Professional Documents
Culture Documents
263
(e-mail: banuelos@math.purdue.edu)
2 Institute of Mathematics, Wrocaw University of Technology, 50-370 Wrocaw, Poland
(e-mail: bogdan@im.pwr.wroc.pl)
(Received: 9 October 2002; accepted: 5 September 2003)
Abstract. We study exponents of integrability of the first exit time from generalized cones for
conditioned rotation invariant stable Lvy processes. Along the way, we introduce the spherical
fractional Laplacian and derive some of its spectral properties.
Mathematics Subject Classifications (2000): Primary: 31B05, 60J45.
Key words: symmetric stable process, -harmonic function, cone, conditional process, exit time.
1. Introduction
For x Rd \{0}, we denote by (x) the angle between x and the point (0, . . . , 0, 1).
The right circular cone of angle (0, ) is the domain = {x Rd \{0} :
(x) < }. The exact moments of integrability of the first exit time, , for
both Brownian motion and conditioned Brownian motion from have been extensively studied in the literature. These investigations began with the work of
Burkholder [14] who showed that there exists a constant p(, d) such that for
p
any x , Ex ( ) < if and only if p < p(, d). This critical exponent, as it
was shown by Burkholder, can be expressed in terms of zeros of confluent hypergeometric function. Extensions of the result were given by DeBlassie [20], Davies
and Zang [19], and Bauelos and Smits [2]. In particular, the results in [19] and [2]
provide the analogue of Burkholders result for conditioned Brownian motion. In
[21], DeBlassie obtained a counterpart of Burkholders result for symmetric stable
processes in R2 . DeBlassies result was extended to all dimensions by MndezHernndez [30]. In [26], Kulczycki gave results on the asymptotics of the critical
exponent in right circular cones of decreasing aperture.
The purpose of the present paper is to obtain an analogue of the results in [19]
and [2] for conditioned symmetric stable processes in generalized cones and to
extend the results in [21] and [30] for the unconditioned processes to more general
The first author was supported in part by NSF grant # 9700585-DMS.
The second author was supported in part by KBN and by RTN contract HPRN-CT-2001-00273-
HARP.
264
cones. Our method is motivated by, but different then, the method in [19]. The key
step in our proof is to identify, for generalized cones, the Martin kernel with pole at
infinity as a homogeneous function of degree [0, ). Once this is done we use
the scaling of the stable process and homogeneity of the kernel, together with tools
related to the boundary Harnack principle of Song and Wu [33], to estimate the
distribution of the exit time of the process from the cone. The exponent depends
on the geometry of the cone and relates the considered problem to the asymptotics
of harmonic functions at the vertex of the cone, see [7] for boundary points of
Lipschitz domains. We use to identify the critical moments of integrability of the
exit time of the stable processes and the conditioned stable process from the cone as
/ and (d + 2)/, respectively, see Theorem 4.1 below. Our method should
apply to other processes with scaling such as the censored processes studied in
[11].
The paper is organized as follows. In Section 2 we recall the basic properties
of the symmetric -stable processes and the definition of -harmonic functions.
In Section 3 we recall the boundary Harnack principle of [33] and state Theorem 3.2 and Theorem 3.3 which identify the Martin kernels at infinity and at
0 as homogeneous functions of degree and d , respectively. We also
give some explicit examples of cones where we identify these exponents in terms
of the parameter . Theorems 3.2 and 3.3 are proved in Section 6. In Section 4
we obtain the above-mentioned critical moments of integrability of the lifetime of
the stable processes in the generalized cones. In Section 5 we study a spherical
fractional Laplacian, which is related in a natural way to the symmetric stable
process. In the classical case of the Brownian motion, the spherical Laplacian plays
a crucial role in understanding the moments of integrability of the corresponding
process in cones. Indeed, the exponent of integration is obtained from the Dirichlet
eigenvalues of the spherical Laplacian on the set of the sphere which generates the
cones, see for example [2]. While in the current case we were not able to obtain this
precise relation to the exponent of integrability, nevertheless the spherical operator
does provide some additional information. We also believe this operator may be
useful in other settings and a more detailed study of its spectral properties will be
of interest. In a sense, we give a polar coordinate decomposition of the fractional
Laplacian, which as far as we know has not been given before.
