Journal of Applied Mathematics Volume 2012, Article ID 761248,19pages doi:10.1155/2012/761248
Research Article
Spectral Approach to Derive the Representation Formulae for Solutions of the Wave Equation
Gusein Sh. Guseinov
Department of Mathematics, Atilim University, Incek, 06836 Ankara, Turkey
Correspondence should be addressed to Gusein Sh. Guseinov,guseinov@atilim.edu.tr Received 28 September 2011; Accepted 30 January 2012
Academic Editor: Pablo Gonz´alez-Vera
Copyrightq2012 Gusein Sh. Guseinov. 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.
Using spectral properties of the Laplace operator and some structural formula for rapidly decreasing functions of the Laplace operator, we offer a novel method to derive explicit formulae for solutions to the Cauchy problem for classical wave equation in arbitrary dimensions. Among them are the well-known d’Alembert, Poisson, and Kirchhoffrepresentation formulae in low space dimensions.
1. Introduction
The wave equation for a functionux1, . . . , xn, t ux, tofnspace variablesx1, . . . , xnand the timetis given by
∂2u
∂t2 Δu, 1.1
where
Δ ∂2
∂x21 · · · ∂2
∂xn2 1.2
is the Laplacian. The wave equation is encountered often in applications. For n 1 the equation can represent sound waves in pipes or vibrations of strings, for n 2 waves on the surface of water, for n 3 waves in acoustics or optics. Therefore, formulae that give the solution of the Cauchy problem in explicit form are of great significance. In the Cauchy
probleminitial value problemone asks for a solutionux, tof 1.1 defined forx ∈ Rn, t≥0 that satisfies1.1forx∈Rn,t >0 and the initial conditions
ux,0 ϕx, ∂ux,0
∂t ψx x∈Rn. 1.3
Ifn1 andϕ∈C2R,ψ∈C1R, then the classical solution of problem1.1,1.3is given by d’Alembert’s formula
ux, t ϕxt ϕx−t
2 1
2 xt
x−tψ y
dy. 1.4
Ifn2 andϕ∈C3R2,ψ ∈C2R2, then the solution of problem1.1,1.3is given by Poisson’s formula
ux, t 1 2π
|y−x|<t ψ
y dy
t2−y−x2 ∂
∂t
⎡
⎢⎣ 1 2π
|y−x|<t ϕ
y dy
t2−y−x2
⎤
⎥⎦, 1.5
wherex x1, x2,y y1, y2, and|y−x|2 y1−x12 y2−x22.
Ifn3 andϕ∈C3R2,ψ ∈C2R2, then the solution of problem1.1,1.3is given by Kirchhoff’s formula
ux, t 1 4πt
|y−x|tψ y
dSy ∂
∂t 1
4πt
|y−x|tϕ y
dSy
, 1.6
wherex x1, x2, x3,y y1, y2, y3,|y−x|2 y1−x12 y2−x22 y3−x32, anddSyis the surface element of the sphere{y∈R3:|y−x|t}.
Passing to an arbitrarynlet us denote byux, t Nϕx, tthe solution of the problem
∂2u
∂t2 Δu, x∈Rn, t >0, 1.7
ux,0 ϕx, ∂ux,0
∂t 0, x∈Rn. 1.8
It is easy to see that then the function
vx, t t
0
ux, τdτ 1.9
is the solution of the problem
∂2v
∂t2 Δv, x∈Rn, t >0, 1.10
vx,0 0, ∂vx,0
∂t ϕx, x∈Rn. 1.11
Indeed, integrating1.7we get t
0
∂2ux, τ
∂τ2 dτ t
0
Δux, τdτ Δ t
0
ux, τdτ Δvx, t. 1.12
Hence,
∂ux, t
∂t −∂ux,0
∂t Δvx, t or ∂ux, t
∂t Δvx, t, 1.13 by the second condition in1.8. On the other hand, from1.9,
∂vx, t
∂t ux, t, ∂2vx, t
∂t2 ∂ux, t
∂t . 1.14
Comparing1.13and1.14, we get1.10. Besides, vx,0 0, ∂vx,0
∂t ux,0 ϕx 1.15
so that initial conditions in1.11are also satisfied.
