In this section, we present an overview of the index of reducibility and
its uniform bound on several classes of modules studied in this thesis,
including Cohen-Macaulay, generalized Cohen-Macaulay, and sequentially
Cohen-Macaulay modules.
Recall that a proper R-submodule N of M is called irreducible if N
cannot be written as the intersection of two larger R-submodules of M.
Every R-submodule N of M can be expressed as an irredundant intersection
of irreducible R-submodules of M, and the number of irreducible
R-submodules appearing in such an expression depends only on N and not
on the expression (see [60]).
Definition 1.3.1. The number of irreducible R-submodules of M that
appear in an irredundant irreducible decomposition of N is called the index
of reducibility of N on M and denoted by irM(N). For a parameter ideal
q of M, the index of reducibility of q on M is the index of reducibility qM
on M and denoted by irM(q).
117 trang |
Chia sẻ: Kim Linh 2 | Ngày: 09/11/2024 | Lượt xem: 11 | Lượt tải: 0
Bạn đang xem trước 20 trang tài liệu Luận án Các hệ số hilbert trong vành cohen-macaulay và một số lớp mở rộng, để xem tài liệu hoàn chỉnh bạn click vào nút DOWNLOAD ở trên
→ 0
82
and S is Cohen-Macaulay, we get the following commutative diagram
0 // u/qu ι //R/q
ϕdq,R
ϵ // S/qS //
_
ϕdq,S
0
Hdm(R)
∼= //Hdm(S)
with the exact first row. Let x ∈ 0 :S/qS m. Then, since ϕdq,R is surjective
on the socles, we get an element y ∈ 0 :S/qS m such that ϕdq,S(x) = ϕdq,R(y).
Thus ϵ(y) = x, because the canonical map ϕdq,S is injective. Therefore
[u+ q] : m = q : m+ u. Thanks to Lemma 2.1.8, it can be proven that
[qn + u] :R m = qn :R m+ u (4.1)
for all n ≥ 1. Since S is Cohen-Macaulay, we have the following exact
sequence
0→ u/qn+1u→ R/qn+1R→ S/qn+1S → 0.
It follows from (4.1) and applying HomR(k, ∗) to the above sequence that
we obtain the exact sequence
0→ HomR(k, u/qn+1u)→ HomR(k,R/qn+1)→ HomR(k, S/qn+1S)→ 0.
Thus
ℓR([qn+1 :R m]/qn+1) = ℓR([qn+1u :u m]/qn+1u)+ℓR([(qn+1+u) :R m]/qn+1+u).
Since q is m-primary, by [13, Lemma 4.2], the degree of the polynomials
ℓR([qn+1u :u m]/qn+1u) is dim u − 1. Since dim u < dimR and the degree
of the polynomials ℓR([qn+1 :R m]/qn+1), ℓR([(qn+1 + u) :R m]/(qn+1R+ u))
are dimR − 1, the conclusion follows by comparing the coefficients of the
polynomials in the above inequality, as required.
By [53, Lemma 2.1], we get that
f0(q, R) = f0(q, S) = rd(S) = rd(R).
83
On the other hand, by [53, Theorem 1.1], we have
e1(qS : mS, S)− e1(qS, S) = rd(S) = rd(R).
Since (q : m)S ⊆ qS : mS, we have ℓR(S/(q : m)n+1S) ≥ ℓR(S/(qS :
mS)n+1), for all n ≥ 0. Since S is a regular local ring and q ⊆ m2, by [5,
Theorem 2.1], we have qS : mS is integral over qS and so
e0(qS : mS, S) = e0((q : m)S, S) = e0(qS, S).
Therefore, we have
e1((q : m)S, S) ≤ e1(qS : mS, S).
Hence
e1((q : m)S, S)− e1(qS, S) ≤ rd(R).
On the other hand, by Lemma 1.1.8, we have
e1(q, R) =
e1(q, S) if dim u ≤ d− 2,
e1(q, S)− e0(q, u) if dim u = d− 1,
e1(q : m, R) =
e1(q : m, S) if dim u ≤ d− 2,
e1(q : m, S)− e0(q : m, u) if dim u = d− 1.
If dim u < d − 1, then we have e1(q : m, R) − e1(q, R) = e1(q : m, S) −
e1(q, S) ≤ rd(R) = f0(q, R), as required.
Now, we can assume that dim u = d− 1. Then we have
e1(q : m, R)− e1(q, R) ≤ rd(R)− e0(q : m, u) + e0(q, u).
By the Prime Avoidance Theorem, we can choose a parameter element
x ∈ q of R such that (x) ∪ u = 0. Put R = R/(x), and S = S/(x). Then
we have the following exact sequence
0→ u→ R→ S → 0.
84
Consequently,
e0(m, R) = e0(m, S) + e0(m, u) ≥ 2,
so that by [24, Proposition 2.3], qR : m = (q : m)R is integral over qR.
Thus, e0(q : m, R) = e0(q, R). Since x is regular in S, e0(q : m, S) =
e0(q, S). Hence, we have
e1(q : m)− e1(q) ≤ rd(R)− e0(q : m, u) + e0(q, u)
= rd(R)− (e0(q : m, R)− e0(q : m, S)) + (e0(q, R)− e0(q, S))
= rd(R) = f0(q, R).
