Abstract
We determine the Weierstrass semigroup \(H(P_\infty ,P_1,\ldots ,P_m)\) at several rational points on the maximal curves which cannot be covered by the Hermitian curve introduced in Tafazolian et al. (J Pure Appl Algebra 220(3):1122–1132, 2016). Furthermore, we present some conditions to find pure gaps. We use this semigroup to obtain AG codes with better relative parameters than comparable one-point AG codes arising from these curves.
Similar content being viewed by others
References
Bartoli D., Masuda A.M., Montanucci M., Quoos L.: Pure gaps on curves with many rational places. Finite Fields Appl. 53, 287–308 (2018). https://doi.org/10.1016/j.ffa.2018.07.001.
Bartoli D., Montanucci M., Zini G.: AG codes and AG quantum codes from GGS curves. Des. Codes Cryptogr. 86(10), 2315–2344 (2018).
Bartoli D., Montanucci M., Zini G.: Multi-point AG codes on the GK maximal curve. Des. Codes Cryptogr. 86(1), 161–177 (2018).
Beelen P., Montanucci M.: A new family of maximal curves. J. Lond. Math. Soc. 98, 573–592 (2018).
Bras-Amorós M.: On numerical semigroups and the redundancy of improved codes correcting generic errors. Des. Codes Cryptogr. 53(2), 111–118 (2009). https://doi.org/10.1007/s10623-009-9297-8.
Bras-Amorós M., O’Sullivan M.E.: The correction capability of the Berlekamp-Massey-Sakata algorithm with majority voting. Appl. Algebra Energy Commun. Comput. 17(5), 315–335 (2006). https://doi.org/10.1007/s00200-006-0015-8.
Bras-Amorós M., Lee K., Vico-Oton A.: New lower bounds on the generalized Hamming weights of AG codes. IEEE Trans. Inform. Theory 60(10), 5930–5937 (2014). https://doi.org/10.1109/TIT.2014.2343993.
Campillo A., Farrán J.I.: Computing Weierstrass semigroups and the Feng-Rao distance from singular plane models. Finite Fields Appl. 6(1), 71–92 (2000). https://doi.org/10.1006/ffta.1999.0266.
Campillo A., Farrán J.I., Munuera C.: On the parameters of algebraic-geometry codes related to Arf semigroups. IEEE Trans. Inform. Theory 46(7), 2634–2638 (2000). https://doi.org/10.1109/18.887872.
Carvalho C., Torres F.: On Goppa codes and Weierstrass gaps at several points. Des. Codes Cryptogr. 35(2), 211–225 (2005).
Castellanos A.S., Tizziotti G.: Two-points AG codes on the GK maximal curves. IEEE Trans. Inform. Theory 62(9), 4867–4872 (2016). https://doi.org/10.1109/TIT.2015.2511787.
Castellanos A.S., Tizziotti G.: Weierstrass semigroup and pure gaps at several points on the GK curve. Bull. Braz. Math. Soc. (N.S.) 49(2), 419–429 (2018). https://doi.org/10.1007/s00574-017-0059-3.
Duursma I.M., Park S.: Delta sets for divisors supported in two points. Finite Fields Appl. 18(5), 865–885 (2012). https://doi.org/10.1016/j.ffa.2012.06.005.
Fanali S., Giulietti M.: One-point AG codes on the GK maximal curves. IEEE Trans. Inform. Theory 56(1), 202–210 (2010).
Farrán J.I., Munuera C.: Goppa-like bounds for the generalized Feng-Rao distances. Discret. Appl. Math. 128(1), 145–156 (2003). https://doi.org/10.1016/S0166-218X(02)00441-9. International Workshop on Coding and Cryptography (WCC 2001) (Paris)
Feng G.L., Rao T.R.N.: A simple approach for construction of algebraic-geometric codes from affine plane curves. IEEE Trans. Inform. Theory 40(4), 1003–1012 (1994). https://doi.org/10.1109/18.335972.
Fuhrmann R., Garcia A., Torres F.: On maximal curves. J. Number Theory 67(1), 29–51 (1997).
Fulton W.: Algebraic curves. An introduction to algebraic geometry. W. A. Benjamin Inc., New York-Amsterdam (1969). Notes written with the collaboration of Richard Weiss, Mathematics Lecture Notes Series
Garcia A., Stichtenoth H.: A maximal curve which is not a galois subcover of the hermitian curve. Bull. Braz. Math. Soc. (N.S.) 37(1), 139–152 (2006).
Garcia A., Güneri C., Stichtenoth H.: A generalization of the giulietti-korchmáros maximal curve. Adv. Geom. 10(3), 427–434 (2010).
