory and Related Fields , vol. 145, no. 1-2, pp. 75–118, 2009. A. Begyn, “ Asymptotic expansion and central limit theorem for quadratic variations of Gaussian processes ,” Bernoulli , vol. 13, no. 3, pp. 712–753, 2007. //

/lp/hindawi-publishing-corporation/remarks-on-confidence-intervals-for-self-similarity-parameter-of-a-Z8dUabDCWq