Now we consider the case (a). Since dimR ≥ 2, by [24, Proposition
2.3], we have mIn = mqn for all n. Therefore In ⊆ qn : m for all n.
Consequently, we obtain
ℓR(R/qn+1)− ℓR(R/In+1) = ℓR(In+1/qn+1) ≤ ℓR((qn+1 : m/qn+1).
Hence we have
e1(q : m, R)− e1(q, R) ≤ f0(q, R),
as required.
In [52] and [53], H.L. Truong provided the characterizations of Cohen-
Macaulay rings in terms of their Chern numbers, irreducible multiplicities,
index of reducibility of a parameter of R, and the Cohen-Macaulay type
provided R is unmixed. Notice that the necessary and sufficient conditions
of these characterizations need to hold true for all parameter ideals of R.
Recently, N.T.T. Tam and H.L. Truong [48] gave the characterizations of
Cohen-Macaulay rings in terms of their Chern numbers, irreducible mul-
tiplicities, and the type, which was introduced by S. Goto and N. Suzuki
([25, Definition (01)]). Let us now state the main results of this section
and its corollaries.
85
Theorem 4.1.4. Assume that R is unmixed of dimension d ≥ 2. Then
the following statements are equivalent.
i) R is Cohen-Macaulay.
ii) For all C-parameter ideals q ⊆ m2, we have
N (R) = f0(q) = e1(q : m)− e1(q) = rd(R).
iii) For some C-parameter ideal q ⊆ m2 of R, we have
N (R) ≤ e1(q : m)− e1(q).
iv) For some C-parameter ideal q ⊆ m2 of R , we have
N (R) ≤ f0(q, R).
v) For some C-parameter ideal q ⊆ mg(R)+1 of R, we have
f0(q) ≤ rd(R).
vi) For some C-parameter ideal q ⊆ mg(R)+1 of R, we have
e1(q : m)− e1(q) ≤ rd(R).
Proof. i)⇒ ii), i)⇒ iii) and i)⇒ v) follow from Lemma 3.1.7.
iii)⇒ iv) and v)⇒ vi) are trivial.
iv)⇒ i) and vi)⇒ i) are immediate from Lemma 4.1.1 and 2.2.5.
The following result establishs the relationship among the stable
value N (R/u), Chern numbers e1(q : m), e1(q), irreducible multiplicities
f0(q), and the Cohen-Macaulay type rd(R), and also shows that the differ-
ence e1(q : m)− e1(q) ranges among only finitely many value, where q runs
over all C-parameter ideals of R.
86
Proposition 4.1.5. Assume that dimR ≥ 2 and q is a C-parameter ideal
of R. Then we have
N (R/u) ≥ f0(q) ≥ e1(q : m)− e1(q) ≥ rd(R),
if either R is non-regular or q ⊆ mg(R)+1.
Proof. Put S = R/u, IR = q :R m and IS = qS :S mS. Then S is unmixed
with dimS = d. If R is non-regular or q ⊆ mg(R)+1, then by Lemma 4.1.1
and 3.1.7, we obtain N (S) ≥ f0(q, S). Thus by [48, Lemma 2.1], Lemma
4.1.2 and Corollary 2.3.5, we have
N (S) ≥ f0(q, S) ≥ f0(q, R) ≥ e1(q :R m, R)− e1(q, R) ≥ rd(R),
and this completes the proof.
For every parameter ideal q of R, by [32, Theorem 3.6], we have
e1(q) ≤ 0. Thus we can establish the relationship among the above invari-
ants as follows.
Remark 4.1.6. Assume that dimR ≥ 2 and q is a C-parameter ideal of
R. Then we have
N (R/u) ≥ f0(q) ≥ e1(q : m)− e1(q) ≥ e1(q : m),
if either R is non-regular or q ⊆ mg(R)+1.
Let
Ξi(R) = {ei(q : m) | q is a C-parameter ideal of R},
for all i = 1, . . . , d. Now, we can give an upper bound for the set Ξ1(R).
Corollary 4.1.7. Assume that R is non-regular and dimR ≥ 2. Then
Ξ1(R) ⊆ (−∞,N (R/u)].
87
Theorem 4.1.8. Assume that R is a non-regular unmixed local ring of
dimension d ≥ 2. Then
Ξ1(R) ⊆ (−∞,N (R)].
Moreover, N (R) ∈ Ξ1(R) if and only if R is Cohen-Macaulay.
Proof. R is a non-regular unmixed local ring, by Remark 2.1.8, we have
Ξ1(R) ⊆ (−∞,N (R)]. Now, we assume that R is Cohen-Macaulay. Since
R is non-regular, by Lemma 3.1.7, we get e1(q : m) = N (R). Conversely,
assume that e1(q :R m, R) = N (R). By Prosition 4.1.5, we have
N (R) = e1(q : m) ≤ e1(q : m)− e1(q) ≤ f0(q) ≤ N (R).
Therefore N (R) = f0(q). Thus R is Cohen-Macaulay, because of Lemma
4.1.1. This completes the proof.
The following consequence provides a characterization of Gorenstein
rings.