Consequently, the solutionux, tof problem1.1,1.3is represented in the form ux, t Nϕx, t
t
0
Nψx, τdτ. 1.16
It follows that it is sufficient to know an explicit form of the solutionNϕx, tof problem1.7, 1.8. It is known 1,2that
Nϕx, t 1 2m1πm
∂
∂t 1
t m
|y−x|tϕ y
dSy if n2m1, 1.17
Nϕx, t 1 2mπm
∂
∂t 1 t
m−1
∂ dt
|y−x|<t ϕ
y dy
t2−y−x2 ifn2m, 1.18
wherex x1, . . . , xn,y y1, . . . , yn,|y−x|2 y1−x12· · · yn−xn2, anddSyis the surface element of the sphere{y∈Rn:|y−x|t}.
In the present paper, we give a new proof of formulae1.17,1.18for the solution of problem1.7,1.8. Our method of the proof is based on the spectral theory of the Laplace operator. We hope that such a method may be useful also in some other cases of the equation and space.
The paper consists, besides this introductory section, of three sections. InSection 2, we describe the structure of arbitrary rapidly decreasing function of the Laplace operator, showing that it is an integral operator and giving an explicit formula for its kernel. Next we use these results inSection 3to derive the explicit representation formulae for the classical solution to the initial value problem for the wave equation in arbitrary dimensions. The final Section is an appendix and contains some explanation of several points in the paper.
2. Structure of Arbitrary Function of the Laplace Operator
LetAbe the self-adjoint positive operator obtained as the closure of the symmetric operator Adetermined in the Hilbert spaceL2Rnby the differential expression
−Δ − ∂2
∂x21 · · · ∂2
∂x2n
, x1, . . . , xn∈Rn, 2.1
on the domain of definition DA C∞0 Rnthat is the set of all infinitely differentiable functions on Rn with compact support. Let Eμ denote the resolution of the identity the spectral projectionfor A:
Af ∞
0
μdEμf, f ∈DA. 2.2
Next, letgt be any infinitely differentiable even function on the axis−∞ < t < ∞with compact support and
gλ
∞
−∞gteiλtdt 2.3
its Fourier transform. Note that the functiongλ tends to zero as|λ| → ∞λ∈Rfaster than any negative power of|λ|. Consider the operatorgA 1/2defined according to the general theory of self-adjoint operatorssee 3:
g
A1/2 f
∞
0
g
μ
dEμf, f ∈L2Rn. 2.4
The following theorem describes the structure of the operatorgA 1/2showing that it is an integral operator and giving an explicit formula for its kernel in terms of the function gt.
Theorem 2.1. The operatorgA 1/2is an integral operator
g
A1/2 fx
RnK x, y
f y
dy, f∈L2Rn. 2.5
Further, there is a smooth functionktdefined on the interval 0≤t <∞such that
K x, y
kx−y2
. 2.6
The functionktdepends on the functiongtas follows. If one sets Qt g√
t
, that is, Q t2
gt, 0≤t <∞, 2.7
then
kt
⎧⎪
⎪⎪
⎨
⎪⎪
⎪⎩
−1m
πm Qmt if n2m1,
−1m πm
∞
t
Q√mw
w−t dw if n2m,
2.8
where Qmt denotes the mth order derivative of Qt. Further, if suppgt ⊂ −a, a, then suppkt⊂ 0, a2. For any solutionψx, λof the equation
−Δψx, λ λ2ψx, λ, 2.9
the equality
Rnkx−y2 ψ
y, λ
dygλψx, λ 2.10
holds forλ∈R.
Proof. First we consider the casen 1. In this case, the statements of the theorem take the following form: kt Qt g√
t for 0 ≤ t < ∞; the operator gA 1/2 is an integral operator of the form
g
A1/2 fx
∞
−∞g x−y
f y
dy, 2.11
and for any solutionψx, λof the equation
−ψx, λ λ2ψx, λ, 2.12
the equality
∞
−∞g x−y
ψ y, λ
dygλψx, λ 2.13
holds.
