WorksMathematics
CRCs in four-dimensional lattice graphs
Abstract
We obtain series of new completely regular codes in Cayley graphs of the groups with the Lee metric. Combined with binary linear programming methods, the constructions lead to enumeration of all completely regular codes in such graphs for the groups , up to parameter matrices.
Introduction
The undirected Cayley graph of an abelian group with a generating set such that , , is the graph with vertex set whose edges are the pairs , , . The graphs considered in this paper are Cayley graphs of abelian groups. By we denote the word with the one in the th position and all other coordinates equal to zero. Below we define the classes of graphs used in the paper.
is the Hamming graph;
and are the lattices with the Lee metric (-metric);
and are the lattices with the -metric.
For the graph is regular of degree , and the graph is regular of degree ; is isomorphic to (the cycle is the square ), , . For even the graph is bipartite: the parts are formed by the words with even and with odd Lee weight.
The notion of a completely regular code arose in the 1970s in connection with the solution of the problem of the existence of perfect codes with arbitrary covering radius in the Hamming metric over field alphabets [19, 20, 17]. Perfect codes in the Lee metric have been studied since the seminal paper of Golomb and Welch [10]. One of the classical problems for completely regular codes is the classification of their parameters in a given family of graphs. In particular, classifications of the parameters of all completely regular codes (CRCs) in the Hamming graphs were obtained [14 Section 7] for (in , respectively) and in the Hamming graphs for .
For the Lee metric graphs, the problem of classifying CRCs has been considered in several papers. In the graph of the infinite square lattice , the case of covering radius 1 was studied in [4], and the case of covering radius 2 by Puzynina in [13]. For covering radius 1, several restrictions on the parameter matrix were obtained in [7]. A complete classification of CRCs of arbitrary covering radius in was obtained in [3]. Thus, the first open cases are the lattices and .
In [12], CRCs in whose first two diagonal elements of the parameter matrix are equal to zero were studied. In particular, all such codes in the graph were enumerated. We also mention an important paper [2], in which, for CRCs in the graphs and , the monotonicity of the superdiagonal and subdiagonal elements of the parameter matrix was established, as well as the upper bound on the covering radius.
The main definitions concerning completely regular codes and the basic constructions are given in § 2. In Section 3 we propose new series of CRCs with covering radius 2 in the graphs (two constructions) and for any , and in the graph for even if there exists a perfect code in . Mostly, the constructions are based on the composite nature of the alphabet size 6 and combine binary or ternary perfect codes with special symbol functions or with classical methods such as the Solov’eva construction [16]. Moreover, the new codes are not obtained from binary and ternary codes of small length in the Hamming metric by an inverse homomorphism, which follows from the enumeration of the parameters of these CRCs [14 Section 7]. In the final Section 4 we list the parameter matrices of all CRCs in , , , obtained using binary linear programming; each matrix is confirmed by a construction of a CRC.
Basic definitions and results
Perfect colorings and codes in graphs
Perfect colorings and CRCs A partition of the vertex set of a graph is called a perfect coloring with parameter matrix if every vertex of has exactly neighbors in . For a code (a nonempty set of vertices) we put as the set of vertices at distance from in the graph. The number is the covering radius of the code . A code is called completely regular (briefly, a CRC) if the partition is a perfect coloring. The parameter matrix of a CRC is tridiagonal (i.e. for ) and if the graph is -regular, it is recovered from the intersection array
The intersection array is mostly used for the compact notation of CRCs in the tables of Section 4 and in the rest of the text CRCs are represented by their parameter matrices . A classic example of CRC is a one-vertex code in a distance-regular graph and its intersection array coincides with the intersection array of the graph. For a one-vertex code in the Hamming graph we have , , , .
A perfect code (one-error-correcting perfect code) in a -regular graph is a CRC with intersection array , i.e. with the matrix
thus, for the singleton in from the example above is a perfect code with the matrix
Such codes exist in for all [10].
More generally, for a -fold perfect code in a -regular graph is a CRC with parameter matrix
i.e. a code meeting every ball of radius 1 in exactly vertices; a 1-fold perfect code is a perfect code.
Linear codes. A code is called linear if it is a subgroup of the group . Every linear code is given by a parity-check matrix, i.e. a matrix over with columns such that . For Theorems 3–4, in the case when all input codes are linear, we briefly indicate the description of the resulting code in terms of the parity-check matrices of the input codes. We note that there are also nonlinear perfect codes in Lee metric [8], [11 Remarks 2, 4].
