Module theory over Leavitt path algebras and \(K\)-theory. (Q2267426)

For a field \(k\) and an integer \(n\geq 2\), the Leavitt algebra \(L(1,n)\) of type \((1,n)\) is the algebra with generators \(x_i,y_j\), \(1\leq i,j\leq n\), with defining relations given by \[ (x_1,\dots,x_n)(y_1,\dots,y_n)^t=1,\quad (y_1,\dots,y_n)^t(x_1,\dots,x_n)=I_n, \] where \(I_n\) is the \(n\times n\) identity matrix. These algebras were studied by \textit{W. G. Leavitt} [in Proc. Am. Math. Soc. 8, 322-328 (1957; Zbl 0073.02402) and Trans. Am. Math. Soc. 103, 113-130 (1962; Zbl 0112.02701)]. They provide us with universal examples of algebras without the invariant basis number property: observe that right multiplication by the row \((x_1,\dots,x_n)\) gives an isomorphism from the free left \(L(1,n)\)-module of rank one onto the free left \(L(1,n)\)-module of rank \(n\). They are algebraic analogues of the Cuntz algebras \(\mathcal O_n\), which were introduced independently by \textit{J. Cuntz} [in Commun. Math. Phys. 57, 173-185 (1977; Zbl 0399.46045)]. \textit{P. Ara} analyzed [in J. Pure Appl. Algebra 191, No. 1-2, 1-21 (2004; Zbl 1072.16012)] the structure of the finitely presented modules over \(L(1,n)\) in connection with the structure of certain classes of finitely presented modules over the free algebras \(k\langle x_1,\dots,x_n\rangle \) and \(k\langle y_1,\dots,y_n\rangle\). Both free algebras embed in \(L(1,n)\), and the Abelian category \(\mathcal S\) of finitely presented left \(L(1,n)\)-modules of finite length is equivalent to a quotient category of the Abelian category of finite-dimensional \(k\langle y_1,\dots,y_n\rangle\)-modules by a certain Serre subcategory. Let \(\Sigma\) be the class of all the square matrices over \(k\langle x_1,\dots,x_n\rangle\) that are sent to an invertible matrix by the augmentation map. Then \(\mathcal S\) is identified with the category of finitely presented \(\Sigma\)-torsion modules by \textit{P. Ara}, [loc. cit., Zbl 1072.16012, Theorem 6.2], and this is used to give a formula for \(K_1(Q_n)\), where \(Q_n=L(1,n)\Sigma^{-1}\) is the universal localization of \(L(1,n)\) with respect to \(\Sigma\), which was shown by \textit{P. Ara, K. R. Goodearl} and \textit{E. Pardo}, [K-theory 26, No. 1, 69-100 (2002; Zbl 1012.16013)], to be a simple von Neumann regular ring. The objective of this article is to extend these results to the much wider context of path algebras. The main principle dealing with this problem is the idea that free algebras are prototypical examples of path algebras, and many results on free algebras should admit suitable generalizations to this background. The authors study the structure of the finitely presented modules of finite length over the Leavitt path algebra \(L_k(E)\) and show its close relationship with the finite-dimensional representations of the inverse quiver \(\overline E\) of \(E\), as well as with the class of finitely generated \(P_k(E)\)-modules \(M\) such that \(\text{Tor}_q^{P_k(E)}(k^{|E^0|},M)=0\) for all \(q\), where \(P_k(E)\) is the usual path algebra of \(E\). By using these results they compute the higher \(K\)-theory of the von Neumann regular algebra \(Q_k(E)=L_k(E)\Sigma^{-1}\), where \(\Sigma\) is the set of all square matrices over \(P_k(E)\) which are sent to invertible matrices by the augmentation map \(\varepsilon\colon P_k(E)\to k^{|E^0|}\).