2. Preliminaries
We begin by reviewing the notation used in this paper. By | | we denote the
Euclidean norm in Rd . For x Rd , r > 0 and a set A Rd we let B(x, r) =
{y Rd : |x y| < r} and dist(A, x) = inf{|x y| : y A}. A is the closure,
and Ac is the complement of A. We always assume Borel measurability of the
considered sets and functions. The notation c = c(, , . . . , ) means that c is a
constant depending only on , , . . . , . Constants will always be (strictly) positive
and finite. Throughout the paper we use the convention that 0 = 0.
265
For the rest of the paper, unless stated otherwise, is a number in (0, 2) and
d = 1, 2, . . . . By (Xt , Px ) we denote the standard [5] rotation invariant -stable
Rd -valued Lvy process (that is, homogeneous, with independent increments), with
index of stability and characteristic function
Ex ei(Xt x) = et | | ,
x Rd , Rd , t 0.
and it is a strong Markov processes with respect to the standard filtration. For this
and other basic properties, we refer the reader to [5].
/ U }, the first exit time
For an open set U Rd , we put U = inf{t 0; Xt
of U . Given x Rd , the Px distribution of XU , which is given by
Px {XU A, U < +},
for every Borel set A, is a subprobability measure on U c (probability measure if U
is bounded) called the -harmonic measure.
The scaling property of Xt plays a key role in this paper. Namely, for r > 0
we have that for every x Rd the Px distribution of Xt is the same as the Prx
distribution of r 1 Xr t . In particular, the Px distribution of U is the same as the Prx
distribution of r rU . In short, rU = r U in distribution.
When r > 0, |x| < r and B = B(0, r) Rd , the corresponding -harmonic
measure has the density function Pr (x, ) (the Poisson kernel) given by
2
2 /2
d r |x|
|y x|d if |y| > r,
(1)
Pr (x, y) = C
|y|2 r 2
with Cd = (d/2) d/21 sin( /2), and 0 otherwise. The formula for the Poisson kernel for the exterior of the ball {y Rd : |y x| > r} is similar. Namely,
for d and |x| > r we have
2
2 /2
d |x| r
x U,
(3)
266
for every bounded open set U with closure contained in D. It is called regular
-harmonic in D if (3) holds for U = D.
In (3) we always assume that the expectation is absolutely convergent. If D is
unbounded then by the usual convention Ex u(XD ) = Ex [u(XD ); D < ]. By
the strong Markov property a regular -harmonic function is -harmonic. The
converse is not generally true as shown in [8, 16]. An alternative description of
-harmonic functions as those annihilating the fractional Laplacian
f (y) f (x) f (x) (y x)1{|yx|<1}
/2
dy
f (x) = Ad,
|y x|d+
Rd
f (y) f (x)
dy,
= Ad, lim+
d+
0
B(x,)c |y x|
is given in [10]. Here and below, Ad, = 21 d/2 [(d + )/2]/ (1 /2)
is an appropriate normalizing constant ([5, 27]). It follows from (1) and (3) that an
-harmonic function f in D satisfies
Pr (x , y )f (y) dy, x B(, r),
(4)
f (x) =
|y|>r
for every ball B(, r) of closure contained in D. In fact, the condition characterizes
functions -harmonic in D. The integral in (4) is absolutely convergent and, by (1),
f is smooth in D and
|f (y)|
dy < .
(5)
d+
Rd (1 + |y|)
If, in addition, f is nonnegative on Rd and nonzero in D, then it is positive in D,
regardless of the connectedness of D. This is a consequence of Harnack inequality,
see, e.g., Lemma 1, p. 138 in [9].
3. Kernel Functions
The cones described in the introduction are called right circular cones. By a
generalized cone in Rd we shall mean in this paper an open set Rd with the
property that if x and r > 0 then rx . If 0 then = Rd . Otherwise,
is characterized by its intersection with the unit sphere Sd1 Rd . Namely, let
= be a relatively open subset of Sd1. Without loosing generality in what
follows, we assume that 1 = (0, 0, . . . , 0, 1)
. The generalized cone spanned
by
is then
=
= {x Rd : x = 0 and x/|x|
}.
Note that we do not impose any regularity properties on
, in particular
may
be disconnected.
For n = 0, 1, . . . , we let Bn = {|x| < 2n } and n = Bn .
