Regularity of subelliptic \(p\)-harmonic systems with subcritical growth in Carnot group (Q2256007)

The article focuses on the properties of weak solutions of a subelliptic system on a Sobolev space defined on a domain, \(\Omega\), in the Carnot group \(\mathcal{G}\). Precisely, the authors consider the space \[ HW^{k,p}(\Omega)=\{u\in L_p(\Omega)\text{ s.t. } X_{i_1}\dots X_{i_k}u\in L_p(\Omega)\text{ for all }1\leq i_1,\dots,i_k\leq m\}, \] where each \(X=(X_i)_{1\leq i\leq m}\) is a family of vector fields on \(\Omega\) satisfying the Hörmander finite-rank condition. Then, they consider the following class of inhomogenuous subelliptic \(p\)-harmonic systems: \[ \sum_{i=1}^mX_i^*(|Xu|^{p-2}X_iu)=F(x,u,Xu),\text{ and }|F(x,u,Xu)|\leq L|Xu|^q\text{ a.e } x\in\Omega,\tag{\(*\)} \] such that each \(X^*_i\) stands for the adjoint of \(X_i\), \(L\) is a positive constant, and \(q\in [0,p).\) Next, for \(\alpha\in (0,1]\), they define \(\big(\Gamma^{0,\alpha}(\Omega),\|\cdot\|_{\Gamma^{0,\alpha}}(\Omega)\big)\), the Banach subspace of bounded function as \[ \Gamma^{0,\alpha}(\Omega):=\left\{u\in L_\infty(\Omega)\text{ such that }[u]_\alpha:=\displaystyle\sup_{x\neq y}\frac{|u(x)-u(y)|}{d^\alpha_X(x,y)}<\infty\text{ for } |I|\leq k\right\} \] and \[ \Gamma^{k,\alpha}(\Omega):=\{u\in \Gamma^{0,\alpha}(\Omega)\text{ and } X_Iu\in \Gamma^{0,\alpha}(\Omega)\text{ for } |I|\leq k\}, \] where \(X_I=X_{i_1}\dots X_{i_k}\) and \(|I|\) is the length of the multi-index \((i_1,i_2,\dots,i_m)\), and \(d_X\) stands for the Carnot-Carathéodory metric related to \(X\). Then, the authors show two meaningful theorems on giving sufficient conditions for a weak solution of \((*)\) to be in \(\Gamma^{k,\alpha}(\Omega)\). More precisely, for \(p\) near to two, they state that if there is a weak bounded solution of \((*)\) belonging to \(HW^{1,p}(\Omega)\), then it appertains (locally) to \(\Gamma^{0,\alpha}(\Omega)\) and \(\Gamma^{1,\mu}(\Omega)\), respectively, where \(\mu\) is a constant depending on the homogeneous dimension of \(\mathcal{G}, p, q, L\) (Theorems 1 and 2). The proofs are essentially based on a few significant inequalities, see Lemmas 6,7,8 and 9.