Corollary 4.1.9. Assume that R is unmixed of dimension d ≥ 2. Then
the following statements are equivalent.
i) R is Gorenstein.
ii) We have N (R) = 1.
iii) For some C-parameter ideal q ⊆ mg(R)+1 of R, we have f0(q) = 1.
iv) For some C-parameter ideal q ⊆ mg(R) + 1, we have
e1(q : m)− e1(q) = 1.
Proof. Since R is Gorenstein, we obtain N (R) = r(R) = 1, that is i)⇒ ii)
is true. On the other hand, by Proposition 4.1.5, for all C-parameter ideal
q ⊆ mg(R)+1 of R, we get that
1 = N (R) ≥ f0(q, R) ≥ e1(q :R m)− e1(q) ≥ rd(R) = 1.
88
It means that , ii) ⇒ iii) and iii) ⇒ iv) hold. Now, we prove iv) ⇒ i).
Let q be a C-parameter ideal of R such that e1(q :R m) − e1(q) = 1 and
q ⊆ mg(R)+1. Then we have e1(q : m)− e1(q) ≤ rd(R). By Theorem 4.1.4,
R is Cohen-Macaulay. Therefore, we have
1 = e1(q :R m)− e1(q) = rd(R).
Hence, R is Gorenstein, as required.
4.2 Finiteness of sets of Chern numbers
In this section, we analyze finiteness of sets Ξi(R), for each i = 1, . . . , d.
And then we deduced that the local cohomology modules {Him(R)}i<d are
of finite length.
For each n, i ≥ 1, we put
Ωni (R) = {ei(q :R m, R) | q is a C-parameter ideal of R contained in mn}.
Note that Ωti(R) ̸= ∅ and Ωti(R) ⊆ Ωt′i (R) ⊆ Ξi(R), for all t ≥ t′ ≥ 1.
Moreover, we get by Lemma 3.1.7 that if R is a non-regular generalized
Cohen-Macaulay ring, then Ξi(R) and Ωni (R) are finite for all i ≥ 1 and
n ≥ 1.
Let x = x1, x2, . . . , xd be a C-system of parameters of R. Put q[n] =
(xn1 , xn2 , . . . , xnd) and I [n] = q[n] :R m, for all n ≥ 1. For each t ≥ 1, let
Ωtx,1(R) = {e1(I [n], R) | n ≥ t}.
Note that Ωtx,1(R) ̸= ∅ and Ωtx,1(R) ⊆ Ωt′x,1(R) for all t ≥ t′ ≥ 1.
Lemma 4.2.1. Assume that R is unmixed with dimR = d ≥ 2. Suppose
that there exist a C-system of parameters x = x1, . . . , xd of R and an
integer t ≥ 2 such that the set Ωtx,1(R) is finite. Let
h = max{|X| : X ∈ Ωtx,1(R)}+N (R).
89
Then mhHim(R) = 0 for all i ̸= d.
Proof. We use induction on the dimension d of R. Suppose that d = 2.
Since R is unmixed, R is generalized Cohen-Macaulay, and so N (R) =
2r1(R) + r2(R). We chosse n ≥ t. Put q = (xn1 , xn2) and I = q :R m. Then,
since x1, x2 is a C-system of parameters of R, by Lemma 3.1.7 we have
ℓ(H1m(R)) = −e1(I, R) + r1(R) + r2(R)
≤ |e1(I, R)|+N (R) ≤ h.
Hence, mhH1m(R) = 0.
Suppose that d ≥ 3 and our assertion holds true for d − 1. We
put A = R/(x1). Then dimA = d − 1 and by the fact 3.1.4, we have
AssA ⊂ AsshA ∪ {m}. It follows that UA0 = H0m(A). Let B = A/H0m(A).
Then B is unmixed.
Now, by Lemma 3.1.4, x′ = x2, x3, . . . , xd is a C-system of parameters
of A. Thus, x′ is also a C-system of parameters of B. On the other hand,
it follows from Lemma 3.1.4 that
[(xm2 , xm3 , . . . , xmd ) + H0m(A)] :A mA = [(xm2 , xm3 , . . . , xmd ) :A mA] + H0m(A),
for all m ≥ 1. Therefore, we have the following.
Claim 4.2.2. Ωtx′,1(B) ⊆ Ωtx,1(R).
Proof. For each m ≥ t, we put y2 = xm2 , . . . , yd = xmd , q = (y1, y2 . . . yd),
I = q :R m, q′ = (y2, . . . , yd). Since x is a C-system of parameters of R, x1
is a superficial element of R with respect to I and q. Therefore, by Lemma
1.1.7, we have
e1(I, R) = e1(I ′, A),
where I ′ = q′A :A mA = (q :R m)A. It follows from Lemma 1.1.8, d ≥ 3
and [q′ +H0m(A)] :A mA = [q′ :A mA] + H0m(A) that we have
e1(I ′, A) = e1(q′B :B mB,B).
90
Consequently, e1(q′B :B mB,B) = e1(I, R) ∈ Ωtx,1(R). Hence Ωtx′,1(B) ⊆
Ωtx,1(R), as required.
It follows from the above claim that the set Ωtx′,1(B) is finite. Thus
by the hypothesis inductive on d, we have mh′Him(B) = 0, for all i ̸= d− 1,
where h′ = max{|X| : X ∈ Ωtx′,1(B)} + N (B). On the other hand, by
Lemma 3.1.4, we have
N (R) = N (A) = N (B) + r0(A).