Coverings
A surjective map from the vertices of a graph onto the vertices of a graph is called a -fold covering if for every vertex of the map sends the neighbors of to neighbors of and every neighbor of has exactly preimages among the neighbors of (for this is the usual notion of a covering). For Cayley graphs, the homomorphisms described in Proposition 1 are -fold coverings.
Proposition 1. [5 Lemma 4.1] Let be a surjective homomorphism of a group to a group , and let and be such that , , , and every element of has exactly , , preimages in . If is a CRC in with parameter matrix , then is a CRC in with parameter matrix .
A homomorphism has the form for an matrix ; it is surjective, for example, if some minor of order of the matrix is coprime to . Let (or ) be the coordinatewise reduction modulo :
Consider several basic constructions for CRCs arising from group homomorphisms (see also [12]).
K1. Let be any integer, and be a CRC in with parameter matrix . Then is a CRC in (respectively, in ) with matrix (here , since in ).
K2. Let be any integer, , and be a CRC in with parameter matrix . Then is a CRC in (respectively, in ) with matrix (here ).
K3. Let be odd, , and let the rows of the matrix be all the words , , . Then maps bijectively onto . Consequently, if is a CRC in with matrix , then is a CRC in with matrix .
K4. Let , , and let the rows of the matrix be all the words of length over whose first nonzero symbol is . Then maps bijectively onto . Consequently, if is a CRC in with matrix , then is a CRC in with matrix .
Basic CRCs, addition and multiplication
We regard the following codes as the basic, simplest sources of CRCs.
A one-vertex code in a distance-regular graph, for example in the cycle or in a Hamming graph;
Perfect codes in the Hamming metric and in the Lee metric. In particular, a binary perfect code of length in exists if and only if , (necessity follows from the Hamming bound, sufficiency from the Hamming codes). If , then the code is perfect in [1].
Extended binary perfect codes. Appending a parity check to each codeword gives an extended perfect code of length and cardinality with minimum distance (for ). For it is known [14 Proposition 1.38] that this code is a CRC in with covering radius and parameter matrix
For example, for the extended Hamming code is a CRC in with array , and for the extended perfect code of length is a CRC in with array . By Construction K1, for the preimage is a CRC in (and in ) with matrix , i.e. with array ; for example, for we obtain a CRC in with array .
Note that recently, in [18], a criterion for the extension of CRC code to be CRC was obtained in terms of the parameter matrices in a wide family of bipartite graphs, including the Hamming graphs over binary alphabets and the Lee graphs over alphabets of even size.
A criterion for CRCs with covering radius 2
For even each part of the graph is a CRC with matrix
(the even-weight code);
Even-weight subcodes of perfect codes. Let be a perfect code in a bipartite -regular graph with parts that meets both parts. Then (see e.g., [14 Proposition 1.32]) is a CRC with parameter matrix
-fold perfect codes are CRCs with parameter matrix
In particular, if is a perfect code in a Cayley graph , , that is a subgroup of the group , and the group is abelian, then the union of any cosets of , , is a -fold perfect code. As indicated at the beginning of this section, if , then there exists a perfect code of the graph that is a subgroup of , and therefore for every , , there exists a -fold perfect code in .
The following two operations, see [9] provide new CRCs from existing ones. Let and be a CRC in with parameter matrix .
Addition (): is a CRC in with matrix , where is the identity matrix.
Multiplication (): the preimage of under the homomorphism
is a CRC in with matrix .
In the tables of Section 4, the notation means a CRC in obtained by addition from a CRC in , and the notation means a CRC obtained by multiplication by from a CRC in ; the operations can be combined, for example . In all cases is the graph in which the original code is constructed (the construction of the original code is indicated in the table for ; its intersection array does not change under addition, and is multiplied by under multiplication).
A criterion for CRCs with covering radius 2
All the new codes of Section 3 are obtained using the following simple criterion.
Proposition 2. Let be a -regular graph, and let , where and are nonempty, disjoint, and there are no edges between and . Suppose that every vertex of is adjacent to exactly , , vertices of , and every vertex of is adjacent to exactly , , vertices of and to exactly vertices of . Then the following holds.