The following result follows from [33].
267
for all x Rd ,
(6)
v(x) = Ex v(X1 ),
for all x Rd ,
(7)
and
we have
C11 v(x) u(x) C1 v(x),
x B0 .
(8)
We note that by (6) and (7), u and v are regular -harmonic on 1 . By considering the regular -harmonic function v(x) = Px [X1 B1c ] 1 defined by the
boundary condition: v = 1 on B1c and v = 0 on B1 c , we see that (8) implies
that every function u satisfying the assumptions of Lemma 3.1 is bounded on B0
and in fact,
u(x) C2 u(1),
|x| < 1,
(9)
r > 0, x = 0,
x Rd .
(10)
x = 0.
(11)
268
x Rd .
(12)
We shall now describe some examples of generalized cones where the exponent
can be explicitly identified.
EXAMPLE 3.1. Consider = Rd . The only nonnegative -harmonic functions
on the whole of Rd are constants, see, e.g., Lemma 3.2, p. 545, in [12]. Thus M 1
and = 0 for this cone.
EXAMPLE 3.2. We put
/2
if xd > 0,
M(x) = xd
0
if xd 0,
where x = (x1 , x2 , . . . , xd ) Rd . Then M is -harmonic on Rd + = Rd {xd > 0}
by checking that /2 M(x) = 0, x Rd + (see the calculations pp. 2729 in [11]
and notice that (, /2) = 1/, with the notation therein). By Theorem 3.2, it
is the Martin kernel at infinity for = Rd + . In this case, = /2. (This rate of
decay is characteristic for smooth domains [15, 24].)
EXAMPLE 3.3. Let > 1 and consider the function M(x) = |xd |1 , x Rd .
Using the calculations in [11], pp. 2729 and noticing that
1
t 1 (1 + t)1 dt =
0
for every (0, 2), one obtains /2 M(x) = 0, x Rd , xd = 0. Again by
Theorem 3.2, M is the Martin kernel at infinity for cone = Rd \ {xd = 0} and
= 1 in this case. For the same cone but for 1, M is the indicator function
of Rd \ {xd = 0} since {xd = 0} is a polar set if 1 [27]. This time we have
= 0.
Let , be generalized cones in Rd , and let m, M be their respective Martin kernels
with pole at infinity. If then
Px {X1 B1c } Px {X1 B1c },
x Rd .
(13)
x B0 .
(14)
269
x 1 .
(15)
Let
a = inf
x
M(x)
.
m(x)
x 1 .
We note that H is homogeneous of degree ( ) = (), nonnegative and continuous, so if H (y) > 0 for some y (hence on a set of positive Lebesgue measure)
then h(x) > 0 for all x 1 by (4). By BHP we have that h m on B(0, 1)
with some > 0, thus H m everywhere. In particular M am m or
a = infx M(x)/m(x) a +, which is a contradiction. We conclude that H 0
or M1 = am. By the mean value property of m
M(x) = Ex {M(X1 )1 (X1 )},
x 1 .
x = 0,
(16)
270
x Rd .
(17)
1
Ex {K(Xt )f (Xt ); > t}.
K(x)
THEOREM 4.1. Let be the homogeneity degree of the Martin kernel M of the
cone . For p > 0 and x we have
p
if and only if
p < /,
(18)
if and only if
p < (d + 2)/.
(19)
Ex <
and
E0x <
valid for any positive random variable on any probability space, and the following
two lemmas.
271
LEMMA 4.2. There is C3 = C3 (, ) such that for all t > 0 and x Rd
satisfying |x| < t 1/ we have
C31 M(x)t / Px { > t} C3 M(x)t / .
(20)
|x| < 1.
(21)
Bn = {|x| < 2n }, n = 0, 1, . . . ,
and we have
Px { > 1} Px {1 > 1} + Px {1 < },
x Rd .
|x| < 1,
|x| < 1.
272
(22)
x 0 .
(23)
We have
Ex {K(X1 ); > 1} = Ex {K(X1 ); 3 1, > 1} + Ex {K(X1 ); 3 > 1}.
By the -harmonicity of K and Fatous lemma
K(x) Ex K(X3 ),
x Rd .
