2.1,下列事件有什么关系? 试指出并说明理由,
(1) ))(( ntN <与)( tS
n
>
)()(
n
SNntN =<∵
tS
n
>∴
)(tNn
StS ≥>∵
ntN <∴ )(
)())(( tSntN
n
>=<∴
(2) ))(( ntN ≤与)( tS
n

若计数过程具有Possion过程的性质,可不考虑同一时刻有2个以上“顾客”到达的情况,即可忽略0)(
1
==
+nn
SSP的小概率事件,假设
1+
<
nn
SS,则
)()1)(())((
1
tSntNntN
n
>=+<=≤
+

)())(( tSntN
n
≥?≤∴
若是普通的计数过程,则
1+

nn
SS,当tSS
nn
==
+1
时,
nntN >+= 1)(,两者无包含关系,
(3) ))(( ntN >与)( tS
n
<
同上,若
1+
<
nn
SS,则
)())(( tSntN
n
<?>

1+

nn
SS,两者无包含关系,
(4) ))(( xtW >与)0)()(( =?+ tNxtN
)0)()((
)1)()((
)1)()((
})())(( { )(
)())((
1)(
1)(
=?+=
<?+=
+<+=
>?<+>=
>?=>
+
+
tNxtN
tNxtN
tNxtN
tSntNxtS
xtSxtW
ntN
tN

2.3,设)0 ),(( ≥ttN为Poisson过程,参数为λ,求或证明,
(1) 求)]()([ tsNtNE + ;
222
2
2
))](([)]([)]([)]([
)]([)]]()()[([
)]()([
ttst
tNEtNDsNEtNE
tNEtNtsNtNE
tsNtNE
λλλ ++=
++=
+?+=
+
(2) 求})()({ sNtsNE +的分布律;
{( ) () }
{()()()()}
{()()()}{()()}
ENs t Ns m
ENs t Ns Ns Ns m
E Nst NsNs m ENsNs m
tmλ
∴+ =
=+?+ =
=+? =+ =
=+
[ ] ))((})()({})()({ msNPmsNtsNEsNtsNEP ===+=+∵
∴分布律为
[]
s
m
e
m
s
mtsNtsNEP
λ
λ
λ
=+=+
!
)(
})()({
(3) 任给ts ≤≤0,有1)}()({ =≤ tNsNP,
证明,
令0,≥+= st,则
1
!
)(
}0)()({)}()({
0
=
=
≥+=≤

∞+
=

k
k
e
k
sNsNPtNsNP
λ
λ
得证,
(4) 任给0,0 >≤≤ εts,有0})()({lim =>?

εsNtNP
st
,
证明,
}1)()({})()({0 ≥?≤>?≤ sNtNPsNtNP ε
0
lim1
}0)()({lim1
}1)()({lim
)(
=
=
==
≥?




st
st
st
st
e
sNtNP
sNtNP
λ
得证,
2.16 设}1,{ ≥nX
n
iid,
n
X的概率密度函数为

>
=

δ
δρ
δρ
x
xe
xf
x
,0
,
)(
)(
其中0>δ给定,求更新过程中的概率))(( ktNP ≥,
观察δ+'
nn
XX和的分布函数,其中'
n
X服从参数为ρ的指数分布,对应一个Poisson过程)(' tN,
)'(dd)(
0
)(
xXPueuexXP
n
x
u
x
u
n
≤+===≤
∫∫

δρρ
δ
ρ
δ
δρ

两者的分布函数完全相同,可令δ+= '
nn
XX,则


∞+
=

=
>
=
≥?=?≤=
≤=≤=≥∴
ki
kt
ii
k
k
i
ik
ktIe
i
kt
kktNPktSP
ktXPtSPktNP
}{
!
)(
))('()'(
)'()())((
)(
0
δ
δρ
δδ
δ
δρ
2.17 设
n
X的概率密度0,)(
2
≥=
xxexf

λ,求相应的更新函数)(tm,
对分布函数)(tF作Laplace变换,得
+
=
=∴
+
===
∫∫
+∞
+?
+∞
λ
λλ
λ
λ
λ
λ
22
1
)(
~
1
)(
~
)(
~
)(
d)(d)(
~
2
2
0
)(2
0
sssF
sF
sm
s
ttetFesF
tsst
i
i
1
'( ) ( ) d,0,0 ( i )
2πi
st
mt mse s t s
σ
σ
σ σω
+∞

∴= ≥≥=+

null
查Laplace反变换表,得更新函数的导数为,
4
1
42
d)1(
2
)(
0)0(
)1(
2
)('
2
0
2
2
+=?=∴
=
=

t
t
t
t
et
tetm
m
etm
λ
λ
λ
λλ
λ

2.24,令niXXYXY
iii
≤≤?==
2,,
)1()()1(1
,试问
n
YYY,,,
21
null是否独立? 同分布? 并证明你的猜想?
通过计算可知顺序统计量),,,(
)((2))1( n
XXX null的联合p.d.f.