Geil O., Matsumoto R.: Bounding the number of \(\mathbb{F}_q\)-rational places in algebraic function fields using Weierstrass semigroups. J. Pure Appl. Algebra 213(6), 1152–1156 (2009). https://doi.org/10.1016/j.jpaa.2008.11.013.
Giulietti M., Korchmáros G.: A new family of maximal curves over a finite field. Math. Ann. 343, 229–245 (2009).
Goppa V.D.: Codes on algebraic curves. Dokl. Akad. NAUK SSSR 259, 1289–1290 (1981).
Goppa V.D.: Algebraic-geometric codes. Izv. Akad. NAUK SSSR 46, 75–91 (1982).
Heijnen P., Pellikaan R.: Generalized Hamming weights of \(q\)-ary Reed-Muller codes. IEEE Trans. Inform. Theory 44(1), 181–196 (1998). https://doi.org/10.1109/18.651015.
Høholdt T., van Lint J.H., Pellikaan R.: Algebraic geometry codes. In: Handbook of coding theory, Vol. I, II, pp. 871–961. North-Holland, Amsterdam (1998).
Homma M., Kim S.J.: Goppa codes with Weierstrass pairs. J. Pure Appl. Algebra 162(2–3), 273–290 (2001).
Hu C., Yang S.: Multi-point codes from GGS curves. Adv. Math. Commun. 14, 279 (2020). https://doi.org/10.3934/amc2020020.
Kirfel C., Pellikaan R.: The minimum distance of codes in an array coming from telescopic semigroups. IEEE Trans. Inform. Theory 41(6 part I), 1720–1732 (1995). https://doi.org/10.1109/18.476245.
Korchmáros G., Torres F.: On the genus of a maximal curve. Math. Ann. 323(3), 589–608 (2002).
Lachaud G.: Sommes d’Eisenstein et nombre de points de certaines courbes algébriques sur les corps finis. CR. Acad. Sci 305, 729–732 (1987).
Matthews G.L.: The Weierstrass semigroup of an \(m\)-tuple of collinear points on a Hermitian curve. In: Finite fields and applications, Lecture Notes in Comput. Sci., Vol. 2948, pp. 12–24. Springer, Berlin (2004)
Pellikaan R., Torres F.: On Weierstrass semigroups and the redundancy of improved geometric Goppa codes. IEEE Trans. Inform. Theory 45(7), 2512–2519 (1999). https://doi.org/10.1109/18.796393.
Pellikaan R., Stichtenoth H., Torres F.: Weierstrass semigroups in an asymptotically good tower of function fields. Finite Fields Appl. 4(4), 381–392 (1998).
Schmid W.C., Schürer R.: MinT. http://mint.sbg.ac.at/
Tafazolian S., Teherán-Herrera A., Torres F.: Further examples of maximal curves which cannot be covered by the Hermitian curve. J. Pure Appl. Algebra 220(3), 1122–1132 (2016). https://doi.org/10.1016/j.jpaa.2015.08.010.
Acknowledgements
The first author was supported by Fundação de amparo a pesquisa de Minas Gerais under Grant FAPEMIG APQ-00696-18. The second author was supported by the Spanish government under Grant TIN2016-80250-R and by the Catalan government under Grant 2017 SGR 00705.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This is one of several papers published in Designs, Codes and Cryptography comprising the “Special Issue on Codes, Cryptology and Curves”.
Appendices
Appendix A: genus of S
In this section we will determine the genus of the semigroup S generated by \(\{q^3+1-iN-j:0\le i+j\le p^b\},\) with \(q=(p^b)^r\), for a prime number p, some positives integers b, r, with \(r\ge 2\), and \(N=\frac{q^3+1}{q+1}\).
For this purpose we will first consider the numerical semigroup \(S_{P,N,K}\) generated by \(\{KN+aN-j:0\le j\le a\le P\}\), where P, N, K are positive integers with \(P\mid N-1\), \(P\mid K-1\), \(K<N\). Notice that \(S=S_{P,N,K}\) with \(K=q+1-p^b\) and \(P=p^b\), with a playing the role of \(p^b-i\), and j, N playing their own role. The required conditions hold, indeed, \(P\mid N-1\) since \(N-1=\frac{q^3+1}{q+1}-1=(q^2-q+1)-1=q^2-q\) and, similarly, \(P\mid K-1\).