Thus,
Ex {K(X1 ); 3 > 1}
Ex {EX1 K(X3 ); 3 > 1} = Ex {K(X3 ); 3 > 1}
Ex {EX {K(X3 ); 3 > 1}; X1 B(41, dist(1, c ))}.
1
x, y Rd ,
(24)
273
where c3 = c3 (, ), as we shall see in Lemmas 4.5 and 4.6 below. It follows that
Ex {K(X1 ); 3 > 1} =
p 3 (1, x, y)K(y) dy
3
M(x)M(y)K(y) dy c4 M(x),
c3
3
P0x { > t} =
Thus domains of finite volume are Green bounded. It is also well known that there
exist Green bounded domains of infinite volume.
274
LEMMA 4.4. Suppose D Rd is such that there exits a constant C such that
s(x) C0 for all x Rd . Then for n = 1, 2, . . . ,
Ex Dn n!C0n1 s(x),
x Rd ,
Ex exp(D ) 1 s(x)/(1 C0 ),
(25)
and
Px {D > t}
s(x)
,
1 C0 exp(t) 1
Px {D > t} dt =
Px {D > t + r}
0
C0 Px {D > r}.
For n 2 we multiply both sides of the above inequality by (n 1)r n2 and
integrate, applying Fubinis theorem on the left-hand side, to obtain
Ex Dn nC0 Ex Dn1 .
The first asserted inequality follows by induction. The other two inequalities are
even easier and are left to the reader.
2
Let ptD (x, y) be the transition density of the process Xt killed off D:
ptD (x, y) dy.
Px {D > t, Xt A} =
A
Clearly
ptD (x, y) ptR (x, y) = t d/ p1R (xt 1/ , yt 1/ )
d/
d
(2 )
e| | d = ct d/ , x, y D, t > 0,
t
d
Rd
275
From these two observations and the semigroup property of ptD (x, y) we have
D
D
p3t (x, y) =
p2t
(x, z)ptD (z, y) dz
D
D
D
pt (x, w)pt (w, z) dw ptD (z, y) dz
=
D
c
t d/
(26)
x Rd .
Also,
s(x) c1 M(x),
x k1 .
The lemma can be proved in a very similar way as Lemma 4.2, by using the
boundary Harnack principle and the IkedaWatanabe formula. We skip the details.
From (20) and (22) we immediately obtain the following result.
COROLLARY 4.7. There are constants C3 and C4 depending only on , and d,
such that for all x ,
C31 |x| M(x/|x|) lim inf t / Px { > t}
t
t
C3 |x| M(x/|x|)
(27)
and
C41 |x|d+2 lim inf t (2+d)/ P0x { > t}
t
C4 |x|d+2 .
(28)
We also have the following heat-kernel version of this corollary. Recall that
pt (x, y) are the transition densities (heat kernel) of Xt killed off .
276
t
C5 |x| |y| M(x/|x|)M(y/|y|).
(29)
c
t d/
with c = (2 )d d (d/)/. The upper bound follows from this and the previous
result.
For the lower bound, suppose first that |x|, |y| < 1. By domain monotonicity,
we have p1 (x, y) p13 (x, y). However, since 3 is a bounded domain, it is
intrinsically ultracontractive by [25, 17]. Therefore
p13 (x, y) c1 Px {3 > 1}Py {3 > 1}.
From our proof of (21) it follows that for |x| < 1 we have Px {3 > 1} c2 M(x).
From this we conclude that for |x|, |y| < 1,
p1 (x, y) c3 M(x)M(y).
This inequality and scaling gives that for all t > max(|x| , |y| ),
pt (x, y) = t d/ p1 (t 1/ x, t 1/ y)
c3 t d/ M(t 1/ x)M(t 1/ y)
= c3 t (2d)/ M(x)M(y).
The left-hand side of (29) follows from this.
Corollaries 4.7 and 4.8 should be compared with the corresponding results for the
Brownian motion in generalized cones ([2] (1.5), (1.10) and (2.2)). However, the
exact limits are computed in [2]. It would be interesting to have the exact limits in
the current setting as well. In addition to generalized cones, the above limits have
been studied in [3, 34, 28, 29] in the case of Brownian motion in parabolas and in
other parabolic regions of the form D = {(x, y) R Rd1 : x > 0, |y| < Ax p },
0 < p < 1. The asymptotic behavior of the distribution of the exit time and of the
heat kernel for such regions is shown to be subexponential. It would be interesting
to determine the behavior of the distribution of the exit time and the transition
densities for stable processes in these regions. The scaling techniques we use here
do not seem to apply.
