2.2. Auxiliary Lemmas
In the rest of
Section 2,
will always be a number in
. Let
.
(
, resp.) and
(
, resp.) are said to be
adjacent (denoted by
or
) if they are contained in adjacent orthants. The following lemma is of fundamental importance for this section.
Lemma 4. Suppose that are distinct, , and . If ; then, . In particular, a translate of intersects at most two members of , which are adjacent.
Proof. Fix . Clearly, .
Let
. By (
4) and (
5), we have
Since , there are two possible cases.
Case 1.
. Then, for each
, we have
Case 2. There exists a unique
such that
. Then, for each
, we have
In both cases, by Lemma 1, we have .
To complete the proof, it is sufficient to note that among any three distinct orthants, there is a pair that are not adjacent. □
Lemma 5. Let .
- (a)
If there exists a such that , then .
- (b)
If , then, for each , contains at most two points in .
- (c)
For each , cannot be covered by a single translate of .
Proof. We only need to show (a) and (b).
(a). Take
. By Lemma 2, we have
Since a point is in
if and only if its coordinates is a permutation of the coordinates of
x, we have
(b). Assume the contrary that there exists
such that
. Let
. By (
4), we have
. By (
16),
, which is a contradiction. □
Let
and
be adjacent, with the signs of their
-th coordinates being opposite. Set
Clearly, and . Let and . We write if . Otherwise, we write and we have . Thus, if , then they cannot be covered by a single translate of .
Clearly, we have
Lemma 6. Suppose that and B is a translate of .
- (a)
and .
- (b)
If , then there exist and such that , and and .
Lemma 7. Let . Suppose that . If , then
- (a)
,
- (b)
.
Proof. (a). By applying a transformation in
if necessary, we may assume that
and
. Suppose that the
-th coordinates of
and
are with the opposite sign. Let
. Then, one of the coordinates of
x is 1 and
. By Lemma 2, one coordinate of
(see (
9) for the definition) is
and the other two are 1. It follows that
. If
, then, by (
4), we have
, which contradicts (
10). Hence,
.
(b). It follows from Lemma 4 and Lemma 7a. □
Suppose that . Lemma 7 shows that a translate of containing cannot contain any point in . Conversely, a translate of containing a point in cannot contain . Particularly, in the case of Lemma 6b, .
Lemma 8. Let B be a translate of .
- (a)
If , then .
- (b)
If , then .
Proof. By symmetry, we only need to prove (a). By Lemma 4, . Lemma 7 and the fact show that . The distance from to is 14. Thus, . □
Lemma 9. Let .
- (a)
If , then .
- (b)
If , then .
- (c)
If , then .
Proof. (a). By the hypothesis and (
4), we have
Adding these inequalities, we have
By
and (
4),
.
(b). By the hypothesis and (
4), we have
By
and (
4), we have
.
(c). By the hypothesis and (
4), we have
By
and (
4), we have
. □
Lemma 10. Let . If , then Proof. By symmetry, we only need to consider the case when
. By (
5), we have
Let
. By (
4), we have
By (
6), (
7), and Lemma 1, we have
. □
In the rest of this section,
is a set satisfying
,
Let
be distinct. Then,
, which implies that
. Set
. By Lemma 10,
Set .
Recall that A 2-regular graph is a graph where each vertex is connected to exactly two other vertices.
Lemma 11. Let and be two sets satisfying , be the bipartite graph with bipartitions and , where and are connected by an edge if and only if . If
- (a)
,
- (b)
no member of can be covered by a single translate of ,
- (c)
and is 2-regular,
then, for each pair of distinct members and of , no member of can contain two points in and two points in simultaneously.
Proof. By the hypothesis, each member of is covered by exactly two members of , and each member of intersects exactly two members of .
Assume the contrary that there exist
and distinct
,
such that
contains two points in
and two points in
. By Lemma 4,
. Without loss of generality, we may assume that
. By Lemma 6, we may also assume (exchange the first two coordinates if necessary) that
Let
and
be members of
such that
and
. See
Figure 1. By Lemma 7,
Note that
. By (
10), we have
.
Then,
and
. It can be verified that (see also [
15])
From Lemma 8, it follows that
By Lemma 7,
contains at most eight points of
. Clearly,
. Since
, there exists
that contains two points in
and two points in
, where
. Let
and
be two members of
such that
and
. As above, we have
. By (
18), we have
By Lemma 7 again,
can cover at most two points in
Therefore, at least points in remain uncovered, while the number of the remaining translates of is . Thus, .
If , then there exists a member of containing five points in E, which contradicts Lemma 6. If , then the remaining two members and of contain eight points of . By Lemma 7, contains no point in C, which is a contradiction. □
Lemma 12. If , then there exist and such that .