1.1 A.1 Characterization
With the same notation as before, for an integer \(M\ge 0\), let \(S^M_{P,N,K}=\{MKN+aN-j:0\le j\le a\le MP\}\) and let \(\widetilde{S}^M_{P,N,K}=\{MKN+aN-j: \max \{0,(M-1)P-K+1\}\le a \le MP, 0\le j\le \min \{a,N-1\}\}\). It is obvious that \(S_{P,N,K}=\cup _{M\ge 0} S^M_{P,N,K}\) and that \(\widetilde{S}^M_{P,N,K}\subseteq S^M_{P,N,K}\). Now, any element in \(S^M_{P,N,K}\) is in at least one set \(\widetilde{S}^{M'}_{P,N,K}\) for some \(M'\le M\). This can be proved by induction on M. For \(M=0\) and for \(M=1\) it is straightforward. For \(M>1\), suppose that an element of \(S^M_{P,N,K}\) is \(\ell =MKN+aN-j\) for some particular \(0\le j\le a\le MP\). If \(a\ge (M-1)P-K+1\) and \(j\le N-1\) then \(\ell \in \widetilde{S}^M_{P,N,K}\). Otherwise, if \(0\le a<(M-1)P-K+1\), then \(\ell =(M-1)KN+(K+a)N-j=(M-1)KN+a'-j\) with \(a'=K+a\le (M-1)P\) and \(0\le j\le a\le a'\), so \(\ell \in S^{M-1}_{P,N,K}\) and the result follows by induction. If \(a\ge (M-1)P-K+1\) but \(j>N-1\), then suppose that Q is the quotient of the division of j by N. Then \(\ell =MKN+(a-Q)N-(j-QN)=MKN+a'-j'\) with \(a'=a-Q \ge a-j \ge 0\), and \(a'\le a\le MP\). Furthermore, \(j'=j-QN\), which is the remainder of the division of j by N, and which is between 0 and \(N-1\). So, \(\ell \in \widetilde{S}^M_{P,N,K}\). Consequently, we also have \(S_{P,N,K}=\cup _{M\ge 0} \widetilde{S}^M_{P,N,K}\).
Now, \(\widetilde{S}^M_{P,N,K}\subseteq [(M-1)(K+P)N+1,M(K+P)N]\). Indeed, the minimum of \(\widetilde{S}^M_{P,N,K}\) is at least \((M-1)(KN+PN)+1\) since the elements in \(\widetilde{S}^M_{P,N,K}\) satisfy \(MKN+aN-j\ge MKN+((M-1)P-K+1)N-N+1=MKN+MPN-PN-KN+N-N+1=(M-1)(K+P)N+1\). On the other hand, the maximum of \(\widetilde{S}^M_{P,N,K}\) is at most \(M(K+P)N\) since the elements in \(\widetilde{S}^M_{P,N,K}\) satisfy \(MKN+aN-j\le MKN+MPN-0=M(K+P)N\).
In particular, the sets \(\widetilde{S}^M_{P,N,K}\) are disjoint and, so, \(S_{P,N,K}=\sqcup _{M\ge 0}\widetilde{S}^M_{P,N,K}.\)
Let \(M_0=\frac{K-1}{P}+1\), \(M_1=\frac{N-1}{P}\), \(M_2=\frac{K+N-2}{P}\).
1.2 A.2 Conductor of \(S_{P,N,K}\)
Now we are ready to determine the Frobenius number of \(S_{P,N,K}\), that is, its largest gap. The conductor of \(S_{P,N,K}\) is then the non-gap of \(S_{P,N,K}\) right after its Frobenius number.
Observe that \(\widetilde{S}_M=\cup _{a\ge \max \{0,(M-1)P-K+1\}}I_a\) with \(I_a=\{MKN+aN-j: 0\le j\le \min \{a,N-1\}\}\). Since j ranges from 0 to \(\min \{a,N-1\}\), there are gaps between the intervals \(I_a\) and \(I_{a-1}\) if and only if \(\min \{a,N-1\}<N-1\), i.e., if and only if \(a< N-1\). Since in \(\widetilde{S}^M_{P,N,K}\)a ranges from \(\max \{0,(M-1)P-K+1\}\) to PN, the inequality \(a<N-1\) occurs in \(\widetilde{S}^M_{P,N,K}\) if and only if \((M-1)P-K+1<N-1\), that is, if and only if \(M\le \frac{N+K-2}{P}\).
Let \(M_F= \frac{N+K-2}{P}\). The last gap of \(S_{P,N,K}\) will then be the gap previous to \(I_a\) with \(a=N-2\) in \(\widetilde{S}^{M_F}_{P,N,K}\), that is, the Frobenius number will be \(M_FKN+aN-a-1\) for \(a=N-2\), i.e. \(\frac{K+N-2}{P}KN+(N-2)N-N+1=\frac{K+N-2}{P}KN+N^2-3N+1\).
