Finite local nearrings on metacyclic Miller-Moreno p-groups

In this paper the metacyclic Miller-Moreno p-groups which appear as the additive groups of finite local nearrings are classified.

Збережено в:
Бібліографічні деталі
Опубліковано в: :Algebra and Discrete Mathematics
Дата:2012
Автори: Raievska, I.Yu., Sysak, Ya.P.
Формат: Стаття
Мова:Англійська
Опубліковано: Інститут прикладної математики і механіки НАН України 2012
Онлайн доступ:https://nasplib.isofts.kiev.ua/handle/123456789/152191
Теги: Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Цитувати:Finite local nearrings on metacyclic Miller-Moreno p-groups / I.Yu. Raievska, Ya.P. Sysak // Algebra and Discrete Mathematics. — 2012. — Vol. 13, № 1. — С. 111–127. — Бібліогр.: 12 назв. — англ.

Репозитарії

Digital Library of Periodicals of National Academy of Sciences of Ukraine
_version_ 1859660093954457600
author Raievska, I.Yu.
Sysak, Ya.P.
author_facet Raievska, I.Yu.
Sysak, Ya.P.
citation_txt Finite local nearrings on metacyclic Miller-Moreno p-groups / I.Yu. Raievska, Ya.P. Sysak // Algebra and Discrete Mathematics. — 2012. — Vol. 13, № 1. — С. 111–127. — Бібліогр.: 12 назв. — англ.
collection DSpace DC
container_title Algebra and Discrete Mathematics
description In this paper the metacyclic Miller-Moreno p-groups which appear as the additive groups of finite local nearrings are classified.
first_indexed 2025-11-30T09:36:50Z
format Article
fulltext Jo ur na l A lg eb ra D is cr et e M at h.Algebra and Discrete Mathematics RESEARCH ARTICLE Volume 13 (2012). Number 1. pp. 111 – 127 c© Journal “Algebra and Discrete Mathematics” Finite local nearrings on metacyclic Miller-Moreno p-groups I. Yu. Raievska and Ya. P. Sysak Communicated by G. Pilz Abstract. In this paper the metacyclic Miller-Moreno p- groups which appear as the additive groups of finite local nearrings are classified. Introduction Nearrings are a generalization of associative rings in the sense that with respect to addition they need not be commutative and only one distributive law is assumed. In this paper the concept "nearring" means a left distributive nearring with a multiplicative identity. The reader is referred to the books by J. Meldrum [9] or G. Pilz [11] for terminology, definitions and basic facts concerning nearrings. A nearring R is called local if the set of all non-invertible elements of R forms a subgroup of the additive group of R. A study of local nearrings was initiated by Maxson [5] who defined a number of their basic properties and proved in particular that the additive group of a finite zero-symmetric local nearring is a p-group. It follows from [3] that local nearrings with cyclic additive group are commutative local rings. In [6] Maxson described all non-isomorphic zero-symmetric local nearrings with non-cyclic additive group of order p2 which are not nearfields. He also shown in [7] that every non-cyclic abelian p-group of order pn > 4 is the additive group of a zero-symmetric local nearring which is not a ring. This result was 2000 Mathematics Subject Classification: 16Y30. Key words and phrases: Nearring with identity, local nearring, additive group, Miller-Moreno group. Jo ur na l A lg eb ra D is cr et e M at h. 112 Local nearrings extended in [4] to infinite abelian p-groups of finite exponent. However in the case of finite non-abelian p-groups the situation is different. For instance, neither a generalized quaternion group nor a non-abelian group of order 8 can be the additive group of a local nearring, as it was noted in [8] (see also [2]). Some other examples of finite p-groups with this property can be found in [4]. On the other hand, it was proved by Maxson in [8] that each metacyclic Miller-Moreno group of order pn and exponent pn−1 with n ≥ 3 if p > 2 and n ≥ 4 if p = 2 can be the additive group of a local nearring. The purpose of our paper is to give a full classification of the metacyclic Miller-Moreno p-groups which appear as the additive groups of finite local nearrings. Moreover, if G is such an additive group, then we describe all possible multiplications ” · ” on G for which the system (G,+, ·) is a local nearring. 