We have , , and is a CRC with parameter matrix
The code is a CRC in with parameter matrix
Proof. 1. Since there are no edges between and , all neighbors of a vertex of in are in , and the remaining are in ; the same holds for . In particular . Since , every vertex of is adjacent to , so . The vertices of are not adjacent to and have neighbors in , hence . The rows of the matrix are obtained by direct counting.
Follows directly from the parameters established in the previous item. □
New constructions of CRCs over in the Lee metric
Each construction of Sections 3.1–3.3 gives a code in , to which K2 is then applied; therefore the resulting CRCs also exist in for all and in . The construction of Section 3.4 works for any even .
A concatenation construction from ternary punctured perfect codes
The general form of this construction resembles the Solov’eva concatenation construction for perfect and extended perfect codes [16]. We say that a triple of subsets of is admissible if are nonempty and pairwise disjoint, their union is an independent set of the graph , and every vertex outside is adjacent to exactly one vertex of each of . This is equivalent to being a CRC with matrix
each being a CRC with matrix
and . The same definition is used for subsets of and the graph .
Lemma 1. For and there exists an admissible triple in .
Proof. Let be the ternary perfect (for example, Hamming code) in . Put , , and . Using code terms, is the punctured Hamming code, and the sets are shortened codes. Thus is an admissible triple in , because each shortened perfect code , is a CRC with required parameter matrix [15].
The triple under is an admissible triple in , since maps the neighbors of each vertex bijectively onto the neighbors of its image. □
Theorem 1. Let be an admissible triple in , , . Then the code
is a CRC in with parameter matrix
Proof. Let . In the cycle each of the families and partitions , and for each the sets , and have the following properties: a vertex of (of ) has one neighbor in (in ) and none in (in ); a vertex of has one neighbor in and one in , see Figure 1 a, b. In other words, for any , , is the distance partition w.r.t. the CRC in with the parameter matrix

Figure 1. Two partitions of the graph (a) , and (b) , ; (c) the distance partition of with respect to the CRC with parameter matrix .
see Figure 1 c.
We now check the conditions of Proposition 2 for , , where . Since the sets are pairwise disjoint, for we have if and only if ; in particular .
Consider a vertex , , . Its neighbors of the form are not in , since by the independence of the set . By the remark at the beginning of the proof, among the neighbors exactly one lies in and the other is in , so not in . Hence has exactly one neighbor in , and it lies in . The same holds for ; in particular, there are no edges between and , and .
Let and for some . Then belongs to . The neighbors of the vertex of the form do not lie in (again by the independence of ), and of the neighbors one lies in and one in by the remark at the beginning of the proof. Let and . Then the neighbors are not in because is not in . Moreover, by definition of admissible triple , a vertex , , has exactly one neighbor for each , these are all its neighbors in , and the vertex is in (in ) if and only if (), which holds for exactly one . In both cases , and the statement follows from Proposition 2. □
Corollary 1. For all , CRCs with parameter matrix exist in the graphs and .
Proof. For the required code in is the edge , see Figure 1c. For we apply Theorem 1 to the admissible triple from Lemma 1: , . We use to obtain CRCs over and . □
CRCs in the metric
By , every CRC in with metric provides a CRC with the same parameter matrix in . Consider the code . We call the symbols 0, 1 lower and 3, 4 upper. Let () be the set of words of with an even (odd) number of lower symbols. A neighbor , , of a vertex in is obtained by changing each symbol by 0 or .
Theorem 2. The code is a CRC in with parameter matrix
Proof. Let be in . A symbol can be changed by or while staying in in exactly two ways: by leaving it unchanged or by replacing it with its “partner” in this set (that is, or ); in this case a lower symbol remains lower and an upper symbol remains upper. Therefore has exactly neighbors in , and all of them lie in the same one of the codes , as ; in particular, there are no edges between and .
Let , i.e. symbols of are equal to or . The symbol can be replaced by (lower) or (upper), the symbol by (lower) or (upper), and each of the lower and upper symbols in two ways preserving its type. Since , all resulting words are different from , i.e. they are neighbors of in , and, since the type of a symbol equal to or can be chosen freely, exactly half of them have an even number of lower symbols. Thus , , and the statement follows from Proposition 2. □
Corollary 2. For any and there exists a CRC in and in with the parameter matrix of Theorem 2.
Proof. By Theorem 2 and construction K4 (with , ), the preimage of is a CRC in with the same parameter matrix; then construction K2 from § 2.2 gives the CRCs in and .