Simplifying by means of Sage we obtain that the conductor of \(S_{P,N,K}\) is
1.3 A.3 Genus
Let \(M_0=\frac{K-1}{P}+1\), \(M_1=\frac{N-1}{P}\), \(M_2=\frac{K+N-2}{P}=M_0+M_1-1\).
-
If \(0\le M\le M_0\) then \(S_{P,N,K}^M=\widetilde{S}_{P,N,K}^M\) and \(\#\widetilde{S}_{P,N,K}^M=\sum _{a=0}^{MP}(a+1)=\sum _{b=1}^{MP+1} b=\frac{(MP+1)(MP+2)}{2}\).
-
If \(M_0<M\le M_1\) then \(\#\widetilde{S}_{P,N,K}^M=\sum _{a=(M-1)P-K+1}^{MP}(a+1)=\sum _{(M-1)P-K+2}^{MP+1}b=\frac{(MP+1)(MP+2)}{2}-\frac{(MP-P-K+1)(MP-P-K+2)}{2}\). This is because if \(M\le M_1\), then \(MP\le N-1\), so j will always range from 0 to a.
-
If \(M>M_1\) then \(\#([(M-1)(K+P)N+1,M(K+P)N]{\setminus } \widetilde{S}_{P,N,K}^M)=\)\(\sum _{a=(M-1)P-K+1}^{N-2}(N-a-1)=\sum _{b=1}^{N-(M-1)P+K-2}b=\)\(\frac{(N-MP+P+K-2)(N-MP+P+K-1)}{2}\). This is because if \(M>M_1\) then \(MP>N-1\) and so at some point \(a=N-1\). In this case, \([(M-1)(K+P)N+1,M(K+P)N] {\setminus }\widetilde{S}_{P,N,K}^M=\{MKN+aN-j: (M-1)P-K+1 \le a \le N-2, a+1\le j\le N-1\}\).
Now, using the formulas \(\sum _{M=1}^n M=\frac{n(n+1)}{2}\) and \(\sum _{M=1}^n M^2=\frac{n(n+1)(2n+1)}{6}\) we get to the final count of the genus:
1.4 A.4 Back to the originary problem
If we take \(N=\frac{q^3+1}{q+1}\), \(K=q+1-p^b\), \(P=p^b\), then the genus is \(g=\frac{q^5 - q^3p^b - q^3 + q^2}{2p^b},\) while the conductor is \(c=\frac{q^5 - 2q^3p^b + q^2p^{2b} - q^2p^b - qp^{2b} + q^2 + qp^b + p^{2b} - p^b}{p^b}.\)
Appendix B: genus of \(S'\)
Suppose we have \(q=p^a\), \(b\mid a\), \(b\ne a\), n odd, \(n\ge 3\). Let \(M=\frac{q^n+1}{q+1}=q^{n-1}-q^{n-2}+q^{n-3}-q^{n-4}+\dots -q+1\). We want to prove that the genus of the semigroup \(S'=\langle q^n+1-iM-j:0\le iM+jq^2\le q^{n-1}p^b \rangle \) is \(\frac{q^{n+2}-p^bq^n-q^3+q^2}{2p^b}\).
