Small complete caps and saturating sets in Galois spaces, I
Massimo Giulietti
University of Perugia (Italy)
Workshop in Finite Geometry - Szeged 10-14 June 2013
Saturating sets and complete caps
Σ = Σ(N,q)
Galois space of dimension N over the finite field Fq
Σ
Saturating sets and complete caps
Σ = Σ(N,q)
Galois space of dimension N over the finite field Fq
bb b b b
b
b b
b b b
b b S
Σ
S ⊂Σ is asaturatingset if every point in Σ\S is collinear with two points in S
Saturating sets and complete caps
Σ = Σ(N,q)
Galois space of dimension N over the finite field Fq
bb b b b
b
b b
b b b
b b Sb
Σ
S ⊂Σ is asaturatingset if every point in Σ\S is collinear with two points in S
Saturating sets and complete caps
Σ = Σ(N,q)
Galois space of dimension N over the finite field Fq
bb b b b
b
b b
b b b
b b Sb
Σ
S ⊂Σ is asaturatingset if every point in Σ\S is collinear with two points in S
Saturating sets and complete caps
Σ = Σ(N,q)
Galois space of dimension N over the finite field Fq
bb b b b
b
b b
b b b
b b Sb
Σ
S ⊂Σ is asaturatingset if every point in Σ\S is collinear with two points in S
Saturating sets and complete caps
Σ = Σ(N,q)
Galois space of dimension N over the finite field Fq
bb b b b
b
b b
b b b
b b Sb
Σ
S ⊂Σ is asaturatingset if every point in Σ\S is collinear with two points in S
A complete capis a saturating set which does not contain 3 collinear points
Saturating sets and complete caps
Σ = Σ(N,q)
Galois space of dimension N over the finite field Fq
bb b b b
b
b b
b b b
b b Sb
Σ
S ⊂Σ is asaturatingset if every point in Σ\S is collinear with two points in S
A complete capis a saturating set which does not contain 3 collinear points
WhenN = 2 complete caps are usually calledcomplete arcs
Saturating sets and complete caps
Σ = Σ(N,q)
Galois space of dimension N over the finite field Fq
bb b b b
b
b b
b b b
b b Sb
Σ
S ⊂Σ is asaturatingset if every point in Σ\S is collinear with two points in S
A complete capis a saturating set which does not contain 3 collinear points
WhenN = 2 complete caps are usually calledcomplete arcs Problem: construct smallsaturating sets/complete caps
Classical examples of complete caps
in PG(2,q), q odd, an irreducible conic (q+ 1 points)
Classical examples of complete caps
in PG(2,q), q odd, an irreducible conic (q+ 1 points) in PG(2,q), q even, an irreducible conic plus its nucleus (q+ 2 points)
Classical examples of complete caps
in PG(2,q), q odd, an irreducible conic (q+ 1 points) in PG(2,q), q even, an irreducible conic plus its nucleus (q+ 2 points)
in PG(3,q) an elliptic quadric (q2+ 1 points)
Classical examples of complete caps
in PG(2,q), q odd, an irreducible conic (q+ 1 points) in PG(2,q), q even, an irreducible conic plus its nucleus (q+ 2 points)
in PG(3,q) an elliptic quadric (q2+ 1 points) Classification results
(Segre, 1954) In PG(2,q), q odd, every complete arc of size q+ 1 is an irreducible conic