Therefore h′ ≤ h, and so mhHim(B) = 0, for all i ̸= d−1. Hence mhHim(A) =
0, for all 1 ≤ i ≤ d− 2.
Now since R is unmixed, by [33, Lemma 3.4], x1 is R-regular and
x1H1m(R) = 0. Therefore it follows from the exact sequence
0 //R x1 //R //A // 0
that we have the surjective maps Him(A) // // 0 :Hi+1m (R) x1 for all i ≤ d− 2
and the injective map H1m(R)
//H1m(A), since x1H1m(R) = 0. Thus we
have
mh[0 :Hi+1m (R) x1] = 0 and m
hH1m(R) = 0
for all 1 ≤ i ≤ d − 2. It follows from Him(R) = ⋃
n≥1
[0 :Him(R) m
n] that we
have mhHim(R) = 0 for all i ̸= d, as required.
Lemma 4.2.3. Assume that d ≥ 2. Suppose that R/u is generalized
Cohen-Macaulay. Then
q :R m = q :R/u m
for all parameter ideals q of R/u contained in m2.
Proof. Put S = R/u. Let q is a parameter of S, IR = q :R m and IS =
q :S mS. Then q is also a parameter of R. Since S is generalized Cohen-
Macaulay and q ⊆ m2, by Lemma 3.1.6, we have (IS)2 = qIS. Thus by [23,
91
Theorem 1.1], we obtain e1(IS, S) = ℓR(S/IS)− e0(q, S)+ e1(q, S). By [23,
Theorem 3.1], we have
ℓR(S/IS)− e0(q, S) = e1(IS, S)− e1(q, S)
≥ e1(IR, S)− e1(q, S) ≥ ℓR(S/IRS)− e0(q, S).
Then we have IS = IR since IR ⊆ IS.
Theorem 4.2.4. Assume that d ≥ 2. Then the following statements are
equivalent.
i) R/u is generalized Cohen-Macaulay and dim u ≤ d− 2.
ii) There exists an integer t ≥ 2 such that the set Ωt1(R) is finite.
Proof. i) ⇒ ii). Let q be a C-parameter ideals of R contained in mg(R)+1.
We get by Proposition 4.1.5 that
N (R/u) ≥ e1(q :R m, R)− e1(q, R) ≥ rd(R).
On the other hand, the set {e1(q, R) | q is a parameter of R} is finite be-
cause of [18, Theorem 4.5]. Thus Ωt1(R) is finite for all t ≥ g(R) + 1.
ii) ⇒ i). We put S = R/u and u = dim u. Then there exists a C-system
x = x1, . . . , xd of parameters of R such that
1) (x) ⊆ mt,
2) (xu+1, xu+2, . . . , xd)u = 0,
3) x is a C-system of parameters of S.
Let q[n] = (xn1 , . . . , xnd)R, I
[n]
R = q[n] :R m and I
[n]
S = q[n]S :S mS, where n ≥
t. We have I [n]R ⊆ I [n]S , and so we get ℓR(S/(I [n]S )m+1) ≤ ℓR(S/(I [n]R )m+1S)
92
for all m ≥ 0. Thus e1(I [n]S , S) ≥ e1(I [n]R , S), because I [n]S is integral over
q[n]S by [24, Proposition 2.3]. Therefore by Proposition 4.1.5, we have
N (S) ≥ e1(I [n]S , S)− e1(q[n], S) ≥ e1(I [n]R , S)− e1(q[n], S)
= e1(I [n]R , R)− e1(q[n], R) ≥ rd(R) = rd(S).
(4.2)
Since the set Ωt1(R) is finite, so is the set {e1(I [n]R , R) | n ≥ 0}. Thus the
set {e1(q[n]R , R) | n ≥ 0} is finite. By Lemma 1.1.8, we have
0 ≥ e1(q[n]R , S) ≥ e1(q[n]R , R).
Therefore, the set {e1(q[n]R , S) | n ≥ 0} is finite. By (4.2), the set Ωt′x,1(S) is
finite for some t′. It follows from S is unmixed and Lemma 4.2.1 that S is
generalized Cohen-Macaulay.
Now suppose that u = d− 1. It follows from Lemma 3.1.7 that
e1(I [n]S , S) = −
d−1∑
j=1
d− 2
j − 1
ℓR(Hjm(S))− d∑
j=1
d− i
j − 1
rj(S)
,
which is independent of the choice of n. On the other hand, we have
e0(I [n]R , u) = e0(q[n], u) = e0((xn1 , . . . , xnu), u) = nue0(q, u) ≥ nu.
Thus by Lemma 1.1.8, we get that
−e1(I [n]R , R) = −e1(I [n]R , S) + e0(I [n]R , u)
= −e1(I [n]S , S) + nue0(q, u),
which is in contradiction with the finiteness of Ωn1(R) and Ωn1(S). Hence
u ≤ d− 2.
Applying Theorem 4.2.4 we obtain the following result.
Theorem 4.2.5. Assume that R is a non-regular unmixed local ring of
dimension d = dimR ≥ 2. Then the following statements are equivalent.
93
i) R is generalized Cohen-Macaulay.
ii) The set Ξ1(R) is finite.
Proof. It is immediate from Theorem 4.2.4 and Lemma 3.1.7.