1.1 B.1 Definition of \(S'\) revisited
Lemma B.1
If \(n\ge 3\),
Equivalently, by setting \(A=q^{n-1}-q^{n-2}\), \(k_0=q+1-p^b\), \(k_1=q+1\), \(\ell _0=q^{n-3}(q-p^b)+1\), \(\ell _1=\frac{q^{n-2}+1}{q+1}\), the semigroup \(S'\) is \(S'=\left\langle kA+\ell : k_0\le k\le k_1, \ell _0 \le \ell \le k\ell _1\right\rangle .\)
Proof
We can rewrite \(S'\) as \(S'=\langle (q+1-i)M-j:0\le iM+jq^2\le q^{n-1}p^b \rangle .\) The integer i is then bounded as \(0\le i\le \lfloor \frac{q^{n-1}p^b}{M}\rfloor \). The quotient and the remainder of the division of \(q^{n-1}p^b\) by M are, respectively, \(p^b\) and \(p^b(q^{n-2}-q^{n-3}+\dots +q-1)\) (since this remainder is between 0 and \(M-1\)). Consequently, \(0\le i\le p^b\). Now, setting \(k=q+1-i\), the bounds for k are \(q-p^b+1\le k\le q+1\). Hence, since \(iM=(q+1-k)M=(q+1)M-kM=q^n+1-kM\),
Finally, we want to replace the bounds for \(q^n+1-kM+jq^2\) by bounds on j. Reorganizing them, we obtain
Since \(k\le q+1\), the lower bound is non-positive. So, \(0\le jq^2\)
As for the upper bound on j, \(j\le \left\lfloor \frac{kM-q^{n-1}(q-p^b)-1}{q^2}\right\rfloor = \left\lfloor \frac{kq^{n-1}-kq^{n-2}+kq^{n-3}-\dots +kq^2-kq+k-q^{n-1}(q-p^b)-1}{q^2}\right\rfloor \)\(= \left\lfloor k\frac{q^{n-2}+1}{q+1}+\frac{-kq+k-1}{q^2}-q^{n-3}(q-p^b)\right\rfloor =\left\lfloor k\frac{q^{n-2}+1}{q+1} -q^{n-3}(q-p^b)-1 +\frac{q^2-k(q-1)-1}{q^2} \right\rfloor \). By the bounds on k we deduce that \(0\le \frac{q^2-k(q-1)-1}{q^2}\le \frac{p^b(q-1)}{q^2}<1\). So,
Let now \(\ell =k\frac{q^{n-2}+1}{q+1}-j\). Notice that \(kM-j=k(M-\frac{q^{n-2}+1}{q+1})+\ell =k(q^{n-1}-q^{n-2})+\ell \). The bounds of \(\ell \) are \(q^{n-3}(q-p^b)+1 \le \ell \le k\frac{q^{n-2}+1}{q+1}\). \(\square \)
Let \(G=\left\{ kA+\ell : k_0\le k\le k_1, \ell _0 \le \ell \le k\ell _1 \right\} \) and let \(mG=\{a_1+\dots +a_m:a_i\in G\}\). Define \(B_m=[(m-1)(q^n+1)+1,m(q^n+1)]\cap mG\).
Lemma B.2
The following statements hold.
-
(1)
\(S'=\{0\}\cup \bigcup _{m\ge 1} mG\),
-
(2)
\(S'=\{0\}\cup \sqcup _{m\ge 1} B_m\).
Proof
First of all, notice that \(mG=\underbrace{[mk_0A+m\ell _0,mk_0(A+\ell _1)]}_{mk_0\ell _1 -m\ell _0 +1 } \cup \underbrace{[(mk_0+1)A+m\ell _0,(mk_0+1)(A+\ell _1)]}_{(mk_0+1)\ell _1 -m\ell _0 +1 } \cup \dots \cup \underbrace{[mk_1A+m\ell _0,mk_1(A+\ell _1)]}_{mk_1\ell _1-m\ell _0+1}\)
-
(1)
The first part is obvious and follows from the definitions.
-
(2)
For the second part, it is obvious that the sets \(B_m\) are disjoint and it is obvious the inclusion \(\supseteq \). Let us prove for all m the inclusion \(S'\cap [(m-1)(q^n+1)+1,m(q^n+1)]\subseteq B_m\) by induction on m.
First of all we need to see that \(S'\cap [1,q^n+1]=B_1\). The smallest element of 2G is \(2(k_0A+\ell _0)= 2((q+1-p^b)(q^{n-1}-q^{n-2})+q^{n-3}(q-p^b)+1)= 2(q^n+1-p^b(q^{n-1}-q^{n-2}+q^{n-3}))\)\(= q^n+1+(q^n+1-2p^b(q^{n-1}-q^{n-2}+q^{n-3})) > q^n+1+(q^n+1-2p^b\frac{q^n+1}{q+1}) \ge q^n+1\) if \(2p^b\le q+1\), which is a consequence of the fact that \(p^b<q\).
Now suppose \(m>1\). Since the maximum of mG is \(m(q^n+1)\), we have \(mG\subseteq [0,m(q^n+1)]\). Now it will suffice to see that \(mG\cap [0,(m-1)(q^n+1)]\subseteq (m-1)G\) and the result will follow by induction.