Proof. Let be the bipartite graph defined as in Lemma 11 with bipartitions and . By Lemma 4, the total number of edges is at most 16.
Assume the contrary that, for each and , . Then, for each , the degree of in is at least two. Thus, is 2-regular.
Clearly,
. By Lemmas 4 and 6,
. By Lemmas 6 and 11, we only need to consider the case when
. For each
, we have
Then, for each , there exists a unique such that contains one point in and two points in , where .
Let
and
be two members of
such that
. Without loss of generality, we may assume that
We claim that
. Otherwise, there exists
in
such that
. By Lemma 4,
. From (
19), it follows that
, which contradicts Lemma 4 since
, and
are pairwise nonadjacent.
There are two cases.
Case a.
. Then,
. It follows that
. By Lemma 9a,
. Thus,
By Lemma 9b, . Since , . Clearly, , which is a contradiction.
Case b. . Without loss of generality, we may assume (exchange the first two coordinates if necessary) that . By Lemma 6, . Then, either or is contained in .
If
, then, by
and Lemma 7,
. By Lemma 7b and (
19), there exists a member of
containing
that contains at most two points in
E, which is a contradiction.
Now, assume that
. We claim that
. Otherwise, (20) holds, which yields a contradiction as in Case a. From Lemma 9c, it follows that
, which contradicts Lemma 7b and (
19). □
Corollary 1. If , then there exist and distinct such that Proof. Assume the contrary that, for each
and each
satisfying
, we have
Thus, .
Considering the bipartite graph
G defined as in Lemma 11 with bipartitions
and
. Each member of
has degree 1 in
G and, by Lemma 4, each member of
has degree at most 2. Thus, the total number
L of edges of
G satisfies
Each member of has degree at least 1 and each member of has degree at least 2. Therefore, we have . It follows that , and the subgraph of G with bipartitions and is 2-regular.
By Lemma 5, we have . By Lemma 12, there exist and such that . From Lemma 5b, it follows that . Then, there exists an such that . By Lemma 4, can only intersect with the three members of that are adjacent to . Note that these three members of are pairwise nonadjacent. From Lemma 5b and Lemma 4, it follows that contains at most two points in and .
Clearly, . By the pigeon hole principle and Lemma 6, there exist and such that contains two points in and two points in , which contradicts Lemma 11. □
2.3. The Proof of Theorem 1
Suppose the contrary that there exists a set
with
such that
. By (
17), we have
, where
.
By Corollary 1, without loss of generality, we may assume that there exists
such that
Lemma 13. .
Proof. Take
, then
, which is equivalent to
By
, we have
It follows that . Hence, .
It can be verified that (see also [
15])
Set
and
. It can be verified that
, and
A is 14-separated. For each
, there exists
such that
. Clearly, if
are distinct, then
and
are distinct. By
, we have
. By Lemma 4,
. Hence,
and
. Set
Put
where
,
, and
. Since
is a linear isometry on
,
is also 14-separated.
Clearly, if are distinct and , then by Lemma 4, .
Since
is 14-separated,
. From
, it follows that
. Thus,
On the one hand, neither
nor
is adjacent to any of
and
(see
Table 1); thus,
. On the other hand, since
is 14-separated,
. Thus,
. Hence, either
or
By Lemma 4 and
Table 1,
. Since
and
we have
Lemma 14. .
Proof. By Lemma 4 and
Table 1, we obtain all possible locations of points listed in
Table 4.
Suppose the contrary that one of the two points
and
is in
. By
. By
and
,
. Thus,
. Since
intersects both
and
,
and
.
Note that . By , . Since intersects both and , . Thus, .
By
Table 4 and
, we have
.
By , , , and , we have .
Clearly, . By Lemma 13, . Moreover, , , and neither nor is adjacent to . By Lemma 4, . Thus, .
Since , , and , we have . By and , . By , intersects both and , we have , a contradiction. □
Corollary 2. .
Proof. Since
, we have
. From
and
, it follows that
Thus,
is a configuration satisfying (
22), where
is replaced with
. Let
be the point in
satisfying
. Then, since
we have
. Applying Lemma 14 on
, we have
. It follows that
. □
By
and
Table 4, we have
.
Lemma 15. .
Proof. By
Table 1, we only need to show that
.
By and , we have .
Next, we claim that
. Otherwise, since
, we would have
. By
, (
25), and Corollary 2, we have
.
By
Table 1,
is contained in
. By
we have
. Hence,
intersects both
and
. By Lemma 4,
, and therefore
. Since
, we have
. In the current situation,
intersects both
and
, and
intersects both
and
. By Lemma 4 and
Table 1,
.