To prove the last statements note that, in the casen1, the operatorAis generated in the Hilbert spaceL2−∞,∞by the operation−d2/dx2and the operatorA1/2by the operation id/dx. The resolventRμ A−μI−1of the operatorAhas the form
Rμfx i 2√μ
∞
−∞ei|x−y|√μf y
dy, 2.14
while the spectral projectionEμof the operatorAhas the formsee 3, page 201
Eμfx ∞
−∞
sin√μ x−y π
x−y f y
dy, 0≤μ <∞,
Eμ0 for μ <0.
2.15
Therefore,
g
A1/2 fx
∞
0
g
μ
dEμfx
∞
0
g
μ∞
−∞
cos√μ x−y 2π√μ f
y dy
dμ
∞
−∞
1 π
∞
0
gλcosλ x−y
dλ
f y
dy ∞
−∞g x−y
f y
dy,
2.16
where we have used the inversion formula for the Fourier cosine transform. Therefore,2.11 is proved. To prove2.13note that the general solution of2.12is
ψx, λ
⎧⎨
⎩
c1cosλxc2sinλx ifλ /0, c1c2x ifλ0,
2.17
wherec1andc2are arbitrary constants. Then, we have, forλ /0, ∞
−∞g x−y
ψ y, λ
dyc1 ∞
−∞g x−y
cosλy dyc2 ∞
−∞g x−y
sinλy dy c1
∞
−∞gtcosλx−tdtc2
∞
−∞gtsinλx−tdt c1
∞
−∞gtcosλxcosλtsinλxsinλtdt c2
∞
−∞gtsinλxcosλt−sinλtcosλxdt c1cosλx
∞
−∞gtcosλt dtc2sinλx ∞
−∞gtcosλt dt c1cosλxc2sinλx
∞
−∞gtcosλt dtψx, λgλ,
2.18
where we have used the fact that the functiongtis even and therefore ∞
−∞gtsinλt dt0. 2.19
The same result can be obtained similarly forλ0. Thus,2.13is also proved.
Now we consider the casen≥2. We shall use the integral representation
Rμfx
Rnr x, y;μ
f y
dy 2.20
of the resolventRμ A−μI−1 of the operatorA. As is known 4, Section 13.7, Formula 13.7.2,
r x, y;μ
iμn−2/4
2n2/2πn−2/2x−yn−2/2Hn−2/21 x−yμ
, 2.21
whereHν1zis the Hankel function of the first kind of orderν. Next, according to the general spectral theory of self-adjoint operators 3, page 150, Formula11, we have
dEμfx 1 2πi
Rμi0−Rμ−i0
fxdμ. 2.22
Therefore, from2.4it follows that the representation2.5holds with K
x, y 1
2πi ∞
0
g
μ r
x, y;μi0
−r
x, y;μ−i0 dμ. 2.23
Now the representation2.6, which expresses thatKx, yis a function of|x−y|2, follows from2.23by2.21.
To prove2.10we use2.23. By virtue of2.23,
Rnkx−y2 ψ
y, λ dy
RnK x, y
ψ y, λ
dy
lim
ε→0
Rn
1 2πi
∞
0
g
μ r
x, y;μiε
−r
x, y;μ−iε dμ
ψ y, λ
dy
ψx, λlim
ε→0
ε π
∞
0
g√μ μ−λ22
ε2dμψx, λgλ,
2.24
see Appendix. Here we have used the fact that from2.9it follows that
−Δ−zψx, λ λ2−z
ψx, λ, 2.25
that is,
ψx, λ λ2−z
−Δ−z−1ψx, λ, 2.26
and therefore
Rnr x, y;z
ψ y, λ
dy 1
λ2−zψx, λ. 2.27
Finally, to deduce the explicit formulae2.7,2.8, we takeψx, λ eiλx1 in2.10.