Classical examples of complete caps
in PG(2,q), q odd, an irreducible conic (q+ 1 points) in PG(2,q), q even, an irreducible conic plus its nucleus (q+ 2 points)
in PG(3,q) an elliptic quadric (q2+ 1 points) Classification results
(Segre, 1954) In PG(2,q), q odd, every complete arc of size q+ 1 is an irreducible conic
(Barlotti-Panella, 1955) In PG(3,q), q odd, every complete cap of size q2+ 1 is an irreducible quadric
Classical examples of complete caps
in PG(2,q), q odd, an irreducible conic (q+ 1 points) in PG(2,q), q even, an irreducible conic plus its nucleus (q+ 2 points)
in PG(3,q) an elliptic quadric (q2+ 1 points) Classification results
(Segre, 1954) In PG(2,q), q odd, every complete arc of size q+ 1 is an irreducible conic
(Barlotti-Panella, 1955) In PG(3,q), q odd, every complete cap of size q2+ 1 is an irreducible quadric
(Penttila, 2013) In PG(3,q),q an even square, every complete cap of size q2+ 1 is an irreducible quadric
Covering Codes
(Fnq,d) d Hamming distance C ⊂Fnq
Covering Codes
(Fnq,d) d Hamming distance C ⊂Fnq
Covering Radius of C
R(C):= max
v∈Fnq
d(v,C)
Covering Codes
(Fnq,d) d Hamming distance C ⊂Fnq
Covering Radius of C
R(C):= max
v∈Fnq
d(v,C) Covering Density of C
µ(C):= #C ·size of a sphere of radiusR(C) qn
Covering Codes
(Fnq,d) d Hamming distance C ⊂Fnq
Covering Radius of C
R(C):= max
v∈Fnq
d(v,C) Covering Density of C
µ(C):= #C ·size of a sphere of radiusR(C) qn
Problem: findlinearsubspacesC withµ(C) close to 1
Linear Codes with R = 2
k = dimC r =n−k µ(C) =1+n(q−1)+(n2)(q−1)2 qr
Linear Codes with R = 2
k = dimC r =n−k µ(C) =1+n(q−1)+(n2)(q−1)2 qr
Length function
ℓ(r,q) := minn for which there existsC ⊂Fnq with R(C) = 2, n−dim(C) =r
Linear Codes with R = 2
k = dimC r =n−k µ(C) =1+n(q−1)+(n2)(q−1)2 qr
Length function
ℓ(r,q)d := minn for which there existsC ⊂Fnq with R(C) = 2, n−dim(C) =r, d(C) =d
Linear Codes with R = 2
k = dimC r =n−k µ(C) =1+n(q−1)+(n2)(q−1)2 qr
Length function
ℓ(r,q)d := minn for which there existsC ⊂Fnq with R(C) = 2, n−dim(C) =r, d(C) =d Proposition
ℓ(r,q)4 =minimum size of a complete cap in PG(r−1,q) ℓ(r,q)3 =minimum size of a saturating set in PG(r−1,q)
ℓ(r,q)1 =ℓ(r,q)2 =ℓ(r,q)3+ 1
Linear Codes with R = 2
k = dimC r =n−k µ(C) =1+n(q−1)+(n2)(q−1)2
qr ≥1
Length function
ℓ(r,q)d := minn for which there existsC ⊂Fnq with R(C) = 2, n−dim(C) =r, d(C) =d Proposition
ℓ(r,q)4 =minimum size of a complete cap in PG(r−1,q) ℓ(r,q)3 =minimum size of a saturating set in PG(r−1,q)
ℓ(r,q)1 =ℓ(r,q)2 =ℓ(r,q)3+ 1 Trivial Lower Bound (TLB)
ℓ(r,q)≥√ 2qr−22
Linear Codes with R = 2
k = dimC r =n−k µ(C) =1+n(q−1)+(n2)(q−1)2
qr ≥1
Length function
ℓ(r,q)d := minn for which there existsC ⊂Fnq with R(C) = 2, n−dim(C) =r, d(C) =d Proposition
ℓ(r,q)4 =minimum size of a complete cap in PG(r−1,q) ℓ(r,q)3 =minimum size of a saturating set in PG(r−1,q)