}0{
)(
21
21
21
!
),,,(
n
n
xxx
xxxn
n
Ien
xxxf
<<<≤
+++?
=
null
null
null
λ
λ

1?
=
iii
xxy,则变换的雅可比行列式为,
1
11111
0111
0011
0001
),,,(
),,,(
21
21
==
=
nullnullnullnullnull
null
null
null
null
null
n
n
yyy
xxx
J
所以),,,(
21 n
YYY null的联合p.d.f.为
}1,0{
)1(
}1,0{
)1(
21
21
21
)1(
!
),,,(
niy
yynny
niy
yynnyn
n
i
n
i
n
Ieenen
Ieeen
yyyg
≤≤≥

≤≤≥

=
=
λλλ
λλλ
λλλ
λ
null
null
null
计算每个变量的边缘p.d.f.,得,
niIeinyg
i
i
y
yin
i
≤≤?+?=

+
1,)1()(
}0{
)1(λ
λ

)()()(),,,(
2121 nn
ygygygyyyg nullnull =
可知
n
YYY,,,
21
null独立,但不同分布,
2.25,设}0,)({ ≥ttN为时齐Poisson过程,
n
SSS,,,
21
null为事件相继发生的时刻,
(1) 给定ntN =)(,试问
1121
,,,

nn
SSSSS null是否条件独立? 是否同分布? 试证明你的猜想?
已知
n
SSS,,,
21
null在)1( )( ≥= nntN条件下的p.d.f.为,
}0{21
21
!
),,,(
tttt
n
n
n
I
t
n
tttf
≤<<<<
=
null
null
令niSSXSX
iii
≤≤?==
2,,
111
.令
1?
=
iii
ttx,则变换的雅可比行列式为,
1
),,,(
),,,(
21
21
=
=
n
n
xxx
ttt
J
null
null
所以),,,(
21 n
XXX null的联合p.d.f.为
}0,{21
21
!
),,,(
≥≤+++
=
in
xtxxx
n
n
I
t
n
xxxg
null
null
计算边缘p.d.f.,得
∫∫
≥?≤
+?


=
=
0,
1121
1
ddddd
!
)(
ki
n
ik
k
k
i
xxtx
nii
n
iX
xxxxx
t
n
xg
nullnull
下面计算上式中的重积分,
1
1
1
12 1 2
,0
12
,0
1
12 1 1
,0
1
{0 }
(,,,)dd d 1
dd d
!
()
dd d d d
(1)!
()
()
n
kk
k
n
kk
k
n
kik
k
ki
ii
nn
xtx
n
n
xtx
n
i
ii n
xtxx
n
i
Xi xt
n
gxx x x x x
t
xx x
n
tx
xx x x x
n
nt x
gx I
t
=
=
=

≤≥
≤≥
+
≤? ≥
≤≤
=

∴=


∴=?
∫∫
∫∫
∫∫
∵nullnull
null
nullnull
可知,)()()(),,,(
2121 nn
xgxgxgxxxg nullnull ≠,因此
1121
,,,

nn
SSSSS null条件不独立,但同分布,
(2) 求[ ])(
1
tNSE的分布律,
根据习题1.1的结果有
[]

+∞
=>==
0
11
d))(()( sntNsSPntNSE
① 当0=n时,
11
0
0
()
0
() 0 ( () 0)d
(() 0 () 0)d
1
d
st
E SNt PS sNt s
PNs Nt s
est
λ
λ
+∞
+∞
+∞

=?= >=

===
==+



分布律为,
1
1
(() (()0)
t
PESNt t PNt e
λ
λ

=+ = = =


② 当1≥n时,
1
1(1)
00
0
()
(())d()d
d
1
t
n
t
ESNt n
PS sNt n s PY s s
ts t
s
tn
+∞
=?

=>==>

==

+

∫∫

分布律为:
1
()
(() (())
1!
n
t
tt
PESNt PNt n e
nn
λ
λ

== ==

+

(3) 利用(1)及(2),求[ ])(tNSE
k
的分布律,
(可利用(2)的方法,[ ]

+∞
=>==
0
d))(()( sntNsSPntNSE
kk
.)
① 当1 kn≤ ≤时,
-1 -1 -2 2 1 1
-1 -1 -2
1
()
()
[ ()][ ()]
[ ( ) ]
( )
1
k
kk k k
kk k k
ESNt n
E SS S S SSSNtn
ES S Nt n ES S Nt n
ES Nt n
kt
n
=?