Theorem 4.2.6. Assume that R is a non-regular with dimR = d ≥ 2.
Then the following statements are equivalent.
i) R/H0m(R) is Cohen-Macaulay.
ii) The sets Ξi(R) are finite for all i ≥ 2 and there exists a C-parameter
ideal q ⊆ mg(R)+1 of R such that
e1(q : m)− e1(q) ≤ rd(R).
Proof. i) ⇒ ii). Let S = R/u. Then since S is Cohen-Macaulay, R is
generalized Cohen-Macaulay. Therefore since R is non-regular, by Lemma
3.1.7, we obtain Ξi(R) are finite for all i ≥ 2. Let q ⊆ mg(R)+1 be a
C-parameter ideals of R. Then by [16, Lemma 2.4], we can assume that
[u+ q] : m = u+ [q : m].
It follows from the Lemma 3.1.7 and 1.1.8 that
e1(q :R m, R)− e1(q, R) = e1(q :S m, S)− e1(q, S)
= e1(qS :S mS, S)− e1(q, S) = rd(S) = rd(R),
as required.
ii)⇒ i). Since
e1(q :R m, R)− e1(q, R) ≤ rd(R)
for some C-parameter ideals q ⊆ mg(R)+1 of R, by Lemma 2.2.5, S is
Cohen-Macaulay.
Suppose that u = dim u ≥ 1. Then there exists a C-system x =
x1, . . . , xd of parameters of R such that
94
1) (x) ⊆ m2,
2) (xu+1, xu+2, . . . , xd)u = 0,
3) x is a C-system of parameters of S.
Let q[n] = (xn1 , . . . , xnd)R, I
[n]
R = q[n] :R m and I
[n]
S = q[n]S :S mS for all
n ≥ 0. Since S is Cohen-Macaulay, we get by Lemma 4.2.3 that I [n]S = I [n]R
Moreover, it follows Lemma 3.1.7 that
ed−u(I [n]S , S) =
0 if u ≤ d− 2,
rd(S) if u = d− 1.
which is independent of the choice of n. On the other hand, we have
e0(I [n]R , u) = e0(q[n], u) = e0((xn1 , . . . , xnu), u) = nue0(q, u) ≥ nu.
Thus, by Lemma 1.1.8, we get that
(−1)d−ued−u(I [n]R , R) = (−1)d−ued−u(I [n]R , S) + e0(I [n]R , u)
= (−1)d−ued−u(I [n]S , S) + nue0(q, u)
≥ nu + rd(S),
which is in contradiction with the finiteness of Ξd−u(R). Hence u = 0, and
this complete the proof.
Theorem 4.2.7. Assume that R is a non-regular with dimR = d ≥ 2.
Then the following statements are equivalent.
i) R is generalized Cohen-Macaulay.
ii) The sets Ξi(R) are finite for all i = 1, 2, . . . , d.
Proof. i)⇒ ii) follows from Lemma 3.1.7.
ii) ⇒ i). We put S = R/u and u = dim u. Assume that u > 0. Since
95
Ξ1(R) is finite, it follows from Theorem 4.2.5 that S is generalized Cohen-
Macaulay. Then there exists a C-system x = x1, . . . , xd of parameters of
R such that
1) (x) ⊆ m2,
2) (xu+1, xu+2, . . . , xd)u = 0,
3) x is a C-system of parameters of S.
Let q[n] = (xn1 , . . . , xnd)R, I
[n]
R = q[n] :R m and I
[n]
S = q[n]S :S mS for all
n ≥ 0. Since S is generalized Cohen-Macaulay, we get by Lemma 4.2.3
that I [n]S = I
[n]
R . Moreover, it follows Lemma 3.1.7 that
ed−u(I [n]S , S) = (−1)d−u
u∑
j=1
u− 1
j − 1
ℓR(Hjm(S))− u+1∑
j=1
d− i
j − 1
rj(S)
,
which is independent of the integer n. On the other hand, we have
e0(I [n]R , u) = e0(q[n], u) = e0((xn1 , . . . , xnu), u) = nue0(q, u) ≥ nu.
Thus, by Lemma 1.1.8, we get that
(−1)d−ued−u(I [n]R , R) = (−1)d−ued−u(I [n]R , S) + e0(I [n]R , u)
= (−1)d−ued−u(I [n]S , S) + nue0(q, u),
which is in contradiction with the finiteness of Ξd−u(R) and Ξd−u(S). Hence
u = 0, as required.
Remark 4.2.8. i) If for all i = 1, . . . , d, we define
Ξi(R) = {ei(q : m) | q is a C-parameter ideal of M contained in m2},
then Corollary 4.1.7, Theorem 4.1.8, 4.2.5, 4.2.6, and 4.2.7 remains
valid even without the condition that R is non-regular.
96
ii) In [29], A. Koura and N. Taniguchi proved that the Chern numbers of
m-primary ideals in R range among only finitely many values if and
only if dimR = 1 and R/H0m(R) is analytically unramified.
Now we close this chapter with the following example of non-generated
Cohen-Macaulay local rings R with |Ξ1(R)| = 1. Moreover, this example
also shows that the statements (1), (4), and (5) in Theorem 4.1.4 will not
be equivalent if the unmixed condition is removed from the hypothesis.