ℓ(r,q)1 =ℓ(r,q)2 =ℓ(r,q)3+ 1 Trivial Lower Bound (TLB)
ℓ(r,q)≥√
2qr−22 (#S ≥√
2q(N−1)/2 inΣ(N,q))
Naive construction method:
Naive construction method:
S ={P1, P2,
bb
P1 P2
Naive construction method:
S ={P1, P2, P3,
bb
P1 P2
bP3
Naive construction method:
S ={P1, P2, P3, P4,
bb
P1 P2
bP3
b
P4
Naive construction method:
S ={P1, P2, P3, P4, . . . ,
bb
P1 P2
bP3
b
P4 b
b
b b
b b
bb
bb b b
Naive construction method:
S ={P1, P2, P3, P4, . . . , Pn}
bb
P1 P2
bP3
b
P4 b
b
b b
b b
bb
bb b b
Naive vs. Clever
Naive construction method:
S ={P1, P2, P3, P4, . . . , Pn}
bb
P1 P2
bP3
b
P4 b
b
b b
b b
bb
bb b b
Naive vs. Clever : when applicable, the naive method in most cases produces saturating sets/complete caps that are smaller than any theoretically obtained ones (even with probabilistic methods)
Outline
The plane case
The 3-dimensional case
Recursive constructions of complete caps Plane saturating sets
Recursive constructions of saturating sets
Outline
The plane case
The 3-dimensional case
Recursive constructions of complete caps Plane saturating sets
Recursive constructions of saturating sets
Improvements on the TLB for complete arcs
S complete arc in PG(2,q)
Improvements on the TLB for complete arcs
S complete arc in PG(2,q)
TLB:
#S >p 2q+ 1
Improvements on the TLB for complete arcs
S complete arc in PG(2,q)
TLB:
#S >p 2q+ 1 If q =ph withh ∈ {1,2,3}
Improvements on the TLB for complete arcs
S complete arc in PG(2,q)
TLB:
#S >p 2q+ 1 If q =ph withh ∈ {1,2,3}
#S >p 3q+1
2
(Blokhuis, 1994 - Ball, 1997 - Polverino, 1999)
Probabilistic and computational results
In PG(2,q) there exists a saturating S set of size
#S ≤5p qlogq (Kov´acs, 1992)
Probabilistic and computational results
In PG(2,q) there exists a saturating S set of size
#S ≤5p qlogq (Kov´acs, 1992)
In PG(2,q) there exists a complete arcS of size
#S ≤D√qlogCq (Kim-Vu, 2003)
Probabilistic and computational results
In PG(2,q) there exists a saturating S set of size
#S ≤5p qlogq (Kov´acs, 1992)
In PG(2,q) there exists a complete arcS of size
#S ≤D√qlogCq (Kim-Vu, 2003)
Polynomial type algorithm
Probabilistic and computational results
In PG(2,q) there exists a saturating S set of size
#S ≤5p qlogq (Kov´acs, 1992)
In PG(2,q) there exists a complete arcS of size
#S ≤D√qlogCq (Kim-Vu, 2003)
Polynomial type algorithm
For every q≤42013 there exists a complete arcS of size
#S ≤√qlogq
(Bartoli-Davydov-Faina-Marcugini-Pambianco, 2012)
Algebraic method
S parametrized by polynomials defined over Fq
S ={(f(t),g(t))|t ∈Fq} ⊂AG(2,q)
Algebraic method
S parametrized by polynomials defined over Fq
S ={(f(t),g(t))|t ∈Fq} ⊂AG(2,q) P = (a,b) collinear with two points inS if there exist x,y ∈Fq with
det
a b 1
f(x) g(x) 1 f(y) g(y) 1
= 0
Algebraic method
S parametrized by polynomials defined over Fq
S ={(f(t),g(t))|t ∈Fq} ⊂AG(2,q) P = (a,b) collinear with two points inS if there exist x,y ∈Fq withFa,b(x,y) = 0, where