Remark 1. The example of the code from Theorem 2 for in Figure 2 was proposed by S.V. Avgustinovich.

Figure 2. (a) The adjacency in the graph (b) The CRC in with parameter matrix .
Open problem (S.V. Avgustinovich). Obtain a classification of CRCs in the graphs (Cayley graphs of the group with the metric), .
CRCs from binary perfect codes
Let , , and let be a binary perfect code of length .
By K1, the code is a CRC in with matrix
From the parameter matrix of the perfect code in we have the following obvious fact:
(H) if , then the neighbors of in are exactly and for a unique .
We write the vertices of the graph in the form , , , and consider the homomorphism from onto :
Note that the homomorphism maps the graph onto a multigraph which is the complete graph with each edge of multiplicity .
Put
Theorem 3. For each the code is a CRC in with parameter matrix
Moreover, for all , CRCs with the same parameter matrix exist in and in .
Proof. Since is a single vertex of the multigraph (with each edge of multiplicity ), which has no loops, and since is a homomorphism from onto this multigraph, no pair of vertices of is adjacent, i.e. ; in particular, there are no edges between and . Let and . Denote
Here is the word of length with the one in the th position.
For each , exactly one of the neighbors lies in , namely , since and for , while the other neighbors have and . Thus has exactly neighbors in and exactly neighbors in . By the definition of we have
Recall that the code is an independent set in , since the parameter matrix of this CRC has 0 in the upper left corner. We have the following cases.
Case A. Let . Then , and by the independence of , for all . The vertex has exactly neighbors in and exactly one, namely , in .
Case B. Let . Then , and by (H) the neighbors of in are exactly and for a unique ; hence . Therefore and for . In both cases has exactly neighbors in and in . The conditions of Proposition 2 are satisfied with , , , , and the parameter matrix of the code is
We use K2 to obtain a CRC in and . □
Parity-check matrix. Let be the binary Hamming code with an parity-check matrix . The code has the parity-check matrix over , and . Since , the code is linear with parity-check matrix
where is the all-ones row.
``Lengthening'' of a doubled perfect code
Let be even and let be a perfect code in . We write the vertices of in the form , , , and put ; for let . Consider the codes
Thus is one of the parts of the bipartite graph , is the other part, and and differ by the condition or .
Theorem 4. If for even there exists a perfect code in , then the code in (1) is a CRC in with parameter matrix
Proof. The part is independent, so and there are no edges between and . Let be a vertex in the part , . All neighbors of lie in the part ; the neighbors and , , correspond to the word , and the neighbors correspond to the word . If , then by the perfectness of the code , and the neighbors of in are exactly . If , then has a unique neighbor (where ), and the neighbors of in are exactly and . In both cases , , and the statement follows from Proposition 2.
Parity-check matrix. If is a linear code with parity-check matrix over , then is linear with parity-check matrix
Remark 2. If , then a perfect code in exists (§ 2.3). According to Horak, AlBdaiwi and Milazzo [1], perfect codes in exist for all such that the product of all prime divisors of divides . In particular, such a code exists in (for example, the preimage under of the ternary Hamming code of length 4), so Theorem 4 gives a CRC in .
CRCs in the lattice graphs , ,
The enumeration of the parameter matrices of all CRCs in the graphs for , was carried out by solving an integer linear programming problem. The program was written in Python, and the problems were solved with the HiGHS solver.
Let be a CRC with covering radius in a graph with vertex set and adjacency matrix . Its characteristic matrix is the binary matrix with if and otherwise. It is easy to see [14 Proposition 1.4] that a binary matrix is the characteristic matrix of a CRC with a tridiagonal parameter matrix (with nonzero sub- and superdiagonal entries) if and only if
Thus a CRC with parameter matrix exists if and only if this system has a binary solution.
Note that the case of the graph with vertices is already borderline, since the number of variables is already very large. However, the graph has small degree and fairly large diameter, which has a positive effect on the speed of the solver. For instance, already in the Hamming graph with vertices and degree 20, the solver cannot cope with the problem of the existence of CRCs.
Before solving the problem for a candidate matrix with covering radius , we checked that satisfies the following necessary conditions.
Monotonicity of the off-diagonal entries. As proved in [2], the parameter matrix of every CRC in the lattice graphs and satisfies for and for , i.e. and .