1. Preliminaries Recall first some concepts concerned nearrings. Definition 1. A set R with two binary operations + and · is called a (left) nearring if the following statements hold: 1. (R,+) = R+ is a (not necessarily abelian) group with neutral element 0; 2. (R, ·) is a semigroup; 3. x(y + z) = xy + xz for all x, y, z ∈ R. If R is a nearring, then the group R+ is called the additive group of R. As it follows from statement 3, for each subgroup M of R+ and each element x ∈ R the set xM = {x · y|y ∈ M} is a subgroup of R+ and, in particular, x · 0 = 0. If in addition 0 · x = 0, then the nearring R is called zero-symmetric, and if the semigroup (R, ·) is a monoid, i.e. it has an identity element i, then R is a nearring with identity i. In the latter case the group R∗ of all invertible elements of the monoid (R, ·) is called the multiplicative group of R. A subgroup M of R+ is called R∗-invariant if rM ≤ M for each r ∈ R∗, and M is an (R,R)-subgroup, if xMy ⊆ M for arbitrary x, y ∈ R. Definition 2. A nearring R with identity is said to be local if the set L = R \R∗ of all non-invertible elements of R is a subgroup of R+. Some basic properties of local nearrings are described in the following lemma (see [1], Lemmas 3.2, 3.4 and 3.9). Jo ur na l A lg eb ra D is cr et e M at h. I . Yu. Raievska, Ya. P. Sysak 113 Lemma 1. Let R be a local nearring with identity i and L its subgroup of all non-invertible elements of R+. Then the following statements hold: 1) L is an (R,R)-subgroup of R+; 2) each proper R∗-invariant subgroup of R+ is contained in L; 3) if R is finite, then R+ is a p-group for some prime p, the subgroup L is normal in R+ and the factor group R+/L is elementary abelian. Definition 3. A finite non-abelian group whose proper subgroups are abelian is called a Miller-Moreno group. Henceforth for each prime number p and positive integers m,n with m ≥ 2 let G(pm, pn) denote the semidirect product <a> ⋊ <b> of two cyclic groups <a> and <b> of orders pm and pn, respectively, in which b−1ab = a1+pm−1 . It is well-known that each metacyclic Miller-Moreno p-group is isomorphic either to a quaternion group Q8 or to a group G(pm, pn) (see [12]). Recall that the exponent of a finite p-group is the maximal order of its elements. The following assertion is easily verified. Lemma 2. The exponent of G(pm, pn) is equal to pm for m > n and to pn for m ≤ n. Moreover, if x is an element of maximal order in G(pm, pn), then there exist generators a, b of this group such that either a = x or b = x and the relations ap m = bp n = 1 and b−1ab = a1+pm−1 hold. Lemma 3. Let the group G(pm, pn) be additively written. Then for any natural numbers r, s, t the equalities bs + ar = ar(1− spm−1) + bs and (ar + bs)t = ar(t− s ( t 2 ) pm−1) + bst hold. Proof. Let q = 1 − pm−1. Since −b + a + b = a(1 + pm−1) and m ≥ 2, then b+ a = aq + b, so bs+ ar = arqs + bs for arbitrary integers r ≥ 0 and s ≥ 0. Taking into consideration, that qs = (1− pm−1)s ≡ 1− spm−1 ( mod pm) by binomial’s formula, giving bs+ ar = ar(1− spm−1) + bs. Next, (ar + bs)t = ar(1 + qs + · · · + qs(t−1)) + bst by induction on t. Therefore, 1 + qs + · · · + qs(t−1) ≡ 1 + (1 − spm−1) + · · · + (1 − s(t − 1)pm−1) = t− s ( t 2 ) pm−1 ( mod pm), thus (ar + bs)t = ar(t− s ( t 2 ) pm−1) + bst. Lemma 4. Put G = G(pm, pn) where m ≤ n. If A = AutG is the group of all automorphisms of G and <x> is a cyclic subgroup of G, then the following statements hold: Jo ur na l A lg eb ra D is cr et e M at h. 114 Local nearrings 1) if m = n and <x> is a normal subgroup of order pm in G, then |xA| ≤ p2m−2(p− 1); 2) if p > 2 and < x > is a non-normal subgroup of order pn, then x−1 /∈ xA. Proof. If < x > is a normal subgroup of order pm in G , then either < x > is a central subgroup of G or ap m−1 ∈ < x >. Since the center Z(G) =< ap > × < bp > and the derived subgroup G′ =< ap m−1 > are characteristic subgroups of G, it follows that in the first case xA ⊆ Z(G) and so |xA| ≤ p2m−2(p− 1) and in the second case G =<x> ⋊ <b> and ap m−1 ∈ <xα> for each α ∈ A. The latter implies | <x><xα> | = |x||xα| | <x> ∩ <x>α | ≤ p2m−1, so that the subgroup <x><xα> is abelian and thus xα ∈ CG(x). Since CG(x) =<x> × <bp>, for m = n the number of all elements of order pm in CG(x) is equal to p2m−2(p− 1) and hence |xA| ≤ p2m−2(p− 1), as desired. Now