Fa,b(x,y) :=det
a b 1
f(x) g(x) 1 f(y) g(y) 1
Algebraic method
S parametrized by polynomials defined over Fq
S ={(f(t),g(t))|t ∈Fq} ⊂AG(2,q) P = (a,b) collinear with two points inS if there exist x,y ∈Fq withFa,b(x,y) = 0, where
Fa,b(x,y) :=det
a b 1
f(x) g(x) 1 f(y) g(y) 1
P = (a,b) collinear with two points inS if the algebraic curve CP :Fa,b(X,Y) = 0
has a suitableFq-rational point (x,y)
Appying Hasse-Weil Theorem
Theorem
Let C be an irreducible curve defined over Fq of degree d . Then
#CP(Fq)≥q+ 1−2g√q, where g ≤(d−1)(d−2)/2 is the genus of C
Appying Hasse-Weil Theorem
Theorem
Let C be an irreducible curve defined over Fq of degree d . Then
#CP(Fq)≥q+ 1−2g√q, where g ≤(d−1)(d−2)/2 is the genus of C The method works when, apart from few P’s,
CP is irreducible (or admits an irreducible component defined over Fq)
The genus g of CP (or of its component) is not too large with respect to q
Example
d a divisor of q−1
S ={(
f(t)
z}|{
td ,
g(t)
z}|{
t3d )
| {z }
Pt
|t∈Fq}, #S = q−1 d
Example
d a divisor of q−1
S ={(
f(t)
z}|{
td ,
g(t)
z}|{
t3d )
| {z }
Pt
|t∈Fq}, #S = q−1 d
P = (a,b) is collinear withPx andPy if and only if Fa,b(x,y) :=b−ay3d−x3d
yd−xd +xdydy2d−x2d yd−xd = 0
Example
d a divisor of q−1
S ={(
f(t)
z}|{
td ,
g(t)
z}|{
t3d )
| {z }
Pt
|t∈Fq}, #S = q−1 d
P = (a,b) is collinear withPx andPy if and only if Fa,b(x,y) :=b−ay3d−x3d
yd−xd +xdydy2d−x2d yd−xd = 0 the curve CP then isFa,b(X,Y) = 0
Example
d a divisor of q−1
S ={(
f(t)
z}|{
td ,
g(t)
z}|{
t3d )
| {z }
Pt
|t∈Fq}, #S = q−1 d
P = (a,b) is collinear withPx andPy if and only if Fa,b(x,y) :=b−ay3d−x3d
yd−xd +xdydy2d−x2d yd−xd = 0 the curve CP then isFa,b(X,Y) = 0
Fa,b(X,Y) =b−a(Y2d+YdXd +X2d) +Y2dXd+YdX2d
Applying Segre’s criterion
(Segre, 1962)
Let C:f(X,Y) = 0.If there exists a pointP ∈ C and a tangentℓ of C atP such that
ℓcounts once among the tangents of C atP,
the intersection multiplicity of C andℓat P equals deg(f), C has no linear components through P,
then C is irreducible.
Applying Segre’s criterion
(Segre, 1962)
Let C:f(X,Y) = 0.If there exists a pointP ∈ C and a tangentℓ of C atP such that
ℓcounts once among the tangents of C atP,
the intersection multiplicity of C andℓat P equals deg(f), C has no linear components through P,
then C is irreducible.
Fa,b(X,Y) =b−a(Y2d+YdXd+X2d) +Y2dXd+YdX2d
Applying Segre’s criterion
(Segre, 1962)
Let C:f(X,Y) = 0.If there exists a pointP ∈ C and a tangentℓ of C atP such that
ℓcounts once among the tangents of C atP,
the intersection multiplicity of C andℓat P equals deg(f), C has no linear components through P,
then C is irreducible.
Fa,b(X,Y) =b−a(Y2d+YdXd+X2d) +Y2dXd+YdX2d At P =Y∞ the tangents are ℓ:X =α withαd =a
Applying Segre’s criterion
(Segre, 1962)
Let C:f(X,Y) = 0.If there exists a pointP ∈ C and a tangentℓ of C atP such that
ℓcounts once among the tangents of C atP,