Notice that mG is the union of the sets of the form \(S_{m,\tilde{k}}=[\tilde{k} A+m\ell _0,\tilde{k}(A+\ell _1)]\) for some \(\tilde{k}\) satisfying \(mk_0\le \tilde{k}\le m k_1\), while \((m-1)G\) is the union of sets of the form \(S_{(m-1),\tilde{\tilde{k}}}[\tilde{\tilde{k}} A+m\ell _0,\tilde{\tilde{k}}(A+\ell _1)]\) for some \(\tilde{\tilde{k}}\) satisfying \((m-1)k_0\le \tilde{\tilde{k}}\le (m-1) k_1\). Suppose that \(a\in mG\cap [0,(m-1)(q^n+1)]\). If \(a\in S_{m\tilde{k}}\) with \(mk_0\le \tilde{k}\le (m-1)k_1\), then, since \(S_{m,\tilde{k}}\subseteq S_{m-1,\tilde{k}}\), we have \(a\in S_{m-1,\tilde{k}}\subseteq (m-1)G\). On the other hand, if \(a\in S_{m,\tilde{k}}\cap [0,(m-1)(q^n+1)]\) with \((m-1)k_1<\tilde{k} \le mk_1\), then \(a\ge \tilde{k} A+ m\ell _0 >(m-1)k_1 A +(m-1)\ell _0\). So, \(a\in S_{m-1,(m-1)k_1}\subseteq (m-1)G\). \(\square \)
1.2 B.2 Number of gaps by intervals
Let \(C_m=[(m-1)(q^n+1)+1,m(q^n+1)]{\setminus } B_m\). In this section we wonder what are the elements in \(C_m\). As before, we split the elements in mG into (not necessarily disjoint) blocks of the form \(S_{m,\tilde{k}}=[\tilde{k} A+m\ell _0,\tilde{k}(A+\ell _1)]\) for some \(\tilde{k}\) satisfying \(mk_0\le \tilde{k}\le m k_1\).
Lemma B.3
-
(1)
Suppose that \(mk_0< k\le mk_1\). Then the gaps between \(S_{m,(k-1)}\) and \(S_{m,k}\) are contained in \([(m-1)(q^n+1)+1,m(q^n+1)]\) if and only if \( k\ge \max \left( mk_0+1,mq+m-q\right) \)
-
(2)
\(\max \left( mk_0+1,mq+m-q\right) =mk_0+1\) if and only if \(m\le M_1:=p^{a-b}\).
Proof
-
(1)
Suppose that \(mk_0< k\le mk_1\). Then the gaps between \(S_{m,(k-1)}\) and \(S_{m,k}\) are contained in \([(m-1)(q^n+1)+1,m(q^n+1)]\) if and only if \((k-1)\frac{q^n+1}{q+1}\ge (m-1)(q^n+1)\), that is, if and only if \(k\ge (m-1)(q+1)+1=mq+m-q\).
-
(2)
\(\max \left( mk_0+1,mq+m-q\right) =mk_0+1\) if and only if \(mq+m-q\le m(q+1-p^b)+1\), that is, if and only if \(-q\le -mp^b+1\), i.e., \(mp^b\le q+1\). Now observe that the quotient of the Euclidean division of \(q+1\) by \(p^b\) is \(p^{a-b}\) while the remainder is 1. So, the statement follows.\(\square \)
Lemma B.4
-
(1)
Suppose that \(mk_0< k\le mk_1\). Then there are gaps between \(S_{m,(k-1)}\) and \(S_{m,k}\) if and only if \(k\le \min \left( mk_1,\frac{q^{n}-q-1+m(q+1)(q^{n-3}(q-p^b)+1)}{q^{n-2}+1}\right) .\)
-
(2)
If \(n>3\), \(\min \left( mk_1, \frac{q^{n}-q+m(q+1)(q^{n-3}(q-p^b)+1)}{q^{n-2}+1}\right) =mk_1\) if and only if \(m\le M_2:=(q-1)p^{a-b}\).
-
(3)
If \(n=3\), \(\min \left( mk_1, \frac{q^{n}-q+m(q+1)(q^{n-3}(q-p^b)+1)}{q^{n-2}+1}\right) =mk_1\) if and only if \(m\le \tilde{M_2}:=(q-1)p^{a-b}-1\).
Proof
-
(1)
Suppose that \(mk_0< k\le mk_1\). Then there are gaps between \(S_{m,( k-1)}\) and \(S_{m, k}\) if and only if \(( k-1)(A+\ell _1)\le k A+m\ell _0-2\), equivalently, \(( k-1)\ell _1\le m\ell _0-2+A\), equivalently, \( k\le \frac{m\ell _0-2+A}{\ell _1}+1= (q+1)\frac{m(q^{n-3}(q-p^b)+1)-2+q^{n-1}-q^{n-2}}{q^{n-2}+1}+1\)\(= \frac{m(q+1)(q^{n-3}(q-p^b)+1)-q-2+q^{n}-q^{n-2}+q^{n-2}+1}{q^{n-2}+1} = \frac{ q^{n} -q-1 +m((q+1)q^{n-3}(q-p^b)+1) }{q^{n-2}+1}\).