let < x > be a non-normal subgroup of order pn in G. Then < a> ∩ <x>= 1 and so x = asbt for some natural numbers s, t such that (t, p) = 1. In particular, G =< a> ⋊ < x> and [a, x] = [a, b]t = atp m−1 . Assume that xα = x−1 for some automorphism α ∈ A. Since < ap m−1 >α=< ap m−1 >, it follows that aα = arxu for some natural numbers r, u such that (r, p) = 1 and p is a divisor of u. Therefore artp m−1 = (atp m−1 )α = [aα, xα] = [arxu, x−1] = [a, x−1]r = a−rtpm−1 and hence a2rtp m−1 = 1. But then pm divides 2rtpm−1 what is impossible if p> 2. Therefore x−1 /∈ xA, as claimed. Lemma 5. The exponent of the additive group of a finite nearring R with identity i is equal to the additive order of i which coincides with additive order of every invertible element of R. Proof. Indeed, if i·k = 0 for some natural number k, then x·k = (x·i)·k = x · (i · k) = x · 0 = 0 for each x ∈ R. On the other hand, if r · l = 0 for r ∈ R∗ and for a natural number l, then i · l = r−1(r · l) = 0, so that the additive orders of r and i coincide. The following assertion is a direct consequence of Lemmas 2 and 5. Corollary 1. Let R be a nearring with identity i whose additive group R+ is isomorphic to a group G(pm, pn). Then there exist generators a, b Jo ur na l A lg eb ra D is cr et e M at h. I . Yu. Raievska, Ya. P. Sysak 115 of R+ satisfying relations apm = bpn = 0 and a + b = b + a(1 + pm−1) such that either a or b coincides with i. Lemma 6. Let R be a local nearring whose additive group R+ is isomor- phic to a group G(pm, pn) and let L be the subgroup of all non-invertible elements of R. Then L is a subgroup in R+ of index |R+ : L| = p. Proof. Assume that |R+ : L| = pk with k ≥ 2. Since R = R∗ ∪ L and R∗ ∩ L = ∅, we have |R∗| = pm+n − pm+n−k = pm+n−k(pk − 1). In particular, the order of R∗ is divisible by pk − 1. On the other hand, for each element r ∈ R∗ the mapping x 7→ rx with x ∈ R is an automorphism of R+ because r(x + y) = rx + ry for every x, y ∈ R. Therefore R∗ can be viewed as a subgroup of the group of automorphisms AutR+. As it follows from [10], Corollary of Theorem 2, the order of the group AutG(pm, pn) and so AutR+ is a number of the form pl(p− 1) for some l ≥ m+ n, and therefore it is divisible by pk − 1 only if k = 1. This contradiction shows that |R+ : L| = p. 2. Nearrings with identity on metacyclic Miller-Moreno p-groups As it was mentioned in the Introduction, the quaternion group Q8 cannot be isomorphic to the additive group of a nearring R with identity. Therefore in what follows the additive group of R is isomorphic to a group G(pm, pn). Thus it follows from Corollary 1 that R+ =<a> + <b> with elements a, b one of which coincides with identity element of R and the relations apm = bpn = 0 and a+ b = b+ a(1 + pm−1) are valid. Moreover, each element x ∈ R is uniquely written in the form x = ax1 + bx2 with coefficients 0 ≤ x1 < pm and 0 ≤ x2 < pn. Consider first the case when a coincides with identity element of R, so that xa = ax = x for each x ∈ R. Then R+ is of exponent pm by Lemma 5 and so m ≥ n. Furthermore, for each x ∈ R there exist integers α(x) and β(x) such that xb = aα(x) + bβ(x). It is clear that modulo pm and pn, respectively, these integers are uniquely determined by x and so some mappings α : R → Zpm and β : R → Zpn are determined. Lemma 7. Let x = ax1 + bx2 and y = ay1 + by2 be elements of R. If a coincides with identity element of R, then xy = a[(x1y1 + α(x)y2)− (x1x2 ( y1 2 ) + α(x)x2y1y2 + α(x)β(x) ( y2 2 ) )pm−1] + b(x2y1 + β(x)y2). Jo ur na l A lg eb ra D is cr et e M at h. 116 Local nearrings Moreover, for the mappings α : R → Zpm and β : R → Zpn the following statements hold : (0) α(0) = β(0) = 0 if and only if the nearring R is zero-symmetric; (1) α(a) = 0 and β(a) = 1; (2) α(x) ≡ 0 ( mod pm−n), except the case p = 2 = m and n = 1; (3) x1(β(x)− 1) ≡ x2α(x) ( mod p ); (4) α(xy) = x1α(y) + α(x)β(y) − [x1x2 ( α(y) 2 ) + α(x)x2α(y)β(y) + α(x)β(x) ( β(y) 2 ) ]pm−1; (5) β(xy) = x2α(y) + β(x)β(y). Proof. By the left distributive law, we have xy = (xa)y1 + (xb)y2 = (ax1 + bx2)y1 + (aα(x) + bβ(x))y2. Furthermore, Lemma 3 implies that (ax1 + bx2)y1 = ax1(y1 − x2 ( y1 2 ) pm−1) + bx2y1, (aα(x) + bβ(x))y2 = aα(x)(y2 − β(x) ( y2 2 ) pm−1) + bβ(x)y2 and