Then, puttingx x2, . . . , xn, we can write
Rnkx1−y12x−y2
eiλy1dy1dygλeiλx1. 2.28
If we set
x1−y12
w, 2.29
then the left-hand side of2.28equals ∞
−∞
Rn−1k
wx−y2 dy
eiλy1dy1. 2.30
On the other hand,
Rn−1k
wx−y2 dy
∞
0
|x− y|rk
wx−y2 dS
dr
∞
0
k
wr2
|x− y|rdS
drσn−1 ∞
0
rn−2k wr2
dr
1 2σn−1
∞
w
t−wn−3/2ktdt,
2.31
where
σn 2πn/2
Γn/2 2.32
is the surface area of then−1-dimensional unit sphereΓis the gamma functionanddS denotes the surface element of the sphere{y∈Rn−1:|x−y| r}. Therefore, setting
Qw 1
2σn−1 ∞
w
t−wn−3/2ktdt, 2.33
we get that2.28takes the form ∞
−∞Qweiλy1dy1gλeiλx1. 2.34 Substituting here the expression ofwgiven in2.29and making then the change of variables x1−y1t, we obtain
∞
−∞Q t2
eiλtdtgλ ∞
−∞gteiλtdt. 2.35
Hence2.7follows. Further, it is not difficult to check that the formula2.33forn ≥ 2 is equivalent to2.8, see Appendix.
Sincegtis smooth and has a compact support, it follows from2.7,2.8that the functionktalso is smooth and has a compact support; more precisely, if suppgt⊂−a, a, then suppkt ⊂ 0, a2. This implies, in particular, convergence of the integral in2.10for each fixedx. The theorem is proved.
3. Derivation of Formulae 1.17, 1.18
Consider the Cauchy problem1.7,1.8:
∂2u
∂t2 Δu, x∈Rn, t >0, 3.1
ux,0 ϕx, ∂ux,0
∂t 0, x∈Rn, 3.2
whereuux, t,t≥0,x x1, . . . , xn∈Rn,ϕx∈C∞0Rn. Forν ν1, . . . , νn,x x1, . . . , xn∈Rn, let us set
|ν|2ν21· · ·νn2, ν, x ν1x1· · ·νnxn. 3.3
Since
−Δeiν,x|ν|2eiν,x, 3.4
applying2.9,2.10, we get
Rnkx−y2
eiν,ydyg|ν|e iν,x ν∈Rn. 3.5
Hence, by the inverse Fourier transform formula,
kx−y2 1
2πn
Rng|ν|e iν,xe−iν,ydν. 3.6 Multiplying both sides of the last equality byϕyand then integrating ony∈Rn, we get
Rnkx−y2 ϕ
y
dy 1
2πn
Rng|ν|eiν,x
!
Rnϕ y
e−iν,ydy
"
dν. 3.7
Substituting here forg|ν| its expression
g|ν| 2 ∞
0
gtcos|ν|tdt 3.8
and setting
ux, t 1 2πn
Rncos|ν|teiν,x
!
Rnϕ y
e−iν,ydy
"
dν, 3.9
we obtain
Rnkx−y2 ϕ
y dy2
∞
0
gtux, tdt. 3.10
Obviously, the functionux, tdefined by3.9is the solution of problem3.1,3.2. Next we will transform the left-hand side of3.10usingTheorem 2.1.
First we consider the casen1. In this case,3.10takes the form ∞
−∞kx−y2 ϕ
y dy2
∞
0
gtux, tdt 3.11
and from2.7,2.8we have
k t2
Q t2
gt. 3.12
Therefore, making the change of variablesy−xtand taking into account the evenness of the functiongt, we can write
∞
−∞kx−y2 ϕ
y dy
∞
−∞k t2
ϕxtdt
∞
−∞gtϕxtdt ∞
0
gt
ϕxt ϕx−t dt.
3.13
Substituting this in the left-hand side of3.11, we obtain ∞
0
gt
ϕxt ϕx−t dt2 ∞
0
gtux, tdt. 3.14
Hence, by the arbitrariness of the smooth even functiongtwith compact support, we get
ux, t ϕxt ϕx−t
2 . 3.15
Further assume thatn≥2. Making the change of variables
y−xtω, 0≤t <∞, |ω|1, ω ω1, . . . , ωn, dytn−1dt dSω, 3.16 wheredSωis the surface element of the unit sphere{ω∈Rn:|ω|1}, we get
Rnkx−y2 ϕ
y dy
∞
0
tn−1k t2
|ω|1ϕxtωdSω
dt. 3.17
Further, making in the right-hand side of3.17the change of variables
xtωy, dSytn−1dSω, 3.18
wheredSyis the surface element of the sphere{y∈Rn:|y−x|t}, we have
tn−1
|ω|1ϕxtωdSω
|y−x|tϕ y
dSy:Pϕx, t. 3.19
Therefore,
Rnkx−y2 ϕ
y dy
∞
0
k t2
Pϕx, tdt, 3.20
and3.10becomes
∞
0
k t2
Pϕx, tdt2 ∞
0
gtux, tdt. 3.21
Consider the cases of odd and evennseparately.