+?++?+ =?

=? =+? =+
+=
=
+
null
null

因为同分布
分布律为,
()
(() (()),1
1!
n
t
k
kt t
PESNt PNt n e k n
nn
λ
λ

= === ≤≤

+

② 当0 nk≤ <时,
() () 1 () 1
() () 1 ()
1
1
()
()
[ ()][ ()]
[ ( ) ]
[]
k
Nt Nt Nt k k
Nt Nt Nt
kk
kn
i
i
ESNt n
E SS S SSNtn
ES Nt n ES S Nt n
ES S Nt n
tEX
kn
t
λ
+?
+
=
=?

+?++? =

= =+? =+
+? =
=+
=+

null
null

分布律为,
()
(() (()),0
!
n
t
k
kn t
PESNt t PNt n e n k
n
λ
λ
λ

= += == ≤<


2.26,设}0,)({ ≥ttN是参数为λ的时齐Poisson过程,
n
SS,0
0
=为第n个事件发生的时刻,求,
(1) ),(
52
SS的联合p.d.f,
用微元法来求解,

52
tt <,取充分小的0>h,使得
2222
555222
h
tt
h
t
h
tt
h
t +<<?<+<<?,

222555
222
2
52 55
()
22
{( )1,( ) ( )1,
( ) ( ) 2,( ) ( ) 1} ( )
22 22
hhhh
Pt S t t S t
hhh
PNt Nt Nt
hh hh
Nt Nt Nt Nt Oh
< ≤+<?< ≤+
=?=+=
+= +=+
2
52
5
22
()
()52
2
2
2
()
52 2 2
2
252
()
()
22
( )
1
()( ) ()
22
h
t
tthhh
h
t
tthh
te he e he
Oh
h
ht t t he Oh
λ
λλ λ
λ
λ
λλ
λ


+

=
+
= +
}0{
2
252
5
52
52
5
2
)(
),(
tt
t
I
ettt
ttf
<<
=∴
λ
λ
(2) [ ]1)(
1
≥tNSE

[] [ ] [ ] )1)((1)()0)((0)(
111
≥?≥+=?== tNPtNSEtNPtNSESE∵
根据习题2.25的结果,
[]
[]
t
t
tt
e
te
tNSE
etNSEet
λ
λ
λλ
λ
λλ

=≥∴
≥+?+=∴
1
1
1)(
)1(1)()
1
(
1
1
1

}{}0{
}{}0{
1
1
11
1
1
1
)(
)(
)(
}1)({
txtx
t
x
txtx
II
e
e
II
tSP
xSP
tSxSP
tNxSP
>≤≤
>≤≤
+?
=
+?


=
≤≤=
≥≤
λ
λ

≤≤
=
≥∴
其他,0
0,
1
)1)((
1
tx
e
e
tNxf
t
x
S
λ
λ
λ
t
t
t
x
t
t
t
x
e
te
ex
e
x
e
ex
tNSE
λ
λ
λ
λλ
λ
λ
λ

=
=
=
≥∴
∫∫
1
1
)d(
1
1
d
1
}1)({
00
1
(3) ),(
21
SS在1)( =tN下的条件p.d.f,
注意到只有在
21
0 ttt <≤<的条件下,
0)1)(,(
211)(),(
21
≠=
=
tNttf
tNSS
用微元法来求解,对任意一个给定的t,令
21
tt <,取充分小的0>h,使得
2222
222111
h
tt
h
tt
h
tt
h
t +<<?<≤+<<?,

)
2
(
2
)
2
()
2
()
2
(
2
22
21
111
222111
222111
2
211
)](}1)
2
()
2
(
,0)()
2
(,0)
2
()(
,1)
2
()
2
(,0)
2
({[
1
)1)((
)
2222
(
)1)(
2222
(
tt
h
t
h
h
tt
h
tt
h
h
t
t
eh
t
te
heeehee
hO
h
tN
h
tN
tN
h
tN
h
tNtN
h
tN
h
tN
h
tNP
te
tNP
h
tS
h
tt
h
tS
h
tP
tN
h
tS
h
t
h
tS
h
tP
+?


=

=
+=+
==+?
=+==
=
+≤<?<≤+≤<?
=
=+≤<?<+≤<?
λ
λ
λ
λλ
λ
λ
λ
λ
λ
λλ
λ

}0{
)(
211)(),(
21
2
21
)1)(,(
ttt
tt
tNSS
Ie
t
tNttf
<≤<

=
==∴
λ
λ