the intersection multiplicity of C andℓat P equals deg(f), C has no linear components through P,
then C is irreducible.
Fa,b(X,Y) =b−a(Y2d+YdXd+X2d) +Y2dXd+YdX2d At P =Y∞ the tangents are ℓ:X =α withαd =a
Fa,b(α,Y) = b−a(Y2d+aYd+a2) +aY2d+a2Yd =b−a36= 0
CP :Fa,b(X,Y) = 0 is irreducible of genus g ≤(3d −1)(3d −2)/2
CP :Fa,b(X,Y) = 0 is irreducible of genus g ≤(3d −1)(3d −2)/2 Hasse-Weil bound
#CP(Fq)≥q+ 1−√q(3d −1)(3d −2)
CP :Fa,b(X,Y) = 0 is irreducible of genus g ≤(3d −1)(3d −2)/2 Hasse-Weil bound
#CP(Fq)≥q+ 1−√q(3d −1)(3d −2) We need
CP :Fa,b(X,Y) = 0 is irreducible of genus g ≤(3d −1)(3d −2)/2 Hasse-Weil bound
#CP(Fq)≥q+ 1−√q(3d −1)(3d −2) We need
q+ 1−√q(3d −1)(3d −2) >3d2+ 9d
CP :Fa,b(X,Y) = 0 is irreducible of genus g ≤(3d −1)(3d −2)/2 Hasse-Weil bound
#CP(Fq)≥q+ 1−√q(3d −1)(3d −2) We need
q+ 1−√q(3d −1)(3d −2) >3d2+ 9d
This is implied by
d <
√17√4q 6
CP :Fa,b(X,Y) = 0 is irreducible of genus g ≤(3d −1)(3d −2)/2 Hasse-Weil bound
#CP(Fq)≥q+ 1−√q(3d −1)(3d −2) We need
q+ 1−√q(3d −1)(3d −2) >3d2+ 9d
This is implied by
d <
√17√4q 6
Other points? Covered if some extra conditions on d andq are satisfied ((q−1)/d even and 3|d) andX∞is added toS
Example
d a divisor of q−1
S ={(
f(t)
z}|{
td ,
g(t)
z}|{
t3d )
| {z }
Pt
|t∈Fq}, #S = q−1 d S is saturating if
d <
√17√4q 6
Example
d a divisor of q−1
S ={(
f(t)
z}|{
td ,
g(t)
z}|{
t3d )
| {z }
Pt
|t∈Fq}, #S = q−1 d S is saturating if
d <
√17√4q 6 In the best case,
S ∼ 6
√17q3/4
Cubic curves
X plane irreducible cubic curve
Cubic curves
X plane irreducible cubic curve
•Q
•P
G =X(Fq)\Sing(X)
•
•P⊕Q T
• O
If O is an inflection point ofX, then P,Q,T ∈G are collinear if and only if
P ⊕Q⊕T = 0
Cubic curves
X plane irreducible cubic curve
•Q
•P
G =X(Fq)\Sing(X)
•
•P⊕Q T
• O
If O is an inflection point ofX, then P,Q,T ∈G are collinear if and only if
P ⊕Q⊕T = 0
For a subgroup K of indexm with (3,m) = 1, no 3 points in a coset
S =K ⊕Q, Q ∈/ K are collinear
Classification ( p > 3)
Classification ( p > 3)
Classification ( p > 3)
•
Classification ( p > 3)
•
Classification ( p > 3)
Y =X3 XY = (X −1)3
•
Y(X2−β) = 1 Y2 =X3+AX+B
Cuspidal case: Y = X
3G is an elementary abelianp-group q =ph
Cuspidal case: Y = X
3G is an elementary abelianp-group q =ph K ={(tp−t,(tp−t)3)|t ∈Fq}
Cuspidal case: Y = X
3G is an elementary abelianp-group q =ph K ={(tp−t,(tp−t)3)|t ∈Fq}
S ={(
f(t)
z }| { tp−t+ ¯t,(
g(t)
z }| { tp−t+ ¯t)3)
| {z }
Pt
|t ∈Fq}
Cuspidal case: Y = X
3G is an elementary abelianp-group q =ph K ={(tp−t,(tp−t)3)|t ∈Fq}
S ={(
f(t)
z }| { tp−t+ ¯t,(
g(t)
z }| { tp−t+ ¯t)3)
| {z }
Pt
|t ∈Fq}
P = (a,b) is collinear withPx andPy if and only if Fa,b(x,y) := a+ (xp−x+t)(yp−y+t)2+
(xp−x+t)2(yp−y+t)−b((xp−x+t)2 +(xp−x+t)(yp−y+t) + (yp−y+t)2) = 0
Cuspidal case: Y = X
3G is an elementary abelianp-group q =ph K ={(tp−t,(tp−t)3)|t ∈Fq}
S ={(
f(t)
z }| { tp−t+ ¯t,(
g(t)
z }| { tp−t+ ¯t)3)
| {z }