bx2y1 + aα(x)(y2 − β(x) ( y2 2 ) pm−1) = aα(x)(y2 − β(x) ( y2 2 ) pm−1)(1− x2y1p m−1) + bx2y1. Thus (∗) xy = a[(x1y1 + α(x)y2)− (x1x2 ( y1 2 ) + α(x)x2y1y2 + α(x)β(x) ( y2 2 ) )pm−1] + b(x2y1 + β(x)y2), as desired. As 0 · a = a · 0 = 0, the nearring R is zero-symmetric if and only if 0 = 0 · b = aα(0) + bβ(0) whence α(0) = β(0) = 0. Similarly, from the equality b = ab = aα(a) + bβ(a) it follows that α(a) = 0 and β(a) = 1. Since (xb)pn = x(bpn) = 0 and xb = aα(x) + bβ(x), we have 0 = (aα(x) + bβ(x))pn = aα(x)(pn − β(x) ( pn 2 ) pm−1) by Lemma 3 and hence Jo ur na l A lg eb ra D is cr et e M at h. I . Yu. Raievska, Ya. P. Sysak 117 α(x) ≡ 0 ( mod pm−n), except the case p = 2 = m and n = 1 in which the group R+ is dihedral of order 8. Next, if y = a(1− pm−1) + b, then xy = a[α(x) + x1 − (x1 + α(x)x2)p m−1] + b(x2 + β(x)) by formula (*). On the other hand , y = b+ a and so xy = xb+ x = a(α(x) + x1 − x1β(x)p m−1) + b(x2 + β(x)) by Lemma 3. Comparing both results for xy, we obtain (x1 + α(x)x2)p m−1 ≡ x1β(x)p m−1( mod pm). Thus x1(β(x)− 1) ≡ x2α(x) ( mod p ) and so statement (3) holds. Finally, the associativity of multiplication in R implies that x(yb) = (xy)b = aα(xy) + bβ(xy). Furthermore, substituting yb = aα(y) + bβ(y) instead of y in formula (*), we also have x(yb) = a[(x1α(y) + α(x)β(y))− (x1x2 ( α(y) 2 ) + α(x)x2α(y)β(y) + α(x)β(x) ( β(y) 2 ) )pm−1] + b(x2α(y) + β(x)β(y)). Comparing the coefficients under a and b in two expressions obtained for x(yb), we derive statements (4) and (5) of the lemma. Consider now the case when b coincides with identity element of R and so xb = bx = x for each x ∈ R. Then R+ is of exponent pn by Lemma 5 and thus m ≤ n. As above, there exist integers α(x) and β(x) which are uniquely determined by x modulo pm and pn, respectively, such that xa = aα(x) + bβ(x). Lemma 8. If b coincides with identity element in R, then xy = a[α(x)y1 + x1y2 − x1x2 ( y2 2 ) pm−1] + b[β(x)y1 + x2y2]. Also for the mappings α : R → Zpm and β : R → Zpn the following statements hold: (0) α(0) = β(0) = 0 if and only if the nearring R is zero-symmetric; (1) α(b) = 1 and β(b) = 0; Jo ur na l A lg eb ra D is cr et e M at h. 118 Local nearrings (2) β(x) ≡ 0 ( mod pn−m+1); (3) α(x)(1− x2) ≡ 0 ( mod p ); (4) α(xy) = α(x)α(y) + x1β(y)− x1x2 ( β(y) 2 ) pm−1; (5) β(xy) = β(x)α(y) + x2β(y). Proof. Clearly R is zero-symmetric if and only if 0 = 0 ·a = aα(0)+ bβ(0), whence α(0) = β(0) = 0. From the equality a = ba = aα(b) + bβ(b) it follows that α(b) = 1 and β(b) = 0. Using the left distributive law, we have also xy = (xa)y1 + (xb)y2 = (aα(x) + bβ(x))y1 + (ax1 + bx2)y2. Next, Lemma 3 implies that (aα(x) + bβ(x))y1 = aα(x)(y1 − β(x) ( y1 2 ) pm−1) + bβ(x)y1, (ax1 + bx2)y2 = ax1(y2 − x2 ( y2 2 ) pm−1) + bx2y2 and bβ(x)y1 + ax1(y2 − x2 ( y2 2 ) pm−1) = ax1(y2 − x2 ( y2 2 ) pm−1)(1− β(x)y1p m−1) + bβ(x)y1. Therefore (∗∗) xy = a[(α(x)y1 + x1y2)− (α(x)β(x) ( y1 2 ) + x1x2 ( y2 2 ) + + x1β(x)y1y2)p m−1] + b(β(x)y1 + x2y2). Substituting y = a(1− pm−1) + b in formula (**), we derive xy = a[α(x)(1− pm−1) + x1 − (α(x)β(x) ( 1−pm−1 2 ) + x1x2 ( 1 2 ) + x1β(x)(1− pm−1))pm−1] + b(β(x)(1− pm−1) + x2). In the case p > 2 or p = 2 and m ≥ 3 this implies (i) xy = a[x1 +α(x)− (α(x) + x1β(x))p m−1] + b(β(x)(1− pm−1) + x2). If p = 2 and m = 2, then (ii) xy = a[x1 − α(x) + 2(α(x)β(x) + x1β(x))] + b(x2 − β(x)). Jo ur na l A lg eb ra D is cr et e M at h. I . Yu. Raievska, Ya. P. Sysak 119 On the other hand, y = b+ a and thus (iii) xy = x+ xa = a[x1 + α(x)− x2α(x)p m−1] + b(x2 + β(x)). Comparing formulas (i) and (ii) with formula (iii), we obtain in the first case the congruences (α(x) + x1β(x))p m−1 ≡ x2α(x)p m−1( mod pm) and β(x)pm−1 ≡ 0( mod pn) and in the second case the congruences −α(x) + 2α(x)β(x) + 2x1β(x) ≡ α(x)− 2x2α(x)( mod 22) and β(x) ≡ −β(x)( mod 2n). In both cases this gives the congruences α(x)(1− x2) ≡ 0 ( mod p ) and β(x) ≡ 0 ( mod pn−m+1), i.e. statements (3) and (2), respectively. But then (α(x)β(x) ( y1 2 ) + x1β(x)y1y2)p m−1 ≡ 0 ( mod pm) and so formula (**) becomes the equality xy = a[(α(x)y1 + x1y2)− x1x2 ( y2 2 ) pm−1] + b[β(x)y1 + x2y2], as desired. Finally, replacing in this equality the element y by ya = aα(y) + bβ(y) and using that x(ya) = (xy)a, we have got the expressions for α(xy) and β(xy) contained in statements (4) and (5) of the lemma. 