Letn2m1m∈N. Then, by2.8we have
k t2
−1m πm Qm
t2
3.22
and it follows from2.7 by successive differentiationthat
Qm t2
1
2t
∂
∂t m
gt. 3.23
Therefore,
k t2
−1m 2mπm
1 t
∂
∂t m
gt, 3.24
and3.21takes the form
−1m 2mπm
∞
0
1 t
∂
∂t m
gt
Pϕx, tdt2 ∞
0
gtux, tdt. 3.25
Further, integratingmtimes by parts, we get ∞
0
1 t
∂
∂t m
gt
Pϕx, tdt Rx, t|t∞t0 −1m ∞
0
gt ∂
∂t 1 t
m
Pϕx, tdt, 3.26
where
Rx, t #m
k1
−1k−1 t
1 t
∂
∂t m−k
gt∂
∂t 1 t
k−1 Pϕx, t
#m
k1
−1k−1 t
1 t
∂
∂t m−k
gt∂
∂t 1 t
k−1 t2m
|ω|1ϕxtωdSω.
3.27
Sincegtis identically zero for large values oft, we have from3.27thatRx,∞ 0. Also, it follows directly from3.27thatRx,0 0. Therefore,3.25becomes
1 2mπm
∞
0
gt ∂
∂t 1 t
m
Pϕx, tdt2 ∞
0
gtux, tdt. 3.28
Since in3.28gtis arbitrary smooth even function with compact support, we obtain that ux, t 1
2m1πm ∂
∂t 1
t m
Pϕx, t. 3.29
This coincides with1.17by3.19.
Now let us consider the casen2mm∈N. In this case, by2.8we have
k r2
−1m πm
∞
r2
Q√mw
w−r2dw −1m πm
∞
r
Qm t2
√ 2t
t2−r2 dt, 3.30 and therefore
∞
0
k r2
Pϕx, rdr −1m πm
∞
0
∞
r
Qm t2
√ 2t
t2−r2 dt
Pϕx, rdr
−1m πm
∞
0
∞
r
Qm t2
√ 2t
t2−r2 dt
|y−x|rϕ y
dSy
dr
−1m πm
∞
0
⎧⎪
⎨
⎪⎩ ∞
r
Qm t2
2t
⎡
⎢⎣
|y−x|r
ϕ y
dSy
t2−y−x2
⎤
⎥⎦dt
⎫⎪
⎬
⎪⎭dr
−1m πm
∞
0
Qm t2
2t
⎧⎪
⎨
⎪⎩ t
0
⎡
⎢⎣
|y−x|r
ϕ y
dSy
t2−y−x2
⎤
⎥⎦dr
⎫⎪
⎬
⎪⎭dt.