Example 4.2.9. Let S = k[[X, Y, Z,W ]] be the formal power series ring
over a field k. Put R = S/[(X, Y, Z) ∩ (W )]. Then
(1) R is not unmixed with dimR = 3. Moreover, R is not generalized
Cohen-Macaulay.
(2) |Ξ1(R)| = 1.
(3) We have
f(q, R) = e1(q :R m)− e1(q, R) = rd(R),
for all parameter ideals q in R.
Proof. Using similar arguments as in the Example 2.3.8, we have R is
not unmixed with dimR = 3 and R is not generalized Cohen-Macaulay.
However, for all parameter ideals q of R, one has
ℓR(R/qn+1) = ℓR(A/qn+1) + ℓR(B/qn+1)
= ℓR(A/qA)
n+ 3
3
+ ℓR(B/qB)
n+ 1
1
and
ℓR([qn+1 :R m]/qn+1) = ℓR([qn+1 :A m]/qn+1) + ℓR([qn+1 :B m]/qn+1)
=
n+ 1
2
+ 1,
97
for all integers n ⩾ 0. Since ℓR(R/In+1) = ℓR(R/qn+1) − ℓR([qn+1 :R
m]/qn+1), we have e1(q :R m) = 0 + 1 = 1. Thus |Ξ1(R)| = 1. Moreover,
we have f(q, R) = e1(q :R m)− e1(q, R) = rd(R) = 1, as required.
98
Conclusion of Chapter 4
The main results of this chapter include:
• An upper bound for sets of Chern numbers e1(q : m, R) of local
rings R, where q is a C-parameter ideal of R (Corollary 4.1.7).
• Some characterizations of Cohen-Macaulayness of unmixed local
rings R in terms of Hilbert coefficients e1(q : m, R), e1(q, R), irreducible
multiplicity f0(q, R), the stable value N (R) and the Cohen-Macaulay type
r(R), where q is a C-parameter ideal of R (Theorem 4.1.4).
• Some characterizations of generalized Cohen-Macaulayness of local
rings R in terms of the behavior of Hilbert coefficients ei(q : m, R) (i =
1, . . . , d), where q is a C-parameter ideal of R. (Theorem 4.2.5 and 4.2.7).
99
Conclusions
The main results of this dissertation include:
• A characterization of sequential Cohen-Macaulayness of local rings
R in terms of Hilbert coefficients ei(q : m, R), ei(q, R) (i = 1, . . . , d), where
and q is a distinguished parameter ideal of R (Theorem 2.3.1).
• Some characterizations of Cohen-Macaulay local rings R in terms
of sectional genera sg(q : m, R), sg(q, R), where q is a C-parameter ideal
of R (Theorem 3.2.5, 3.2.4, 3.3.2 and 3.3.4)
• An upper bound for sets of Chern numbers e1(q : m, R) of local
rings R, where q is a C-parameter ideal of R (Corollary 4.1.7).
• Some characterizations of Cohen-Macaulayness, and generalized
Cohen-Macaulayness of local rings R in terms of the behavior of Hilbert
coefficients ei(q : m, R) (i = 1, . . . , d), and q is a C-parameter ideal of R.
(Theorem 4.1.8, Theorem 4.2.5, and Theorem 4.2.7).
100
List of author’s related publications
[1] K. Ozeki, H.L. Truong and H.N. Yen, "On Hilbert coefficients and
sequentially Cohen-Macaulay rings", Proc. Amer. Math. Soc. 150
(2022), no. 6, 2367–2383.
[2] S. Kumashiro, H.L. Truong and H.N. Yen, "On the sectional genera
and Cohen-Macaulay rings", to appear in Journal of Commutative
Algebra.
[3] H.L. Truong and H.N. Yen, "On the set of Chern numbers in local
rings", to appear in Acta Math. Vietnam.
101
The results of this dissertation have
been presented at
• “International Workshop on Commutative Algebra”, May 6–8,
2021, Hanoi, Vietnam.
• “The conference Algebra-Geometry-Topology”, October 21–23, 2021,
Thai Nguyen University of Education, Thai Nguyen City, Vietnam.
• The weekly seminar of the Department of Algebra and the Depart-
ment of Number Theory, Institute of Mathematics, Vietnam Academy of
Science and Technology.
• “The 11th Japan-Vietnam Joint seminar on Commutative Algebra,
by and for young mathematicians”, Marth 28–30, 2023, Hanoi, Vietnam.
102
References
In English
[1] M. Brodmann and R.Y. Sharp, (1998), Local cohomology: an algebraic
introduction with geometric applications, Cambridge University Press,
Cambridge.
[2] W. Bruns and J. Herzog, (1998), Cohen-Macaulay rings, Cambridge
University Press, Cambridge.
[3] A. Corso, C. Huneke, W.V. Vasconcelos, "On the integral closure of
ideals", Manuscripta Math., 95 (1998), no.3, 331–347.
[4] A. Corso, C. Polini, M.E. Rossi, "Depth of associated graded rings via
Hilbert coefficients of ideals", J. Pure Appl. Algebra, 201 (2005), no.
1, 126–141.
[5] A. Corso, C. Polini, and W. V. Vasconcelos, "Links of prime ideals",
Math. Proc. Cambridge Philos. Soc., 115 (1994), no. 3, 431–436.