3. Local nearrings on metacyclic Miller-Moreno p-groups Let R be a local nearring whose additive group R+ is isomorphic to a group G(pm, pn) and let L denote the subgroup in R+ of all non- invertible elements from R. As it was mentioned above, R = R∗ ∪ L and R+ =< a > + < b > with elements a, b satisfying the relations apm = bpn = 0 and a+ b = b+ a(1 + pm−1). Furthermore, < a > is a normal subgroup of R+ and each element x ∈ R is uniquely written in the form x = ax1 + bx2 with coefficients 0 < x1 < pm and 0 < x2 < pn. Jo ur na l A lg eb ra D is cr et e M at h. 120 Local nearrings Lemma 9. 1) If a ∈ R∗, then pm+n > 8 and m> n. 2) If b ∈ R∗, then p = 2 and m ≤ n. Proof. If a ∈ R∗, then R+ is a group of exponent pm by Lemma 5 and so m ≥ n. It was shown in [8] that the dihedral group of order 8 is not the additive group of a local nearring, so that pm+n > 8. Assume that m = n. As it was shown in the proof of Lemma 6, the multiplicative group R∗ can be viewed as a subgroup of AutR+ such that R∗x ⊆ xAutR+ for each x ∈ R and, in particular, R∗ = R∗a ⊆ aAutR+ . Since |aAutR+ | ≤ p2m−2(p − 1) by Lemma 4, it follows that |R∗| ≤ p2m−2(p− 1). On the other hand, the subgroup L is of index p in R+, so that |R∗| = |R| − |L| = |R| − 1 p |R| = p2m−p2m−1 = p2m−1(p−1), contrary to the latter inequality. Thus m> n, proving statement 1). Now let b ∈ R∗. Then R+ is a group of exponent pn and so m ≤ n. Since the subgroup <b> is non-normal in R+, it follows that −b 6∈ bAutR+ by Lemma 4. As R∗ = R∗b ⊆ bAutR+ , this means that −b 6∈ R∗. But then −b ∈ L and if p > 2, then b ∈ L, contrary to the assumption. Thus p = 2 and m ≤ n, proving statement 2). Theorem 1. Let R be a local nearring whose additive group R+ is isomor- phic to a group G(pm, pn). Then R+ = <a> + <b>, one of the elements a, b coincides with identity element of R and the following statements hold: 1) apm = bpn = 0 and a+ b = b+ a(1 + pm−1); 2) if a coincides with identity element of R, then pm+n > 8, m > n, L = <ap> + <b> and R∗ = {ax1 + bx2 | x1 6≡ 0 (mod p)}; 3) if b coincides with identity element of R, then p = 2, m ≤ n, L = <a> + <b2> and R∗ = {ax1 + bx2 | x2 ≡ 1 ( mod 2 )}. Proof. By Corollary 1, there exists the decomposition R+ =<a> + <b> in which one of the elements a or b coincides with identity element of R and statement 1) holds. If a coincides with identity element, then m> n and b ∈ L by Lemma 9, so that L = <ap> + <b>. Since xL ⊆ L for each x ∈ R by Lemma 1, we have xb = aα(x) + bβ(x) ∈ L whence α(x) ≡ 0 ( mod p ) for each x ∈ R. Furthermore, R = R∗∪L and so an element x = ax1+ bx2 belongs to R∗ if and only if x1 6≡ 0 ( mod p ). Similarly, if b coincides with identity element of R, then p = 2, m ≤ n and a ∈ L by Lemma 9. Therefore L = <a> + <b2> and β(x) ≡ Jo ur na l A lg eb ra D is cr et e M at h. I . Yu. Raievska, Ya. P. Sysak 121 0 ( mod 2 ) for each x ∈ R. Thus an element x = ax1 + bx2 belongs to R∗ if and only if x2 ≡ 1 ( mod 2 ). Applying statements 2) and 3) of Theorem 1 to Lemmas 7 and 8, respectively, we obtain the following formulas for multiplying elements x = ax1+bx2 and y = ay1+by2 in the local nearring R whose the additive group is isomorphic to a group G(pm, pn). Corollary 2. If a coincides with identity element of R and xb = aα(x) + bβ(x), then xy = a(x1y1 + α(x)y2 − x1x2 ( y1 2 ) pm−1) + b(x2y1 + β(x)y2), where the mappings α : R → Zpm and β : R → Zpn satisfy the conditions: (0) α(0) = β(0) = 0 if and only if the nearring R is zero-symmetric; (1) α(a) = 0 and β(a) = 1; (2) α(x) ≡ 0 ( mod pm−n); (3) x1(β(x)− 1) ≡ 0 ( mod p ); (4) α(xy) = x1α(y) + α(x)β(y)− x1x2 ( α(y) 2 ) pm−1; (5) β(xy) = x2α(y) + β(x)β(y). Corollary 3. If b coincides with identity element of R and xa = aα(x) + bβ(x), then xy = a(α(x)y1 + x1y2 − x1x2 ( y2 2 ) 2m−1) + b(β(x)y1 + x2y2), where mappings α : R → Z2m and β : R → Z2n satisfy the conditions: (0) α(0) = β(0) = 0 if and only if the nearring R is zero-symmetric; (1) α(b) = 1 and β(b) = 0; (2) β(x) ≡ 0 ( mod 2n−m+1); (3) α(x) ≡ 0 ( mod 2 ) if and only if x2 ≡ 0 ( mod 2 ); (4) α(xy) = α(x)α(y) + x1β(y)− x1x2 ( β(y) 2 ) 2m−1; (5) β(xy) = β(x)α(y) + x2β(y). Jo ur na l A lg eb ra D is cr et e M at