3.31
Hence, setting
Hϕx, t: t
0
⎡
⎢⎣
|y−x|r
ϕ y
dSy
t2−y−x2
⎤
⎥⎦dr
|y−x|<t ϕ
y dy
t2−y−x2, 3.32
we get
∞
0
k r2
Pϕx, rdr −1m πm
∞
0
Qm t2
2tHϕx, tdt. 3.33
Substituting this in the left-hand side of3.21 beforehand replacingtbyrin the left side of 3.21, we obtain
−1m πm
∞
0
Qm t2
2tHϕx, tdt2 ∞
0
gtux, tdt 3.34
or, using3.23,
−1m 2m−1πm
∞
0
1 t
∂
∂t m
gt
tHϕx, tdt2 ∞
0
gtux, tdt. 3.35
Further, integratingmtimes by parts, we get ∞
0
1 t
∂
∂t m
gt
tHϕx, tdt Lx, t|t∞t0 −1m ∞
0
gt ∂
∂t 1 t
m
tHϕx, tdt, 3.36
where
Lx, t #m
k1
−1k−11 t
∂
∂t m−k
gt∂
∂t 1
t k−1
tHϕx, t. 3.37
Sincegtis identically zero for large values oft, we have from3.37thatLx,∞ 0. Also, using the expression ofHϕx, t,
Hϕx, t t
0
⎡
⎢⎣
|y−x|r
ϕ y
dSy
t2−y−x2
⎤
⎥⎦dr
t
0
r2m−1
|ω|1
ϕxrω
√t2−r2 dSω
dr
t
0
r2m−1
√t2−r2
|ω|1ϕxrωdSω
dr
t
0
t2−ξ22m−2
|ω|1ϕ
x
t2−ξ2ω
dSω
dξ,
3.38
we can check directly from3.37thatLx,0 0. Therefore,3.35becomes 1
2m−1πm ∞
0
gt ∂
∂t 1
t m
tHϕx, tdt2 ∞
0
gtux, tdt. 3.39
Since in3.39gtis arbitrary smooth even function with compact support, we obtain that ux, t 1
2mπm ∂
∂t 1 t
m
tHϕx, t 1
2mπm ∂
∂t 1 t
m−1
∂
∂tHϕx, t.
3.40
This coincides with1.18by3.32.
Appendix
For reader’s convenience, in this section we give some explanation of several points in the paper.
1Let us show how2.33forn≥2 implies2.8.
Letn2m1, wherem≥1. Then, since 1
2σ2m 1 2
2πm
Γm πm
m−1!, A.1 Equation2.33takes the form
Qw πm ∞
w
t−wm−1
m−1! ktdt. A.2
Hence applying the differentiation formula d
dw ∞
w
Gt, wdt−Gw, w ∞
w
∂Gt, w
∂w dt A.3
repeatedly, we find
Qmw πm−1mkw A.4
which gives2.8forn2m1.
In the casen2mwithm≥1,2.33takes the form
Qw 1
2σ2m−1 ∞
w
t−w2m−3/2ktdt. A.5
Hence,
Qm−1w 1 2σ2m−1
∞
w
−1m−12m−3 2
2m−5 2 · · ·1
2t−w−1/2ktdt. A.6
Therefore, taking into account that by virtue of Γx x−1Γx−1, Γ
1 2
√
π, A.7
we have 1
2σ2m−1 π2m−1/2
Γ2m−1/2 π2m−1/2
2m−3/22m−5/2· · ·1/2Γ1/2 πm−1
2m−3/22m−5/2· · ·1/2,
A.8
we get
Qm−1w −1m−1πm−1 ∞
w
√kt
t−wdt. A.9
In the right-hand side we replacetbyu, then divide both sides by√
w−tand integrate on w∈t,∞to get
∞
t
Q√m−1w
w−t dw −1m−1πm−1 ∞
t
√ 1 w−t
∞
w
√ku u−wdu
dw
−1m−1πm−1 ∞
t
ku u
t
dw
w−tu−w
du
−1m−1πm ∞
t
kudu,
A.10
because for anyt < u, using the change of variables√
w−tξ, we have u
t
dw
w−tu−w 2 √u−t
0
dξ
u−t−ξ2 2 arcsin ξ
√u−t ξ
√u−t
ξ0 2 arcsin 1π.
A.11
Therefore, differentiatingA.10with respect tot, we get
kt −1m πm
d dt
∞
t
Q√m−1w
w−t dw −1m πm
d dt
∞
0
Qm−1√ut
u du
−1m πm
∞
0
Qmut
√u du −1m πm
∞
t
Qmw
√w−t dw.
A.12
Thus,2.8is obtained also forn2mwithm≥1.
2Here we explain2.24. Note that since the spectrum of the operatorAis 0,∞ zero is included into the spectrum, the spectral representation formula2.4 should be understood in the sense of the formula
g
A1/2 f
∞
−δg μ
dEμf, A.13
whereδis an arbitrary positive real number and the integral does not depend onδ >0Eμis zero on−∞,0becauseAis a positive operator. Therefore, for2.24we have to show that
εlim→0
ε π
∞
−δ
g√μ
μ−λ22ε2dμgλ, λ∈R, A.14
for anyδ >0.