By
Table 1,
, and
we have
.
By
Table 1 and Lemma 13,
for some
. Since
intersects
,
intersects
,
, and
, we have, by Lemma 4,
.
Clearly,
. Since
intersects both
and
,
. By
. Hence,
intersects
,
, and
, which contradicts Lemma 4.
In the rest of this section, we show that
. Otherwise, by
and
, we have
. Since
intersects both
and
, we have
. By
we have
. By
and Lemma 4,
. Thus,
. By
Table 4 and
we have
. Since
intersects both
and
,
. By
Table 4,
we have
. Since
intersects
and
, and
intersects
and
,
. By Lemma 13,
. Thus,
.
Clearly,
. Since
intersects both
and
,
. By
and Lemma 4, we have
, contradicting the fact that
intersects both
and
. □
Lemma 16. If , then .
If , then .
Proof. We prove the first statement and the other one can be obtained by arguments as in the proof of Corollary 2.
By Lemma 14, either or .
Case a . By (
4), for each
, we have
By (
4) and
, for each
, we have
By (
4) again, we have
.
Case b . By (
4), we have
From
and (
4), it follows that
By (
4), we have
. □
If
, then by
we have
. Thus,
If
, then one can verify that
Proof. We show that . The inclusion can be proved by similar arguments as in the proof of Corollary 2.
By Lemma 16, we only need to show that . By Lemmas 4 and 13, . Thus, it suffices to show that, if is a subset of , then .
First, we show that
. Otherwise, by
we have
. By Corollary 2, we have
. By Lemma 4 and the facts
,
, and
, we have
which, together with
and (
27), shows that
.
Similarly, by
,
, and (
26), we have
. It follows that
intersects both
and
, which are not adjacent, which is a contradiction.
Next, we show that
. Suppose that this is not true. From
and Lemma 4, it follows that
. Thus,
. By
we have
. By (
26), we have
, which is impossible since
.
In the rest of this section, we show that
. Otherwise, by
we have
. By
Table 1,
. Since
intersects both
and
and
, we have
. By
we have
. By (26),
we have
. Then,
intersects both
and
, which are not adjacent, which is a contradiction. □
Proof. By
Table 1 and Lemma 13, these two points are in
. By Lemma 17 and
, we have
. □
Lemma 18. We have .
Proof. By
Table 1 and Lemma 13,
. We show that
, and
can be proved by similar arguments as in the proof of Corollary 2.
Otherwise,
. By Lemma 4,
may only intersect
and
. Since
we have
. Since
intersects both
and
, and
we have
. Thus,
intersects both
and
. By Lemma 4,
.
Since
intersects
and
,
. By
we have
and
.
Since
, we have
. Thus,
. By Lemma 13,
. By (
28) and
, we have
. By
we have
.
Since
intersects both
and
,
. By
and
, we have
. By (
28) and
this is impossible. □
Lemma 19. If , then .
Proof. From (
4) and
,
, it follows that
Taking the sum of (
30) and (32), the sum of (31) and (33), and the sum of (
30) and (31), and applying the triangle inequality, we obtain
By (
4), we have
. □
Lemma 20. Either or .
Proof. Assume the contrary that
and
. By (
4), we have
By (
16),
. Clearly,
. Thus, by
Table 1, we have
By Lemma 18,
can only intersect
and
. Moreover,
can only intersect
and
, and
can only intersect
and
. It follows that
By
the midpoint
of
and
is in
. If
, then, by (
4), we have
By (
16), we have
. By (
36),
. In this situation, by Lemma 4,
cannot intersect
and
. Thus, by (
36),
If
then, by Lemma 4,
cannot intersect
and
. By (
36) again, we have (
37). It follows that
.
By
and
,
. If
, then, since
,
does not intersect
. By (
34),
. Since
, by (
35),
. Since
,
, and
are not adjacent to
, by Lemma 4,
, which contradicts Lemma 5b. Thus,
. Similarly, we have
.
By Lemma 19,
. By
, (
37), and Lemma 5b, this is impossible. □
Clearly, by (
29) and Lemma 20,
.
Lemma 21. If , then . If , then .
Proof. Suppose that
. By
,
and Lemma 4,
.
By , , and , we have .
Now, suppose that . As in the proof of Corollary 2, we have . By the first case, we have , where is defined as in Corollary 2. Then, either or . If , then and , which contradicts Lemma 4. Hence, . □
If
, then
. Hence, by Lemma 4, (
23) holds. However, this is in contradiction to
. If
, then
, which is in contradiction to
. This completes the proof of Theorem 1.