h. 122 Local nearrings 4. Groups G(pm, pn) as the additive groups of local near- rings The following theorem gives sufficient conditions for existing a finite local nearring whose additive group is isomorphic to a group G(pm, pn). Together with Theorem 1 and remarks mentioned in the Introduction this completes our classification of all metacyclic Miller-Moreno p-groups which are the additive groups of local nearrings. Theorem 2. For an arbitrary prime p and natural numbers m,n such that either pm+n > 8 and m > n or p = 2 and 1 < m ≤ n there exists a local nearring R whose additive group R+ is isomorphic to a group G(pm, pn). Proof. Let R be an additively written group G(pm, pn) with generators a, b satisfying the relations apm = bpn = 0 and a+ b = b+ a(1 + pm−1). Then R = <a> + <b> and each element x ∈ R is uniquely written in the form x = ax1+ bx2 with coefficients 0 < x1 < pm and 0 < x2 < pn. In order to define a multiplication ” · ” on R in such a manner that (R,+, ·) is a local nearring, we separately consider two cases. 1. The case pm+n > 8 and m > n. For arbitrary elements x = ax1+ bx2 and y = ay1+ by2 of R we define the multiplication ” · ” by the rule x · y = a(x1y1 + α(x)y2 − x1x2 ( y1 2 ) pm−1) + b(x2y1 + β(x)y2), where α : R → Zpm and β : R → Zpn are mappings satisfying conditions (1) - (5) of Corollary 2. As an example, we can for instance take α(x) = 0 for all x ∈ R and β(x) = { 1 if x1 6≡ 0 ( mod p ), 0 if x1 ≡ 0 ( mod p ). It is easy to see that the element a is a multiplicative identity for (R, ·) and x ·b = aα(x)+bβ(x) for each x ∈ R. We show that with respect to the operations ” + ” and ” · ” the system (R,+, ·) is a nearring with identity element a. Clearly it suffices to check that if z = az1 + bz2 is an arbitrary element of R, then x · (y + z) = x · y + x · z and (x · y) · b = x · (y · b). Indeed, we have x · z = a(x1z1 + α(x)z2 − x1x2 ( z1 2 ) pm−1) + b(x2z1 + β(x)z2) Jo ur na l A lg eb ra D is cr et e M at h. I . Yu. Raievska, Ya. P. Sysak 123 and b[x2y1 + β(x)y2] + a[x1z1 + α(x)z2 − x1x2 ( z1 2 ) pm−1] = = a[(1− (x2y1 + β(x)y2)p m−1)(x1z1 + α(x)z2− − x1x2 ( z1 2 ) pm−1)] + b[x2y1 + β(x)y2] by Lemma 3. Since α(x) ≡ 0 ( mod p ) by condition (2) of Corollary 2 and so a(α(x)pm−1) = 0, this implies x · y + x · z = a[x1(y1 + z1) + α(x)(y2 + z2)− (x1x2 ( y1 2 ) + x1x2 ( z1 2 ) + + x1x2y1z1 + x1β(x)y2z1)p m−1] + b[x2(y1 + z1) + β(x)(y2 + z2)]. On the other hand, y + z = (ay1 + by2) + (az1 + bz2) = a(y1 + z1 − y2z1p m−1) + b(y2 + z2) because by2 + az1 = a(1 − y2p m−1)z1 + by2 by Lemma 3 and b(y2z1p m−1) = 0 because m > n. Therefore x · (y + z) = a[x1(y1 + z1) + α(x)(y2 + z2)− (x1y2z1+ + x1x2 ( y1+z1−y2z1p m−1 2 ) )pm−1] + b[x2(y1 + z1) + β(x)(y2 + z2)] and hence x · y + x · z − x · (y + z) = a[−(x1x2 ( y1 2 ) + x1x2 ( z1 2 ) + x1x2y1z1+ + x1β(x)y2z1)p m−1 + (x1y2z1 + x1x2 ( y1+z1−y2z1p m−1 2 ) )pm−1] = = ax1(1− β(x))y2z1p m−1. Since x1(β(x)−1) ≡ 0 ( mod p ) by condition (3) of Corollary 2, it follows that ax1(1 − β(x))y2z1p m−1 = 0 and thus x · (y + z) = x · y + x · z, as desired. Next, y · b = aα(y) + bβ(y) and so x · (y · b) = xα(y) + (x · b)β(y). Applying Lemma 3, we also have xα(y) = (ax1 + bx2)α(y) = ax1(α(y)− x2 ( α(y) 2 ) pm−1) + bx2α(y), (x · b)β(y) = (aα(x) + bβ(x))β(y) = aα(x)(β(y)− β(x) ( β(y) 2 ) pm−1)+ + bβ(x)β(y) and bx2α(y) + aα(x)(β(y)− β(x) ( β(y) 2 ) pm−1) = = aα(x)(β(y)− β(x) ( β(y) 2 ) pm−1)(1− x2α(y)p m−1) + bx2α(y). As α(x) ≡ 0 ( mod p ), it follows that α(x)β(x) ( β(y) 2 ) pm−1) = x2α(x)α(y)β(y)p m−1 = = x2β(x)α(y) ( β(y) 2 ) p2m−2 = 0, Jo ur na l A lg eb ra D is cr et e M at h. 124 Local nearrings whence x·(y·b) = a[x1α(y)+α(x)β(y)−x1x2 ( α(y) 2 ) pm−1]+b[x2α(y)+β(x)β(y)]. However x1α(y) + α(x)β(y) − x1x2 ( α(y) 2 ) pm−1 = α(x · y) and x2α(y) + β(x)β(y) = β(x · y) by conditions (4) and (5) of Corollary 2, so that (x · y) · b = aα(x · y) + bβ(x · y) = x · (y · b). Thus, the system (R,+, ·) is a nearring with identity element a. We show finally that R = (R,+, ·) is a local nearring in which the subgroup L =<ap> + <b> consists of all non-invertible elements of R. Obviously, it is enough to prove that the element x = ax1 + bx2 is right invertible if and only if x1 6≡ 0 ( mod p ) because in this case the set R \L coincides with