Integrality of the cell sizes. The sizes of the cells of the distance partition of a graph of order are determined by the relations [14 Lemma 7.2]
and they must be integers.
Integrality of eigenvalues. By Lloyd theorem [14 Lemma 7.4], every eigenvalue of the parameter matrix of a CRC is an eigenvalue of the graph. The eigenvalues of the cycle are , , and the eigenvalues of are all sums of eigenvalues of . For example, the eigenvalues of are , and for the eigenvalues of are all integers from to ; hence for all eigenvalues of must be integers.
These conditions significantly reduce the set of possible parameter matrices. For , up to passing to the opposite code, there are 496 tridiagonal matrices with row sum 8 and positive off-diagonal entries fulfilling the monotonicity condition, 76 of them have integer cells, and only 41 fulfill all three conditions.
In all tables of this section, CRCs are given by their intersection arrays (see § 2), which uniquely determine the parameter matrices. If is a CRC with array , then the opposite code is a CRC with array , of each such pair of arrays, only one is listed. Notation in the column “construction”: pc is a perfect code; k pc is a -fold perfect code obtained as the union of cosets of a linear perfect code (in a -regular graph it is a CRC with array ; 1 pc is pc itself); sing is a one-vertex code; even is the even-weight code; ext pc is an extended perfect code; even pc is the even-weight subcode of a perfect code; rep is the ternary repetition code of length 3, a well-known CRC in with intersection array , see Figure 3.

Figure 3. Hamming graph and ternary code , which is CRC with parameter matrix .
A graph written after a code name, as in “sing ”, means that the code is constructed in that graph and lifted to by one of the coverings K1–K3 of § 2.2; if no graph is given, the code is constructed in itself. The entries and denote addition and multiplication (§ 2.3) applied to a CRC from the table for . For codes obtained by the constructions of Section 3, the corresponding theorem or corollary is given together with its parameters. New constructions are marked with an asterisk.
Note that all CRCs found in for and have small covering radii, namely at most . For every , a CRC with covering radius exists in : take a one-vertex code in and add inessential coordinates (addition ). By [2], the covering radius of a CRC in is at most , and this bound is attained, for example, when is divisible by 4.
Open problem. Is it true that for not divisible by 4 and , the covering radius of every CRC in is at most ?
,
Since , and , these cases reduce to Hamming graphs; see the survey [14 Section 7].
, .
Table 5.
. New constructions are marked with an asterisk.
Table 8.
Table 9.
. New constructions are marked with an asterisk.
Table 10.
| array | construction | array | construction |
| sing | |||
| sing | |||
| even pc |
Table 11.
, .
| array | construction |
| sing |
Table 12.
| array | construction |
| pc | |
Table 13.
| array | construction |
Table 14.
| array | construction |
Table 15.
| array | construction |
Table 16.
| array | construction |
Table 17.
| array | construction |
| K3 for a singleton in | |
| K3 for a cycle of length 4 in (Figure 4) | |
Table 18.

Figure 4. (a) The adjacency in the Hamming graph and (b) CRC with parameter matrix
,
| array | construction |
| sing |
Table 19.
| array | construction |
Table 20.
| array | construction |
| pc | |
| 2 pc | |
| 3 pc |
Table 21.
| array | construction |
Table 22.
| array | construction |
Table 23.
| array | construction |
Table 24.
Acknowledgments
The author is grateful to S. V. Avgustinovich for discussions and pointing out the connection between construction of the code of length 3 given in Section 3.2, to A. Yu. Vasil’eva for fruitful discussion of the material of the paper, to V. R. Danilko for kindly providing multicore computing resources for solving the binary linear programming problems and A. A. Panin for recommending the HiGHS package for Python.
The results of the paper were obtained in the pre-AI era. At the final part of the paper production the proofs were cross-verified in the Lean language using the AI Aristotle agent: aristotle.harmonic.fun, which was also used to draw all the figures and reformat raw data into tables.
Ivan Yurevich Mogilnykh
Sobolev Institute of Mathematics,
pr. Koptyuga, 4,
630090, Novosibirsk, Russia
Email address: [email protected]
References
- [1]B. AlBdaiwi, P. Horak, L. Milazzo, Enumerating and decoding perfect linear Lee codes, Des. Codes Cryptogr., 52:2 (2009), 155–162. DOI: 10.1007/s10623-009-9273-3.DOI
- [2]S.V. Avgustinovich, A.Yu. Vasil’eva, Completely regular codes in the n-dimensional rect-angular grid, Siberian Electronic Mathematical Reports, 19:2 (2022), 861–869. DOI: 10.33048/semi.2022.19.072.