277
/2
Rd \
(31)
(y) (1)
dy.
|y 1|d+
(32)
The proof of (31) is somewhat tedious (because is not symmetric about 1) but
elementary and will be omitted. The formula (32) follows from (31) immediately.
Let be also homogeneous of degree ; (x) = |x| (x/|x|), x = 0. In view
of (30) we will only consider d < < .
By (32) and polar coordinates
(y) (1)
/2
dy
(1) = Ad, lim+
d+
0
Rd \ |y 1|
()r (1)
(d)
r d1
dr
= Ad, lim+
0
|r 1|d+
0
Sd1 \
[()u (d ) (1)u0 (d )] (d).
(33)
= Ad, lim+
0
Here
u (t) =
Sd1 \
1 t 1
(34)
(we drop d and from the notation); and (33) holds because
d+ 1
d
r
|r 1|
dr =
r d+ 1 (r 2 2rd + 1)(d+)/2 dr.
0
d < < ,
(35)
278
1
r d+1 (r + 1)d dr = .
1
(37)
Sd1
The principal value integral (P.V.) above is understood as in (33). We note that the
second integral in (37) vanishes if is homogeneous of degree 0 and the first one
vanishes whenever is constant on the unit sphere. The observation can be used
to show that the integrals converge and the second one converges absolutely (conf.
Lemma 4.5), a fact that can be also verified by a detailed inspection of u at t = 1.
A formula similar to (37) clearly holds for every vector Sd1:
[() ()]u0 ( ) (d) +
/2 () = Ad, P.V.
Sd1
()[u ( ) u0 ( )] (d),
(38)
+ Ad,
Sd1
279
(39)
Sd1 {1>}
will be called the spherical fractional Laplacian. The second integral in (38) will
be called the radial part and denoted R below.
REMARK 1. By the proof of Lemma 5.2 we see that (x) = |x| is -harmonic
on Rd \ {0} if and only if = d.
We will be concerned with nonnegative definiteness of the kernel u ( ).
We first consider dimension d = 2 and we will identify R2 with the complex
plane C. By a rotation, Ox2 axis in R2 will be identified with
z axis in C so that
1 R2 is now represented by 1 C. S1 will be replaced by ei , T
[0, 2 ) in formulas. With this identification in mind we define
u(2)
()
=
r 2+ 1 |rei 1|2 dr.
in
,
cn e
[0, 2 ).
n=1
n in
r e
/n =
n=,n=0
Since |1 z| = exp( log |1 z|), the sequence {cn = cn (r)} is the convolution exponent of the sequence {1{n=0} r |n| /(2|n|)}
. The latter sequence is
2
nonnegative and so {cn } is positive (c0 1).
Let be a test function and let , [0, 2 ). We consider the operator
2
(2)
(2)
()[u(2)
R () =
( ) u0 ( )] d.
0
280
cn ein() ()() d d dr
0
= 4
|an |
2
0
1 1
(r r 2 )(r 2 1 r d )cn (r) dr > 0,
r
m = 0, 1, . . . ,
(41)
where
cm =
2 (d1)/2Ad,
d/21
((d 1)/2))Cm
d/21
(1)
(43)
281
(44)
dm
[(1 t 2 )m+(d3)/2],
dt m
where
dm =
(d 2)m
,
2m m!(d/2 1/2)m
km = 2m
d +
2
1
cm dm r m
m
282
r 2 2rt + 1 =
The primitive function of the last integrand is (x + 1)/(x 2 2tx + 1), which
yields
u0 (t) =
1
.
1t
Sd1 Ym = m Ym ,
where
m = cm
(45)
and cm is given in (43). The proof of (45) follows from the usual FunkHecke
formula via the approximation (39).
d/21
1. For m > 0 we recall a formula for the
Note that 0 = 0 because C0
Gegenbauer polynomials ([22], (10.9.17), p. 175):
Cm (cos ) =
m
()n ()mn
n=0
d/21
n!(m n)!
cos(m 2n).
283
x Rd .
(46)
0 < r 1.