the multiplicative group R∗ by Lemma 1. Indeed, if x is right invertible, then there exists an element y = ay1+by2 such as x ·y = a. From our definition of x ·y it follows that x1y1+α(x)y2− x1x2 ( y1 2 ) pm−1 ≡ 1 ( mod pm). But then x1y1 ≡ 1 ( mod p ) because α(x) ≡ 0 ( mod p ) by condition (2) of Corollary 2, so that x1 6≡ 0 ( mod p ). Conversely, if the latter holds, then β(x) ≡ 1 ( mod p ) by condition (3) of Corollary 2 and so there exist natural numbers y1, y2 such that x1y1 ≡ 1 ( mod pm) and x2y1 + β(x)y2 ≡ 0 ( mod pn). Taking y = ay1 + by2 and z1 = 1 + α(x)y2 − x1x2 ( y1 2 ) pm−1, we have x · y = az1. Since α(x) ≡ 0 ( mod p ), there exists such a natural number s that z1s ≡ 1 ( mod pm), whence x · (ys) = (x · y)s = (az1)s = a. Thus the element x is right invertible and so the nearring R is local. 2. The case p = 2 and m ≤ n. Define now the multiplication ” · ” by the rule x · y = a(α(x)y1 + x1y2 − x1x2 ( y2 2 ) 2m−1) + b(β(x)y1 + x2y2), where the mappings α : R → Z2m and β : R → Z2n satisfy conditions (1) - (5) of Corollary 3. In particular, we can put α(x) = { 1 if x2 6≡ 0 ( mod 2 ), 0 if x2 ≡ 0 ( mod 2 ) and β(x) = 0 for all x ∈ R. Clearly the element b is a multiplicative identity of R and x · a = aα(x) + bβ(x) for each x ∈ R. As in the above case, in order to show that the Jo ur na l A lg eb ra D is cr et e M at h. I . Yu. Raievska, Ya. P. Sysak 125 system (R,+, ·) is a nearring with identity element b, it is enough to verify that x · (y + z) = x · y + x · z and (x · y) · a = x · (y · a). Indeed, x · z = a(α(x)z1 + x1z2 − x1x2 ( z2 2 ) 2m−1) + b(β(x)z1 + x2z2) by the definition and b[β(x)y1 + x2y2] + a[α(x)z1 + x1z2 − x1x2 ( z2 2 ) 2m−1] = = a[(1− (β(x)y1 + x2y2)2 m−1)(α(x)z1 + x1z2 − x1x2 ( z2 2 ) 2m−1)]+ + b[β(x)y1 + x2y2] by Lemma 3. Since β(x) ≡ 0 ( mod 2n−m+1) by condition (2) of Corollary 3, this implies that x · y + x · z = a[α(x)(y1 + z1) + x1(y2 + z2)− (x1x2 ( y2 2 ) + x1x2 ( z2 2 ) + + x2y2α(x)z1 + x1x2y2z2)2 m−1] + b[β(x)(y1 + z1) + x2(y2 + z2)]. On the other hand, y+ z = a(y1 + z1 − y2z12 m−1) + b(y2 + z2) and so x · (y + z) = a[α(x)(y1 + z1 − y2z12 m−1) + x1(y2 + z2)− − x1x2 ( y2+z2 2 ) 2m−1] + b[β(x)(y1 + z1 − y2z12 m−1) + x2(y2 + z2)]. Thus −x · (y + z) + x · y + x · z = −b[β(x)(y1 + z1 − y2z12 m−1)+ + x2(y2 + z2)] + a[(α(x)y2z1 + x1x2 ( y2+z2 2 ) − x1x2 ( y2 2 ) − x1x2 ( z2 2 ) − − α(x)x2y2z1 − x1x2y2z2)2 m−1] + b[β(x)(y1 + z1) + x2(y2 + z2)]. However a[(α(x)y2z1 + x1x2 ( y2+z2 2 ) − x1x2 ( y2 2 ) − x1x2 ( z2 2 ) − − α(x)x2y2z1 − x1x2y2z2)2 m−1] = aα(x)(1− x2)y2z22 m−1 = 0 by condition (3) of Corollary 3 and therefore −x · (y + z) + x · y + x · z = −b[β(x)(y1 + z1 − y2z12 m−1)+ + x2(y2 + z2)] + b[β(x)(y1 + z1) + x2(y2 + z2)] = = bβ(x)y2z12 m−1 = 0, as required. Jo ur na l A lg eb ra D is cr et e M at h. 126 Local nearrings Next, y · a = aα(y) + bβ(y) and so x · (y · a) = (x · a)α(y) + xβ(y) = (aα(x) + bβ(x))α(y) + (ax1 + bx2)β(y). Applying Lemma 3, we have (aα(x) + bβ(x))α(y) = aα(x)(α(y)− β(x) ( α(y) 2 ) 2m−1) + bβ(x)α(y), (ax1 + bx2)β(y) = ax1(β(y)− x2 ( β(y) 2 ) 2m−1) + bx2β(y) and bβ(x)α(y) + ax1(β(y)− x2 ( β(y) 2 ) 2m−1) = = ax1(β(y)− x2 ( β(y) 2 ) 2m−1)(1− β(x)α(y)2m−1) + bβ(x)α(y). From this and the congruence β(x) ≡ 0 ( mod 2n−m+1) it follows that x·(y·a) = a[α(x)α(y)+x1β(y)−x1x2 ( β(y) 2 ) 2m−1]+b[β(x)α(y)+x2β(y)]. In view of conditions (4) and (5) of Corollary 3, the coefficients in brackets under a and b in the right part of the last equality coincide with α(x·y) and β(x · y), respectively. Therefore x · (y ·a) = aα(x · y)+ bβ(x · y) = (x · y) ·a and hence (R,+, ·) is a nearring with identity element b. To complete the proof it remains to show that the set of all non- invertible elements of (R,+, ·) coincides with L =<a> + <b2>. As in the previous case, it suffices to prove that an element x = ax1 + bx2 is right invertible if and only if x2 ≡ 1 ( mod 2 ). Really, if there exists an element y = ay1 + by2 such that x · y = b, then β(x)y1 + x2y2 ≡ 1 ( mod 2n) by our definition of x · y. But then x2y2 ≡ 1 ( mod 2m−1) because β(x) ≡ 0 ( mod 2n−m+1) by condition (2) of Corollary 3, whence x2 ≡ 1 ( mod 2 ). Conversely, if the last congruence holds, then α(x) ≡ 1 ( mod 2 ) by condition (3) of Corollary 3 and so there exist natural