[6] N.T. Cuong and D.T. Cuong, " On sequentially Cohen-Macaulay mod-
ules", Kodai Math. J., 30 (2007), no. 3, 409–428.
[7] N.T. Cuong and D.T. Cuong, "Local cohomology annihilators and
Macaulayfication", Acta Math. Vietnam., 42 (2017), no.1, 37–60.
[8] N.T. Cuong, S. Goto and H.L. Truong, "The equality I2 = qI in
sequentially Cohen–Macaulay rings", J. Algebra, 379 (2013), 50–79.
[9] N.T. Cuong, S. Goto, and H.L. Truong, "Hilbert coefficients and
sequentially Cohen–Macaulay modules", J. Pure Appl. Algebra, 217
(2013), no. 3, 470–480.
[10] N.T. Cuong and L.T. Nhan, "Pseudo Cohen-Macaulay and pseudo
generalized Cohen-Macaulay modules", J. Algebra, 267 (2003), no. 1,
156–177.
[11] N.T. Cuong and P.H. Quy, "A splitting theorem for local cohomology
and its applications", J. Algebra, 331 (2011), 512–522.
[12] N.T. Cuong and P.H. Quy, "On the structure of finitely generated mod-
ules over quotients of Cohen-Macaulay local rings", arXiv:1612.07638
[13] N.T. Cuong, P.H. Quy, and H.L. Truong, "On the index of reducibility
in Noetherian modules", J. Pure Appl. Algebra, 219 (2015), no. 10,
4510–4520.
[14] N.T. Cuong, P.H. Quy, and H.L. Truong, "The index of reducibility of
powers of a standard parameter ideal", J. Algebra Appl. , 18 (2019),
no. 3, 1950048, 17 pp.
[15] N.T. Cuong and P.H. Quy, "On the index of reducibility of parameter
ideals: the stable and limit Values", Acta Math. Vietnam., 45(2020),
no. 1, 249–260.
[16] N.T. Cuong and H.L. Truong, "Asymptotic behavior of parameter ide-
als in generalized Cohen-Macaulay modules", J. Algebra, 320 (2008),
no. 1, 158–168.
104
[17] L. Ghezzi, S. Goto, J.-Y. Hong, K. Ozeki, T.T. Phuong, W.V. Vascon-
celos, "Cohen-Macaulayness versus the vanishing of the first Hilbert
coefficient of parameter ideals", J. Lond. Math. Soc., (2)81 (2010), no.
3, 679–695.
[18] L. Ghezzi, S. Goto, J.-Y. Hong, K. Ozeki, T.T. Phuong, W.V. Vas-
concelos, "The chern numbers and euler characteristics of modules",
Acta Math. Vietnam., 40 (2015), no. 1, 37–60.
[19] S. Goto and K. Ozeki, "Buchsbaumness in local rings possessing con-
stant first Hilbert coefficients of parameters", Nagoya Math. J., 199
(2010), 95–105.
[20] S. Goto and K. Ozeki, "Uniform bounds for Hilbert coefficients of
parameters", Contemp. Math., 555 American Mathematical Society,
Providence, RI, (2011), 97–118.
[21] S. Goto and K. Ozeki, "Sectional genera of parameter ideals", J. Al-
gebra, 446 (2016), 58–76.
[22] S. Goto and Y. Nakamura, "Multiplicity and tight closures of param-
eters", J. Algebra , 244 (2001), no. 1, 302–311.
[23] S. Goto and K. Nishida, "Hilbert coefficients and Buchsbaumness of
associated graded rings", J. Pure Appl. Algebra, 181 (2003), no. 1,
61–74.
[24] S. Goto and H. Sakurai, "The equality I2 = QI in Buchsbaum rings",
Rend. Sem. Mat. Univ. Padova, 110 (2003), 25–56.
[25] S. Goto and N. Suzuki, "Index of reducibility of parameter ideals in a
local ring", J. Algebra, 87 (1984), no. 1, 53–88.
105
[26] A. Grothendieck, (1967), Local homology, Lect. Notes in Math.,
SpringerVerlag Berlin-Heidelberg-New York.
[27] S. Itoh, "Hilbert coefficients of integrally closed ideals", J. Algebra,
176 (1995), no. 2, 638–652.
[28] C. Huneke, "Hilbert functions and symbolic powers", Michigan Math.
J., 34 (1987), no.2, 293–318.
[29] A. Koura and N. Taniguchi, "Bounds for the first Hilbert coefficients
of m-primary ideals", Tokyo J. Math., 38 (2015), no. 1, 125–133.
[30] S. Kumashiro, H.L. Truong and H. N. Yen, "On the sectional genera
and Cohen-Macaulay rings", to appear in Journal of Commutative
Algebra.
[31] H. Matsumura, (1986), Commutative ring theory, Cambridge Univer-
sity Press, Cambridge.
[32] M. Mandal, B. Singh and J.K. Verma, "On some conjectures about
the Chern numbers of filtrations", J. Algebra, 325 (2011), 147–162.
[33] M. Morales and P.H. Quy, "A study of the length function of gener-
alized fractions of modules", Proc. Edinb. Math. Soc. , (2)60 (2017),
no. 3, 721–737
[34] M. Nagata, (1962), Local rings, Interscience New York, New York.