-
(2)
\(\min \left( mk_1, \frac{q^{n}-q-1+m(q+1)(q^{n-3}(q-p^b)+1)}{q^{n-2}+1}\right) =mk_1\) if and only if \(m(q+1)\le \frac{q^{n}-q-1+m(q+1)(q^{n-3}(q-p^b)+1)}{q^{n-2}+1}\), that is, if and only if \(m(q^{n-1}+q^{n-2}+q+1)\le q^{n}-q-1+m(q+1)(q^{n-3}(q-p^b)+1)\), i.e., \(m(q^{n-1}+q^{n-2}+q+1)\le q^{n}-q-1+m(q^{n-2}(q-p^b)+q^{n-3}(q-p^b)+q+1)\), i.e., \(0\le q^{n}-q-1-mp^b(q+1)q^{n-3}\), i.e., \(m\le \lfloor \frac{q^n-q-1}{p^b(q+1)q^{n-3}}\rfloor \)
Here we notice that the Euclidean division of \(q^n-q-1\) by \(q^{n-2}p^b+q^{n-3}p^b\) has quotient \((q-1)p^{a-b}\) and remainder \(q^n-q-1-(q-1)p^{a-b}(q^{n-2}p^b+q^{n-3}p^b)= q^n-q-1 -q^{n}-q^{n-1} +q^{n-1}+q^{n-2} = q^{n-2} -q-1 \).
So, the statement follows.
-
(3)
It can be proved as the previous item.\(\square \)
Lemma B.5
If \((q-1)p^{a-b}+1\le m\le qp^{a-b}-1\) then \(\frac{q^{n}-q-1+m(q+1)(q^{n-3}(q-p^b)+1)}{q^{n-2}+1}\) is not an integer and \(\lfloor \frac{q^{n}-q-1+m(q+1)(q^{n-3}(q-p^b)+1)}{q^{n-2}+1} \rfloor =q^2-q+m(q-p^b+1)\).
Proof
\(\frac{q^{n}-q-1+m(q+1)(q^{n-3}(q-p^b)+1)}{q^{n-2}+1}= \frac{q^{n}-q+1}{q^{n-2}+1}+ m\frac{(q+1)(q^{n-3}(q-p^b)+1)}{q^{n-2}+1} \) \(=q^2-\frac{q^2+q+1}{q^{n-2}+1} +m(q-p^b+1)-mp^b\frac{q^{n-3}-1}{q^{n-2}+1}\) \(= q^2-q+m(q-p^b+1)+q-\frac{q^2+q+1}{q^{n-2}+1}-mp^b\frac{q^{n-3}-1}{q^{n-2}+1}. \)
Now, it is enough to see that \(0\le q-\frac{q^2+q+1}{q^{n-2}+1}-mp^b\frac{q^{n-3}-1}{q^{n-2}+1}<1\).
On one hand, \( q-\frac{q^2+q+1}{q^{n-2}+1} - mp^b( \frac{q^{n-3}-1}{q^{n-2}+1}) \ge q- \frac{q^2+q+1}{q^{n-2}+1} -(qp^{a-b}-1)(\frac{(q^{n-3}-1)p^b}{q^{n-2}+1})\)\( = q- \frac{q^2+q+1}{q^{n-2}+1} - \frac{(q^{n-3}-1)q^2-p^{b}(q^{n-3}-1)}{q^{n-2}+1}\)\( = q- \frac{q^2+q+1}{q^{n-2}+1} - \frac{q^{n-1}-q^2-q^{n-3}p^b+p^b}{q^{n-2}+1} = q+ \frac{-q^{n-1}+q^{n-3}p^b-p^b+q+1}{q^{n-2}+1}\)\( = \frac{q+q^{n-3}p^b-p^b+q+1}{q^{n-2}+1} = \frac{2q+p^b(q^{n-3}-1)+1}{q^{n-2}+1} >0\).
On the other hand, \( q- \frac{q^2+q+1}{q^{n-2}+1} - mp^b\frac{q^{n-3}-1}{q^{n-2}+1} \le \frac{q^{n-1}+q-q^2-q-1-(qp^{a-b}-p^{a-b}+1)(q^{n-3}p^b-p^b) }{q^{n-2}+1}\)\(= \frac{q^{n-1}-q^2-1 -qp^{a-b}(q^{n-3}p^b-p^b) +p^{a-b}(q^{n-3}p^b-p^b) -(q^{n-3}p^b-p^b) }{q^{n-2}+1}\)\( = \frac{q^{n-1}-q^2-1 -q^{n-1}+q^2 +q^{n-2}-q -q^{n-3}p^b+p^b }{q^{n-2}+1} = \frac{ -1 +q^{n-2}-q -q^{n-3}p^b+p^b }{q^{n-2}+1}\)\( = \frac{ q^{n-2}+1-(2+q+(q^{n-3}-1)p^b) }{q^{n-2}+1} <1. \)\(\square \)
Lemma B.6
Suppose that \(mk_0< k\le mk_1\). Then the number of gaps between \(S_{m,(k-1)}\) and \(S_{m,k}\) is \(m(q^{n-3}(q-p^b)+1)-{ k}\frac{q^{n-2}+1}{q+1}+\frac{q^n+1}{q+1}-1\).