Since for anyε >0
ε π
∞
−δ
1
μ−λ22
ε2dμ ε π
∞
−δ−λ2
du u2ε2 1
π π
2 arctanδλ2 ε
, A.15
we have
εlim→0
ε π
∞
−δ
1
μ−λ22ε2dμ1, λ∈R, ε
π ∞
−∞
du u2ε2 1.
A.16
Givenα >0, we can choose aβ >0 such that g
uλ2
−gλ < α foru∈Ω '
u:−δ−λ2< u <∞,|u|< β(
A.17
since the functiongz is continuous forzλwe choose the continuous branch of the square root for which√
11. Further, we choose a numberMsuch that
gz≤M for|Imz| ≤C <∞, A.18
for sufficiently large positive numberC. This is possible by2.3and the fact thatgthas a compact support. Let us setΩ −δ−λ2,∞\Ω. Then,
ε π
∞
−δ
g√μ
μ−λ22ε2dμ−gλ ε π
∞
−δ
1
μ−λ22ε2dμ
≤ ε π
∞
−δ−λ2
g√ uλ2
−gλ u2ε2 du ε
π
Ω
g√ uλ2
−gλ u2ε2 du ε
π
Ω
g√ uλ2
−gλ u2ε2 du.
A.19
Further,
ε π
Ω
g√ uλ2
−gλ u2ε2 du < α
π ∞
−∞
ε
u2ε2duα, ε
π
Ω
g√ uλ2
−gλ
u2ε2 du≤ 2M π
|u|≥β
ε u2ε2du 4M
π ∞
β
ε
u2ε2du 4M π
π
2 −arctanβ ε
.
A.20
For fixedβ, the last expression tends to zero asε → 0; hence, and byA.16,A.19, andA.20we getA.14.
3The formula2.14follows from2.21forn1 noting that
H−1/21 z 2
πz 1/2
eiz. A.21
4 The difference between operators ∂/∂t 1/tm formulae 1.17, 1.18 and 1/t ∂/∂tmformula3.25is given by
∂
∂t 1
t m
∂
∂t 1
t
∂
∂t m−1
1
t. A.22
5The explicit formula for the solution of the wave equation in the caseneven can be derived from the casenodd by a known computation called the “method of descent”see 1.
6Since for suppgt ⊂ −a, a,a > 0, we have suppkt ⊂ 0, a2, and on the left- hand side of2.10the integral is taken in fact over the ball{y∈Rn:|y−x|< a}, for fixedx.
Therefore, this integral is finite for eachx∈Rnand any solutionψx, λof2.9. We proved 2.10forλ∈R. If the solutionψx, λis an analytic function ofλ∈C, then2.10will be held also for complex values ofλby the uniqueness of analytic continuation.
References
1 R. Courant and D. Hilbert, Methods of Mathematical Physics. Vol. II, Interscience Publishers, New York, NY, USA, 1962.
2 V. I. Smirnov, A Course of Higher Mathematics, Vol. II, Addison-Wesley, Reading, Mass, USA, 1964.
3 M. S. Birman and M. Z. Solomjak, Spectral Theory of Self-Adjoint Operators in Hilbert Space, Reidel, Dordrecht, The Netherlands, 1987.
4 E. C. Titchmarsh, Eigenfunction Expansions Associated with Second-Order Differential Equations. Vol. 2, Clarendon Press, Oxford, UK, 1958.
Submit your manuscripts at http://www.hindawi.com
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Mathematics
Journal ofHindawi Publishing Corporation
http://www.hindawi.com Volume 2014
in Engineering
Hindawi Publishing Corporation http://www.hindawi.com
Differential Equations
International Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Probability and Statistics
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Journal of
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Mathematical PhysicsAdvances in
Complex Analysis
Journal ofHindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Optimization
Journal ofHindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Combinatorics
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
International Journal of
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Operations Research
Journal of
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Function Spaces
Abstract and Applied Analysis
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
The Scientific World Journal
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Decision Sciences
Discrete Mathematics
Journal ofHindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014
Stochastic Analysis
International Journal of