Pt
|t ∈Fq}
P = (a,b) is collinear withPx andPy if and only if Fa,b(x,y) := a+ (xp−x+t)(yp−y+t)2+
(xp−x+t)2(yp−y+t)−b((xp−x+t)2 +(xp−x+t)(yp−y+t) + (yp−y+t)2) = 0 the curve CP then isFa,b(X,Y) = 0
Applying Segre’s criterion
Fa,b(X,Y) := a+ (Xp−X+t)(Yp−Y +t)2+ (Xp−X+t)2(Yp−Y +t)−b((Xp−X +t)2 +(Xp−X +t)(Yp−Y +t) + (Yp−Y +t)2) = 0
Applying Segre’s criterion
Fa,b(X,Y) := a+ (Xp−X+t)(Yp−Y +t)2+ (Xp−X+t)2(Yp−Y +t)−b((Xp−X +t)2 +(Xp−X +t)(Yp−Y +t) + (Yp−Y +t)2) = 0 At P =X∞ the tangents areℓ:Y =β withβp−β+t=b
Applying Segre’s criterion
Fa,b(X,Y) := a+ (Xp−X+t)(Yp−Y +t)2+ (Xp−X+t)2(Yp−Y +t)−b((Xp−X +t)2 +(Xp−X +t)(Yp−Y +t) + (Yp−Y +t)2) = 0 At P =X∞ the tangents areℓ:Y =β withβp−β+t=b Fa,b(X, β) = a+ (Xp−X+t)β2+ (Xp−X+t)2β−
b((Xp−X +t)2+ (Xp−X +t)β+β2) =a−b3
Applying Segre’s criterion
Fa,b(X,Y) := a+ (Xp−X+t)(Yp−Y +t)2+ (Xp−X+t)2(Yp−Y +t)−b((Xp−X +t)2 +(Xp−X +t)(Yp−Y +t) + (Yp−Y +t)2) = 0 At P =X∞ the tangents areℓ:Y =β withβp−β+t=b Fa,b(X, β) = a+ (Xp−X+t)β2+ (Xp−X+t)2β−
b((Xp−X +t)2+ (Xp−X +t)β+β2) =a−b3
If P ∈ X/
CP is irreducible of genus g ≤3p2−3p+ 1
Applying Segre’s criterion
Fa,b(X,Y) := a+ (Xp−X+t)(Yp−Y +t)2+ (Xp−X+t)2(Yp−Y +t)−b((Xp−X +t)2 +(Xp−X +t)(Yp−Y +t) + (Yp−Y +t)2) = 0 At P =X∞ the tangents areℓ:Y =β withβp−β+t=b Fa,b(X, β) = a+ (Xp−X+t)β2+ (Xp−X+t)2β−
b((Xp−X +t)2+ (Xp−X +t)β+β2) =a−b3
If P ∈ X/
CP is irreducible of genus g ≤3p2−3p+ 1 CP has at least q+ 1−(6p2−6p+ 2)√q points
Cuspidal case: Y = X
3G is elementary abelian, isomorphic to (Fq,+)
Cuspidal case: Y = X
3G is elementary abelian, isomorphic to (Fq,+) S ={(L(t) + ¯t,(L(t) + ¯t)3)
| {z }
Pt
|t∈Fq}
L(T) = Y
α∈M
(T −α), M <(Fq,+), #M =m
Cuspidal case: Y = X
3G is elementary abelian, isomorphic to (Fq,+) S ={(L(t) + ¯t,(L(t) + ¯t)3)
| {z }
Pt
|t∈Fq}
L(T) = Y
α∈M
(T −α), M <(Fq,+), #M =m P = (a,b) is collinear withPx andPy if and only if
Fa,b(x,y) := a+ (L(x) +t)(L(y) +t)2+ (L(x) +t)2(L(y) +t)−b((L(x) +t)2 +(L(x) +t)(L(y) +t) + (L(y) +t)2) = 0
Cuspidal case: Y = X
3G is elementary abelian, isomorphic to (Fq,+) S ={(L(t) + ¯t,(L(t) + ¯t)3)
| {z }
Pt
|t∈Fq}
L(T) = Y
α∈M
(T −α), M <(Fq,+), #M =m P = (a,b) is collinear withPx andPy if and only if
Fa,b(x,y) := a+ (L(x) +t)(L(y) +t)2+ (L(x) +t)2(L(y) +t)−b((L(x) +t)2 +(L(x) +t)(L(y) +t) + (L(y) +t)2) = 0 If P ∈ X/
CP is irreducible of genus g ≤3m2−3m+ 1
Cuspidal case: Y = X
3G is elementary abelian, isomorphic to (Fq,+) S ={(L(t) + ¯t,(L(t) + ¯t)3)
| {z }
Pt
|t∈Fq}
L(T) = Y
α∈M
(T −α), M <(Fq,+), #M =m P = (a,b) is collinear withPx andPy if and only if
Fa,b(x,y) := a+ (L(x) +t)(L(y) +t)2+ (L(x) +t)2(L(y) +t)−b((L(x) +t)2 +(L(x) +t)(L(y) +t) + (L(y) +t)2) = 0 If P ∈ X/
CP is irreducible of genus g ≤3m2−3m+ 1 CP has at least q+ 1−(6m2−6m+ 2)√q points
(Sz˝onyi, 1985 - Anbar, Bartoli, G., Platoni, 2013) Let P = (a,b) be a point in AG(2,q)\ X. If
m<p4 q/36 then there is a secant of S passing through P.