- [3]S.V. Avgustinovich, A.Yu. Vasil’eva, I.V. Sergeeva, Distance regular colorings of an infinite rectangular grid, J. Appl. Ind. Math., 6:3 (2012), 280–285. DOI: 10.1134/S1990478912030027.DOI
- [4]M. Axenovich, On multiple coverings of the infinite rectangular grid with balls of constant radius, Discrete Math., 268:1–3 (2003), 31–48. DOI: 10.1016/S0012-365X(02)00744-6.DOI
- [5]E.A. Bespalov, D.S. Krotov, A.A. Mati ushev, A.A. Taranenko, K.V. Vorob’ev, Perfect 2-colorings of Hamming graphs, J. Combin. Des., 29:6 (2021), 367–396. DOI: 10.1002/jcd.21771.DOI
- [6]J. Borges, J. Rifa, V.A. Zinoviev, On completely regular codes, Probl. Inf. Transm., 55:1 (2019), 1–45. DOI: 10.1134/S0032946019010010.DOI
- [7]P. Dorbec, S. Gravier, I. Honkala, M. Mollard, Weighted codes in Lee metrics, Des. Codes Cryptogr., 52:2 (2009), 209–218. DOI: 10.1007/s10623-009-9277-z.DOI
- [8]T. Etzion, Product constructions for perfect Lee codes, IEEE Trans. Inform. Theory, 57:11 (2011), 7473–7481. DOI: 10.1109/TIT.2011.2161133.DOI
- [9]D.G. Fon-Der-Flaass, Perfect 2-colorings of a hypercube, Sib. Math. J., 48:4 (2007), 740–745. DOI: 10.1007/s11202-007-0075-4.
- [10]S.W. Golomb, L.R. Welch, Perfect codes in the Lee metric and the packing of polyominoes, SIAM J. Appl. Math., 18:2 (1970), 302–317. DOI: 10.1137/0118025.DOI
- [11]I. Yu. Mogilnykh, F. I. Solov’eva, Constructions and invariants of optimal codes in the Lee metric, Probl. Inf. Transm., 59:2 (2023), 71–85. doi:10.1134/S0032946023020011.
- [12]I.Yu. Mogilnykh, A.Yu. Vasil’eva, On null completely regular codes in Manhattan metric, Siberian Electronic Mathematical Reports, 22:2 (2025), 1311–1333. DOI: 10.33048/semi.2025.22.079.
- [13]S.A. Puzynina, Perfect colorings of graph G(Z²) into three colors, Diskretn. Anal. Issled. Oper., Ser. 1, 12:2 (2005), 37–54 (in Russian).
- [14]M. Shi, P. Solé (eds.), Completely Regular Codes in Distance Regular Graphs, Chapman and Hall/CRC, Boca Raton, 2025. DOI: 10.1201/9781003393931.DOI
- [15]M. Shi, R. Wu, D. S. Krotov, On q-ary shortened-1-perfect-like codes, IEEE Trans. Inf. Theory, 68:11 (2022), 7100–7106. doi:10.1109/TIT.2022.3187004.DOI
- [16]F.I. Solov’eva, On binary nongroup codes, in: Methods of Discrete Analysis in the Study of Boolean Functions and Graphs, vol. 37, Institute of Mathematics SB AS USSR, Novosibirsk, 1981, 65–76 (in Russian).
- [17]A. Tietäväinen, On the nonexistence of perfect codes over finite fields, SIAM J. Appl. Math., 24:1 (1973), 88–96. doi:10.1137/0124010.DOI
- [18]A.Yu. Vasil’eva, I.Yu. Mogilnykh, On K₂-extensions of completely regular codes in bipartite graphs, Probl. Inf. Transm., submitted.
- [19]V. A. Zinoviev, V. K. Leontiev, On perfect codes, Problems Inform. Transmission, 8:1 (1972), 17–24. Math-Net.Ru: ppi772.
- [20]V. A. Zinoviev, V. K. Leontiev, The nonexistence of perfect codes over Galois fields, Probl. Control Inform. Theory, 2:2 (1973), 123–132.