(47)
The argument verifying (47) is given in the proof of Lemma 5 in [7] and we refer
the reader to that paper for details (take Q = 0, As (Q) = r1 and Ar (Q) = 1 in the
notation of [7]). The fact that is strictly less that is important and distinguishes
the present situation from that of the potential theory of Brownian motion. By
scaling we have hs (1) c1 s , s 1. We define
us (x) =
hs (x)
,
hs (1)
s > 0, x Rd ,
1
= c11 s ,
c1 s
s 1, x Rd .
(48)
We claim that
ut (x) 2 C1 c11 [|x| 1] ,
t 2, x Rd .
(49)
284
To verify (49), assume that t 2. If |x| t/2 then we put s = 2(1 |x|) and by
Lemma 3.1 we have
ut (x) C1 us (x) C1 c11 2 [|x| 1] .
By Harnack inequality and our normalization ut (1) = 1, the functions ut are
uniformly bounded on F for any compact F and equicontinuous on F for
all large t. The last assertion follows from the Poisson formula for the ball or the
gradient estimates of [12]. Therefore there is a sequence tn and a function
M such that
M(x) = lim utn (x),
n
x Rd
x Rd .
(50)
Let x and set B = B(0, r), where r > |x|. We have for tn > r,
x
utn (y)B
(dy),
utn (x) =
(B)c
x
x
the -harmonic measure of B. Since B
Bx on
where we denote by B
c
B we have by (2)
x
(dy) 2d+ 3/2 Cd r |y|d dy,
B
(51)
Since utn (y) M(y) for all y, by (49), (51) and dominated convergence we have
x
x
utn (y)B (dy) =
M(y)B
(dy),
M(x) = lim
n
(B)c
(B)c
x Rd ,
285
Assume that H (x) > 0 for some, and therefore for every, x . Once again by
Lemma 3.1 and scaling
H (x) M(x),
x Rd ,
a = inf
which is a contradiction.
Thus H 0 and hence m = aM. The normalizing condition m(1) = M(1) = 1
yields a = 1 and the uniqueness of M is verified.
It remains to prove the homogeneity property of M. By the scaling of Xt , for
every k > 0 the function M(kx)/M(k1) satisfies the hypotheses used to construct
M. By uniqueness this function is equal to M, that is, M(kx) = M(x)M(k1) for
x Rd . In particular, M(kl1) = M(l1)M(k1) for every positive k, l. By continuity
there exists such that M(k1) = k M(1) = k and
M(kx) = k M(x),
x Rd .
(52)
(53)
286
For every n,
K(x) = Ex {K(XUn ); O} + Ex {K(XUn ); P }
= Ex {K(XU ); O, Un = U } + Ex {K(XUn ); O, Un < U } +
+ Ex {K(XUn ); P }.
(54)
K(x)
= Ex {K(X
x Rd .
(55)
for every x U , where Pr (x, y) is given by (2) for d (see Theorem C in [6]
for the case d = 1 < ). If d then
Pr (x, y) dy C|x|d , x U,
{|y|<r}
287
as can be shown from [4], Lemma 2.2. From this it follows that K(x)
C|x|d
for all sufficiently large x. Hence,
2
)C
T K(x)
= |x|d K(x/|x|
Acknowledgement
We are very grateful to an anonymous referee for the careful reading of this paper
and the many useful comments which have made the paper easier to read.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Bauelos, R., Lataa, R. and Mndez-Hernandez, P.J.: A BrascampLiebLuttinger-type inequality and applications to symmetric stable processes, Proc. Amer. Math. Soc. 129(10)
(2001), 29973008 (electronic).
Bauelos, R. and Smits, R.: Brownian motion in cones, Probab. Theory Related Fields 108
(1997), 299319.
Bauelos, R., DeBlassie, R.D. and Smits, R.: The first exit time of planar Brownian motion
from the interior of a parabola, Ann. Probab. 29 (2001), 882901.
Bass, R.F. and Cranston, M.: Exit times for symmetric stable processes in Rn , Ann. Probab.
11(3) (1983), 578588.
Blumenthal, R.M. and Getoor, R.K.: Markov Processes and Potential Theory, Springer-Verlag,
New York, 1968.