(Sz˝onyi, 1985 - Anbar, Bartoli, G., Platoni, 2013) Let P = (a,b) be a point in AG(2,q)\ X. If
m<p4 q/36 then there is a secant of S passing through P.
m is a power ofp
(Sz˝onyi, 1985 - Anbar, Bartoli, G., Platoni, 2013) Let P = (a,b) be a point in AG(2,q)\ X. If
m<p4 q/36 then there is a secant of S passing through P.
m is a power ofp
the points inX \S need to be dealt with
(Sz˝onyi, 1985 - Anbar, Bartoli, G., Platoni, 2013) Let P = (a,b) be a point in AG(2,q)\ X. If
m<p4 q/36 then there is a secant of S passing through P.
m is a power ofp
the points inX \S need to be dealt with Theorem
If m<p4
q/36, then there exists a complete arc in AG(2,q) with size
(2√m−3)q
m, if m is a square, (√m/p+√mp−3)q
m, otherwise.
(Sz˝onyi, 1985 - Anbar, Bartoli, G., Platoni, 2013) Let P = (a,b) be a point in AG(2,q)\ X. If
m<p4 q/36 then there is a secant of S passing through P.
m is a power ofp
the points inX \S need to be dealt with Theorem
If m<p4
q/36, then there exists a complete arc in AG(2,q) with size
(2√m−3)q
m, if m is a square, (√m/p+√mp−3)q
m, otherwise.
∼q7/8
Nodal case: XY = ( X − 1)
3G is isomorphic to (F∗q,·)
Nodal case: XY = ( X − 1)
3G is isomorphic to (F∗q,·)
the subgroup of index m(ma divisor of q−1):
K =n
tm,(tm−1)3 tm
|t ∈F∗q
o
Nodal case: XY = ( X − 1)
3G is isomorphic to (F∗q,·)
the subgroup of index m(ma divisor of q−1):
K =n
tm,(tm−1)3 tm
|t ∈F∗q
o
a coset:
S =n
¯ttm,(¯ttm−1)3
¯ttm
|t∈F∗q
o
Nodal case: XY = ( X − 1)
3G is isomorphic to (F∗q,·)
the subgroup of index m(ma divisor of q−1):
K =n
tm,(tm−1)3 tm
|t ∈F∗q
o
a coset:
S =n
¯ttm,(¯ttm−1)3
¯ttm
|t∈F∗q
o
the curve CP:
Fa,b(X,Y) = a(t3X2mYm+t3XmY2m−3t2XmYm+ 1)
−bt2XmYm−t4X2mY2m+ 3t2XmYm
−tXm−tYm = 0
Nodal case: XY = ( X − 1)
3G is isomorphic to (F∗q,·)
the subgroup of index m(ma divisor of q−1):
K =n
tm,(tm−1)3 tm
|t ∈F∗q
o
a coset:
S =n
¯ttm,(¯ttm−1)3
¯ttm
|t∈F∗q
o
the curve CP:
Fa,b(X,Y) = a(t3X2mYm+t3XmY2m−3t2XmYm+ 1)
−bt2XmYm−t4X2mY2m+ 3t2XmYm
−tXm−tYm = 0 Problem: Segre’s criterion does not apply
(Sz˝onyi, 1988)
Let P = (b,a) be a point in AG(2,q)\ X. If m<p4
q/36 then there is a secant of S passing through P
(Sz˝onyi, 1988)
Let P = (b,a) be a point in AG(2,q)\ X. If m<p4
q/36 then there is a secant of S passing through P
m is a divisor ofq−1
(Sz˝onyi, 1988)
Let P = (b,a) be a point in AG(2,q)\ X. If m<p4
q/36 then there is a secant of S passing through P
m is a divisor ofq−1
the points inX \S need to be dealt with
(Sz˝onyi, 1988)
Let P = (b,a) be a point in AG(2,q)\ X. If m<p4
q/36 then there is a secant of S passing through P
m is a divisor ofq−1
the points inX \S need to be dealt with Theorem
If m is a divisor of q−1with m<p4
q/36, and in addition (m,q−m1) = 1, then there exists a complete arc in AG(2,q) with size
