Undistorted fillings in subsets of metric spaces (Q6042802)

The authors relate Lipschitz connectedness in spaces of finite Nagata dimension with Euclidean isoperimetric inequalities and coning inequalities. A metric space \(X\) is \(c\)-quasiconvex, \(c\geq1\), if all \(x,x'\in X\) can be connected by a curve \(\gamma:[0,1]\to X\) such that \(\ell(\gamma)\leq cd(x,x')\). \(X\) is Lipschitz \(k\)-connected if Lipschitz maps from \(k\)-spheres to \(X\) can be extended to Lipschitz maps from \((k+1)\)-balls. A covering \(\mathcal B=(B_i)_{i\in I}\) of \(X\) is \(D\)-bounded, \(D\geq0\), if every set \(B\in\mathcal B\) has diameter less than or equal to \(D\). A covering \(\mathcal B\) has \(s\)-multiplicity at most \(n\in\mathbb N\), \(s\geq0\), if every subset \(B\subset X\) with \(\text{diam}(B)\leq s\) meets at most \(n\) members of \(\mathcal B\). \(X\) has Nagata dimension \(\leq n\) with constant \(c\) if for every \(s>0\), \(X\) admits a \(cs\)-bounded covering \(\mathcal B_s\) having \(s\)-multiplicity at most \(n+1\). Denote by \(\mathbf I_k(X)\) the abelian group of \(k\)-dimensional metrical integral currents in \(X\). The filling volume in \(X\) for \(T\in\mathbf I_k(X)\) is defined by \[\text{Fillvol}_X(T)=\inf\{\mathbf M(S):S\in\mathbf I_{k+1}(X)\;\text{and}\;\partial S=T\},\] where \(\mathbf M(S)\) is the mass of \(S\). For \(k\geq0\), \(X\) has \((\text{EI}_k)\) if there is \(D>0\) such that for \(0\leq m\leq k\) and every cycle \(T\in\mathbf I_m(X)\), \(\text{Fillvol}_X(T)\leq D\mathbf M(T)^{(m+1)/m}\). Also, \(X\) has \((\text{LC}_k)\) if there is \(c\geq1\) such that for \(0\leq m\leq k\) each \(L\)-Lipschitz map from the Euclidean \(m\)-sphere to \(X\) extends to a \(cL\)-Lipschitz map defined on the \((m+1)\)-ball. A closed subset \(X\subset Y\) of a complete metric space \(Y\) is undistorted in \(Y\) up to dimension \(k+1\geq1\) if there is \(C>0\) such that for all cycles \(T\in\mathbf I_m(X)\), \(0\leq m\leq k\), \(\text{Fillvol}_X(T)\leq C\text{Fillvol}_Y(T)\). One of the main results is given in the following theorem. Theorem. Let \(Y\) be a complete metric space and \(X\subset Y\) a closed quasiconvex subset of finite Nagata dimension. If \(X\) has \((\text{LC}_k)\) or \((\text{EL}_k)\) for some \(k\geq0\), then \(X\) is undistorted in \(Y\) up to dimension \(k+1\). The distortion constant only depends on the data of \(X\).