Abstract:The use of quadratic residues to construct matrices with specific determinant values is a familiar problem with connections to many areas of mathematics and statistics. Our research has focused on using cubic residues to construct matrices with interesting and predictable determinants.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to construct matrices with specific and predictable determinant values by using cubic residues. Specifically, the authors hope to imitate the work of Paley and Chapman in the field of quadratic residues, explore the behavior of cubic residues when constructing matrices, and look for patterns in the determinants of these matrices.
### Problem Background
1. **Paley's Work**:
- Paley constructed two types of Hadamard matrices using the Legendre symbol, which are respectively applicable to primes of the form \(4k + 1\) and \(4k+3\).
- For a prime \(p\) of the form \(4k + 3\), the element in the \(i\)-th row and \(j\)-th column of the matrix is defined as \(\left(\frac{i - j}{p}\right)\), that is, the quadratic residue symbol.
2. **Chapman's Work**:
- Chapman studied matrices generated by different types of Legendre symbols and observed that for a small prime \(p\) of the form \(4k + 3\), the determinant of the matrix \(M\) is 1.
- This conjecture was later proved by Vsemirnov.
### Research Objectives
The authors hope to generalize the methods of Paley and Chapman to cubic residue symbols in order to construct matrices with interesting and predictable determinants. Specifically:
- Use the cubic residue symbol \(\left[\frac{j - i + c}{p}\right]\) to construct the matrix \(M_n\).
- Calculate the determinants of these matrices and organize the data to look for patterns.
### Main Results
The authors mainly focus on primes \(p\) of the form \(3k + 2\) and have obtained the following conclusions:
1. **Theorem 3.1**: For any prime \(p\) of the form \(3k + 2\), the determinant of the matrix \(M_n\) is \((-1)^{n - 1}(n - 1)\), where \(1\leq n\leq p\) and \(c = 0\).
2. **Theorem 3.3**: For any prime \(p\) of the form \(3k + 2\), the determinant of the matrix \(M_n\) is 1, where \(1 < n\leq p - 1\) and \(c=\pm1\).
3. **Theorem 3.5**: For any prime \(p\) of the form \(3k + 2\), the determinant of the matrix \(M_n\) is \(p - 1\), where \(n = p\) and \(1\leq c\leq p - 1\).
4. **Theorem 3.7**: For any prime \(p\) of the form \(3k + 2\), the determinant of the matrix \(M_n\) is 0, where \(2\leq n\leq p - 2\) and \(2\leq c\leq p - 2\).
5. **Theorem 3.9**: For any prime \(p\) of the form \(3k + 2\), the determinant of the matrix \(M_n\) is 1, where \(n = p - 1\) and \(1\leq c\leq p - 1\).
### Future Work
The authors plan to generalize these results to the Eisenstein integer ring \(Z[\omega]\), where \(\omega\) is a primitive cube root of unity. This will involve more complex cubic residue symbols.
### Summary
The main contribution of this paper lies in exploring the application of cubic residues in matrix construction and discovering several rules for the determinants of these matrices. This provides a theoretical basis for further research on higher - order residues.