numbers y1, y2 such that x2y2 ≡ 1 ( mod 2n) and α(x)y1 + x1y2 ≡ 0 ( mod 2m). Therefore for y = ay1 + by2 and z2 = 1 + β(x)y1 we have x · y = bz2. Since β(x) ≡ 0 ( mod 2n−m+1 ), there is a natural number s such that z2s ≡ 1 ( mod 2n). But then x · (ys) = (x · y)s = (bz2)s = b and so the element x is right invertible. Thus the nearring R is local and this completes the proof. References [1] B. Amberg, P. Hubert, Ya. Sysak, Local nearrings with dihedral multiplicative group, Journal of Algebra, 273, 2004, pp. 700-717. Jo ur na l A lg eb ra D is cr et e M at h. I . Yu. Raievska, Ya. P. Sysak 127 [2] J. R. Clay, C.J. Maxson The near-rings with identities on generalized quaternion groups, Ist. Lombardo Accad. Sci. Lett. Rend., A 104, 1970, pp. 525–530. [3] J. R. Clay, J.J. Malone, Jr. The near-rings with identities on certain finite groups, Math. Scand., 19, 1966, pp. 146-150. [4] S. Feigelstock, Additive Groups of Local Near-Rings, Communications in Algebra, 34, 2006, pp.743 - 747. [5] C. J. Maxson, On local near-rings, Math. Zeitschr., 106, 1968, pp. 197-205. [6] C. J. Maxson, Local nearrings of cardinality p 2, Canad. Math. Bull., 11, 1968, pp. 555-561. [7] C. J. Maxson, On the construction of finite local near-rings (I): on non-cyclic abelian p-groups, Quart. J. Math. Oxford (2), 21, 1970, pp. 449-457. [8] C. J. Maxson, On the construction of finite local near-rings (II): on non-abelian p-groups, Quart. J. Math. Oxford (2), 22, 1971, pp. 65-72. [9] J. D. P. Meldrum, Near-rings and their links with groups, Pitman, Boston, 1985, 275 p. [10] V. T. Nagrebeckii, The automorphism groups of finite Miller-Moreno groups (Russian), Dokl. Akad. Nauk Ukrain. SSR Ser. A, No.5, 1980, pp. 75-77. [11] G. Pilz, Near-rings. The theory and its applications (Second edition), North- Holland, Amsterdam, 1983, 470 p. [12] L. Redei, Das “schiefe Produkt” in der Gruppentheorie mit Anwendung auf die endlichen nichtkommutativen Gruppen mit lauter kommutativen echten Untergrup- pen und die Ordnungszahlen, zu denen nur kommutative Gruppen gehören (German), Comment. Math. Helv., 20, 1947, pp. 225–264. Contact information I. Yu. Raievska Institute of Mathematics Ukrainian National Academy of Science, Tereshchenkivska st. 3, Kiev, 01601, Ukraine E-Mail: raemarina@rambler.ru Ya. P. Sysak Institute of Mathematics Ukrainian National Academy of Science, Tereshchenkivska st. 3, Kiev, 01601 Ukraine E-Mail: sysak@imath.kiev.ua Received by the editors: 23.09.2011 and in final form 30.03.2012. I. Yu. Raievska, Ya. P. Sysak
id nasplib_isofts_kiev_ua-123456789-152191
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
issn 1726-3255
language English
last_indexed 2025-11-30T09:36:50Z
publishDate 2012
publisher Інститут прикладної математики і механіки НАН України
record_format dspace
spelling Raievska, I.Yu.
Sysak, Ya.P.
2019-06-08T11:09:34Z
2019-06-08T11:09:34Z
2012
Finite local nearrings on metacyclic Miller-Moreno p-groups / I.Yu. Raievska, Ya.P. Sysak // Algebra and Discrete Mathematics. — 2012. — Vol. 13, № 1. — С. 111–127. — Бібліогр.: 12 назв. — англ.
1726-3255
2000 Mathematics Subject Classification:16Y30.
https://nasplib.isofts.kiev.ua/handle/123456789/152191
In this paper the metacyclic Miller-Moreno p-groups which appear as the additive groups of finite local nearrings are classified.
en
Інститут прикладної математики і механіки НАН України
Algebra and Discrete Mathematics
Finite local nearrings on metacyclic Miller-Moreno p-groups
Article
published earlier
spellingShingle Finite local nearrings on metacyclic Miller-Moreno p-groups
Raievska, I.Yu.
Sysak, Ya.P.
title Finite local nearrings on metacyclic Miller-Moreno p-groups
title_full Finite local nearrings on metacyclic Miller-Moreno p-groups
title_fullStr Finite local nearrings on metacyclic Miller-Moreno p-groups
title_full_unstemmed Finite local nearrings on metacyclic Miller-Moreno p-groups
title_short Finite local nearrings on metacyclic Miller-Moreno p-groups
title_sort finite local nearrings on metacyclic miller-moreno p-groups
url https://nasplib.isofts.kiev.ua/handle/123456789/152191
work_keys_str_mv AT raievskaiyu finitelocalnearringsonmetacyclicmillermorenopgroups
AT sysakyap finitelocalnearringsonmetacyclicmillermorenopgroups