Proof
The number of gaps between \(S_{m,(k-1)}\) and \(S_{m, k}\) is \(kA+m\ell _0-(k-1)(A+\ell _1)-1= m\ell _0-k\ell _1 +A+\ell _1-1\), which yields the formula in the statement. \(\square \)
Lemma B.7
-
(1)
There are gaps that are at least \((m-1)(q^n+1)+1\) and which are smaller than the elements in \(S_{m,mk_0}\) if and only if \(m\le p^{a-b}=M_1\).
-
(2)
If \(m\le p^{a-b}=M_1\), then the number of gaps between \((m-1)(q^n+1)+1\) and \(S_{m,mk_0}\) is \(q^n-mp^b(q^{n-1}-q^{n-2}+q^{n-3})\).
Proof
-
(1)
There are gaps that are at least \((m-1)(q^n+1)+1\) and which are smaller than the elements in \(S_{m,mk_0}\) if and only if \(mk_0A+m\ell _0\ge (m-1)(q^n+1)+2\). This is equivalent to \(m(q+1-p^b)(q^{n-1}-q^{n-2})+m (q^{n-3}(q-p^b)+1)\ge (m-1)(q^n+1)+2\), that is, if and only if \( mq^n-mq^{n-1}+mq^{n-1}-mq^{n-2}-mq^{n-1}p^b+mq^{n-2}p^b + mq^{n-2}-mq^{n-3}p^b+m \ge mq^n +m -q^n-1 +2\), which is equivalent to \(q^n-1\ge mp^b(q^{n-1}-q^{n-2}+q^{n-3})\) i.e., \(m\le \lfloor \frac{q^n-1}{p^b(q^{n-1}-q^{n-2}+q^{n-3})}\rfloor \).
Here we remark that the Euclidean division of \(q^n-1\) by \(p^b(q^{n-1}-q^{n-2}+q^{n-3})\) has quotient \(p^{a-b}\) and remainder \(q^n-1-(q^{n}-q^{n-1}+q^{n-2}) =q^{n-1}-q^{n-2}-1\). So, the result follows.
-
(2)
It follows from the formula \(mk_0A+m\ell _0-(m-1)(q^n+1)-1\) and a similar simplification as before. \(\square \)
Lemma B.8
Let \(M_3=qp^{a-b}-1\). The set \(C_m\) is not empty if and only if \(m\le M_3\).
Proof
It is clear that for \(m<M_2\), \(C_m\ne \emptyset \). For \(m\ge M_2\), \(C_m\ne \emptyset \) if and only if
This is equivalent to \(q^n-q-1+m(q+1)(q^{n-3}(q-p^b)+1-(q^{n-2}+1))\ge -q(q^{n-2}+1)\), which in turn is equivalent to \(q^n-1-m(q+1)(p^bq^{n-3})\ge -q^{n-1}\), i.e., \(m\le \lfloor \frac{(q+1)q^{n-1}-1}{(q+1)p^bq^{n-3}}\rfloor =qp^{a-b}-1\). \(\square \)
Corollary 1
\(S'=\sqcup _{m=1}^{M_3}B_m\)
Theorem 1
The genus of \(S'\) is \(\frac{q^{n+2}-p^bq^n-q^3+q^2}{2p^b}\).
Proof
By Lemmas B.1, B.2, B.3, B.4, B.5, B.6, B.7, B.8, and Corollary 1, it easily follows that the genus of \(S'\) is
where
A Sage simplification of this leads to
\(\square \)
We remark here that the result does not vary if we replace \(M_2=(q-1)p^{a-b}\) by \(M_2'=(q-1)p^{a-b}-1\).
Rights and permissions
About this article
Cite this article
Castellanos, A.S., Bras-Amorós, M. Weierstrass semigroup at \(m+1\) rational points in maximal curves which cannot be covered by the Hermitian curve. Des. Codes Cryptogr. 88, 1595–1616 (2020). https://doi.org/10.1007/s10623-020-00757-4
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10623-020-00757-4