[35] D.G. Northcott and D. Rees, "Extensions and simplifications of the
theory or regular local rings", J. London Math. Soc., 32 (1957), 367–
374.
[36] D.G. Northcott, "On irreducible ideals in local rings", J. London Math.
Soc., 32 (1957), 82–88.
106
[37] D.G. Northcott, "A note on the coefficients of the abstract Hilbert
function", J. London Math. Soc., 35 (1960), 209–214.
[38] A. Ooishi, "∆-genera and sectional genera of commutative rings", Hi-
roshima Math. J., 17 (1987), no. 2, 361–327.
[39] K. Ozeki, H.L. Truong and H.N. Yen, "On Hilbert coefficients and
sequentially Cohen-Macaulay rings", Proc. Amer. Math. Soc., 150
(2022), no. 6, 2367–2383.
[40] M.E. Rossi and G. Valla, (2010), "Hilbert function of filtered modules",
Lect. Notes UMI 9, Springer-Verlag.
[41] J.D. Sally, "Hilbert coefficients and reduction number 2", J. Algebraic
Geom., 1 (1992), no. 2, 325–333.
[42] J.D. Sally, "Ideals whose Hilbert function and Hilbert polynomial agree
at n = 1", J. Algebra, 157 (1993), no. 2, 534–547.
[43] P. Schenzel, "On the dimension filtration and Cohen-Macaulay filtered
modules", Lecture Notes in Pure and Appl. Math., 206 Marcel Dekker,
Inc., New York, (1999), 245–264.
[44] R.P. Stanley, (1996), Combinatorics and Commutative Algebra, second
edition, Birkha¨user Boston, Boston, MA.
[45] J. Stu¨ckrad and W. Vogel, (1986), Buchsbaum rings and applications,
Springer-Verlag.
[46] N. Suzuki, "On quasi-buchsbaum modules. An application of theory
of flc-modules", Adv. Stud. Pure Math., 11 (1987), 215–243.
[47] I. Swanson and C. Huneke, (2006) Integral closure of ideals, rings and
modules, Cambridge University Press, Cambridge.
107
[48] N.T.T. Tam and H.L. Truong, "A note on Chern coefficients and
Cohen-Macaulay rings", Ark. Mat., 58 (2020), no. 1, 197–212.
[49] N. V. Trung, "Absolutely superficial sequence", Math. Proc. Cambrige
Phil. Soc., 93 (1983), 35–47.
[50] N.V. Trung, "Toward a theory of generalized Cohen-Macaulay mod-
ules", Nagoya Math. J., 102 (1986), 1–49.
[51] H.L. Truong, "Index of reducibility of distinguished parameter ideals
and sequentially Cohen-Macaulay modules", Proc. Amer. Math. Soc.,
141 (2013), no. 6, 1971–1978.
[52] H.L. Truong, "Index of reducibility of parameter ideals and Cohen-
Macaulay rings", J. Algebra, 415 (2014), 35–49.
[53] H.L. Truong, "Chern coefficients and Cohen-Macaulay rings", J. Alge-
bra, 490 (2017), 316–329.
[54] H.L. Truong, "The eventual index of reducibility of parameter ideals
and the sequentially Cohen-Macaulay property", Arch. Math. (Basel),
112 (2019), no. 5, 475–488.
[55] H.L. Truong and H. N. Yen, "On the set of Chern numbers in local
rings", to appear in Acta Math. Vietnam.
[56] W.V. Vasconcelos, "The Chern cofficients of local rings", Michigan
Math. J. , 57 (2008), 725–713.
[57] K. Yamagishi, "The associated graded modules of Buchsbaum mod-
ules with respect to Y-primary ideals in the equi-I-invariant case", J.
Algebra, 225 (2000), pp. 1–27.
108
[58] K. Yamagishi, "Buchsbaumness in Rees modules associated to ideals
of minimal multiplicity in the equi-I-invariant case", J. Algebra, 251
(2002), pp. 213–255.
[59] O. Zariski, P. Samuel, (1960), Commutative Algebra, Vol II, Van Nos-
trand, New York.
In German
[60] E. Noether, "Idealtheorie in Ringbereichen", Math. Ann., 83 (1921),
no. 1-2, 24–66.
[61] P. Schenzel, N.V. Trung, and N.T. Cuong, "Verallgemeinerte Cohen-
Macaulay-Moduln", Math. Nachr., 85 (1978), 57–73.
[62] P. Schenzel, (1982), "Dualisierende Komplexe in der lokalen Alge-
bra und Buchsbaum-Ringe", Lecture Notes in Mathematics, Springer-
Verlag, Berlin-New York.
[63] J. Stu¨ckrad and W. Vogel, "Toward a theory of Buchsbaum singulari-
ties", Amer. J. Math., 100 (1978), 727-746.
[64] W. Vogel, "u¨ber eine Vermutung von D. A. Buchsbaum", J. Algebra,
23 (1973), 106-112.
In Vietnamese
[65] P.H. Quý, Về tính chẻ ra của môđun đối đồng điều địa phương và một
số ứng dụng trong vành Noether, Luận án Tiến sỹ Toán học, Viện
Toán học - Viện Hàn lâm KHCN Việt Nam, 2013.