m+ q−1 m −3
(Sz˝onyi, 1988)
Let P = (b,a) be a point in AG(2,q)\ X. If m<p4
q/36 then there is a secant of S passing through P
m is a divisor ofq−1
the points inX \S need to be dealt with Theorem
If m is a divisor of q−1with m<p4
q/36, and in addition (m,q−m1) = 1, then there exists a complete arc in AG(2,q) with size
m+ q−1
m −3 ∼q3/4
Isolated double point case: Y ( X
2− β ) = 1
G cyclic of orderq+ 1, isomorphic to the subgroup of order q+ 1 of (Fq2,·)
Isolated double point case: Y ( X
2− β ) = 1
G cyclic of orderq+ 1, isomorphic to the subgroup of order q+ 1 of (Fq2,·)
For v ∈F∗q2 let
Qv =
v+1 v−1
√β,(v4v−1)β2
ifv 6= 1
X∞ ifv = 1
Isolated double point case: Y ( X
2− β ) = 1
G cyclic of orderq+ 1, isomorphic to the subgroup of order q+ 1 of (Fq2,·)
For v ∈F∗q2 let
Qv =
v+1 v−1
√β,(v4v−1)β2
ifv 6= 1
X∞ ifv = 1
the group G G =n
Qt|t = u+√ β u−√
β, u ∈Fq
o
∪ {X∞}
Isolated double point case: Y ( X
2− β ) = 1
G cyclic of orderq+ 1, isomorphic to the subgroup of order q+ 1 of (Fq2,·)
For v ∈F∗q2 let
Qv =
v+1 v−1
√β,(v4v−1)β2
ifv 6= 1
X∞ ifv = 1
the group G G =n
Qt|t = u+√ β u−√
β, u ∈Fq
o
∪ {X∞} the subgroup of index mof G:
K =n
Qtm|t= u+√ β u−√
β, u∈Fq
o∪ {X∞}.
Isolated double point case: Y ( X
2− β ) = 1
G cyclic of orderq+ 1, isomorphic to the subgroup of order q+ 1 of (Fq2,·)
For v ∈F∗q2 let
Qv =
v+1 v−1
√β,(v4v−1)β2
ifv 6= 1
X∞ ifv = 1
the group G G =n
Qt|t = u+√ β u−√
β, u ∈Fq
o
∪ {X∞} the subgroup of index mof G:
K =n
Qtm|t= u+√ β u−√
β, u∈Fq
o∪ {X∞}. a coset:
S =n
Q¯ttm |t = u+√ β u−√
β, u ∈Fq
o
∪ {Q¯t}.
CP : det
a b 1
¯txm+1
¯txm−1
√β (¯tx4¯txm−m1)β2 1
¯tym+1
¯tym−1
√β (¯ty4¯tym−m1)β2 1
= 0
CP : det
a b 1
¯txm+1
¯txm−1
√β (¯tx4¯txm−m1)β2 1
¯tym+1
¯tym−1
√β (¯ty4¯tym−m1)β2 1
= 0
Problems:
- the curveCP is not defined over Fq
- we are not looking for anFq-rational point, asx andy must be of type uu+√β
−
√β
CP : det
a b 1
¯txm+1
¯txm−1
√β (¯tx4¯txm−m1)β2 1
¯tym+1
¯tym−1
√β (¯ty4¯tym−m1)β2 1
= 0
Problems:
- the curveCP is not defined over Fq
- we are not looking for anFq-rational point, asx andy must be of type uu+√β
−
√β
Solution: look for another curve, birationally equivalent to CP, which is defined overFq
(Anbar-Bartoli-G.-Platoni, 2013)
Let P = (a,b) be a point in AG(2,q) offX. If m<p4
q/36
then, apart from s ≤3 exceptions,P is collinear with two distinct points of S.
(Anbar-Bartoli-G.-Platoni, 2013)
Let P = (a,b) be a point in AG(2,q) offX. If m<p4
q/36
then, apart from s ≤3 exceptions,P is collinear with two distinct points of S.
m is a divisor ofq+ 1
(Anbar-Bartoli-G.-Platoni, 2013)
Let P = (a,b) be a point in AG(2,q) offX. If m<p4
q/36
then, apart from s ≤3 exceptions,P is collinear with two distinct points of S.
m is a divisor ofq+ 1
the points inX \S need to be dealt with