Blumenthal, R.M., Getoor, R.K. and Ray, D.B.: On the distribution of first hits for the
symmetric stable process, Trans. Amer. Math. Soc. 99 (1961), 540554.
Bogdan, K.: The boundary Harnack principle for the fractional Laplacian, Studia Math.
123(1) (1997), 4380.
Bogdan, K.: Representation of -harmonic functions in Lipschitz domains, Hiroshima Math.
J. 29(2) (1999), 227243.
Bogdan, K. and Byczkowski, T.: Probabilistic proof of boundary Harnack principle for
-harmonic functions, Potential Anal. 11 (1999), 135156.
Bogdan, K. and Byczkowski, T.: Potential theory for the -stable Schrdinger operator on
bounded Lipschitz domains, Studia Math. 133(1) (1999), 5392.
Bogdan, K., Burdzy, K. and Chen, Z.-Q.: Censored stable processes, to appear in Probab.
Theory Related Fields (2001).
288
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
Bogdan, K., Kulczycki, T. and Nowak, A.: Gradient estimates for harmonic and q-harmonic
functions of symmetric stable processes, Illinois J. Math. 46(2) (2002), 541556.
Burdzy, K.: Multidimensional Brownian Excursions and Potential Theory, Pitman Research
Notes in Mathematics 164, Longman Scientific & Technical, 1987.
Burkholder, D.L.: Exit times of Brownian motion, harmonic majorization, and Hardy spaces,
Adv. Math. 26(2) (1977), 182205.
Chen, Z.Q. and Song, R.: Estimates on Green functions and Poisson kernels for symmetric
stable process, Math. Ann. 312(3) (1998), 465501.
Chen, Z.-Q. and Song, R.: Martin boundary and integral representation for harmonic functions
of symmetric stable processes, J. Funct. Anal. 159(1) (1998), 267294.
Chen, Z.Q. and Song, R.: Intrinsic ultracontractivity, conditional lifetimes and conditional
gauge for symmetric stable processes on rough domains, Illinois J. Math. 44(1) (2000), 138
160.
Chung, K.L. and Zhao, Z.: From Brownian Motion to Schrdingers Equation, Springer-Verlag,
New York, 1995.
Davis, B. and Zhang, B.: Moments of the lifetime of conditioned Brownian motion in cones,
Proc. Amer. Math. Soc. 121(3) (1994), 925929.
DeBlassie, R.: Exit times from cones in R n of Brownian motion, Probab. Theory Related
Fields 74 (1987), 129.
DeBlassie, R.: The first exit time of a two-dimensional symmetric stable process from a
wedge, Ann. Probab. 18 (1990), 10341070.
Erdlyi, A. (ed.): Higher Transcendental Functions, Vols. I, II, McGraw-Hill, New York, 1953.
Ikeda, N. and Watanabe, S.: On some relations between the harmonic measure and the Lvy
measure for a certain class of Markov processes, J. Math. Kyoto Univ. 2(1) (1962), 7995.
Kulczycki, T.: Properties of Green function of symmetric stable process, Probab. Math.
Statist. 17(2) (1997), 339364.
Kulczycki, T.: Intrinsic ultracontractivity for symmetric stable processes, Bull. Polish Acad.
Sci. Math. 46(3) (1998), 325334.
Kulczycki, T.: Exit time and Green function of cone for symmetric stable processes, Probab.
Math. Statist. 19(2) (1999), 337374.
Landkof, N.S.: Foundations of Modern Potential Theory, Springer-Verlag, New York, 1972.
Li, W.: The first exit time of Brownian motion from unbounded domain, Preprint.
M. Lifshitz and Z. Shi, The first exit time of Brownian motion from parabolic domain,
Bernoulli 8(6) (2002), 745767.
Mndez-Hernndez, P.J.: Exit times from cones in Rn of symmetric stable processes, Preprint,
2000.
Rubin, B., Fractional Integrals and Potentials, Pitman Monographs 82, Addison-Wesley, 1996.
Ryznar, M.: Estimates of Green function for relativistic -stable process, Potential Anal. 17
(2002), 123.
Song, R. and Wu, J.-M.: Boundary Harnack principle for symmetric stable processes, J.
Funct. Anal. 168(2) (1999), 403427.
van den Berg, M.: Subexponential behavior of the Dirichlet heat kernel, Preprint.