Compactness of the gain parts of the linearized Boltzmann operator with weakly cutoff kernels (Q989409)

It is considered the linear operator \( {L} \) defined by \[ Lf= \iint_{\mathbb{S}^{D-1}\times \mathbb{R}^D}(f+f_1-f'-f'_1) b(\omega,v_1-v)\,d\omega M_1\,dv_1. \] Here \(\omega\in \mathbb{S}^{D-1}\) is a unit vector, \(v\in \mathbb{R}^D,\) \(f_1,\) \(f',\) \(f'_1\) and \(f\) evaluated at the \(v_1,\) \(v',\) \(v'_1\) and \(v\) respectively, the kernel \(b(\omega,v_1-v)\) has the factored form \[ b(\omega,v_1-v)=\hat{b}(\omega\cdot n)\mid v_1-v\mid^\beta,\qquad n= \frac{v_1-v}{|v_1-v|}, \] where \(\hat{b}\) and \(\beta\) satisfy \[ \hat{b}(\omega\cdot n)\in \mathbf{L}^1(d\omega),\quad -{D}<\beta, \] \(M\) denote the Maxwellian given by \[ M(v)=\frac{1}{(2\pi)^{\frac{D}{2}}}\exp \bigg(-\frac{1}{2}|v|^2\bigg). \] The authors decompose the operator \(L\) as \[ Lf =a(v)(I+{K}_1-{K}_2-{K}_3)f, \] where the coefficient \(a(v)\) is given by \[ a(v)=\iint_{\mathbb{S}^{D-1}\times \mathbb{R}^D}b(\omega,v_1-v)\,d\omega M_1\,dv_1, \] while the operator \(K_1\) and the operators \(K_2\) and \(K_3\) are given by \[ \begin{aligned} K_1f&= \frac{1}{a(v)} \iint_{\mathbb{S}^{D-1}\times \mathbb{R}^D} f_1 b(\omega,v_1-v)\,d\omega M_1\,dv_1,\\ K_2f&=\frac{1}{a(v)} \iint_{\mathbb{S}^{D-1}\times \mathbb{R}^D} f'b(\omega,v_1-v)\,d\omega M_1\,dv_1,\\ K_3f&=\frac{1}{a(v)} \iint_{\mathbb{S}^{D-1}\times \mathbb{R}^D}f_1' b(\omega,v_1-v)\,d\omega M_1\,dv_1. \end{aligned} \] From the authors the main result of this paper is the following: For every \(p\in(1,\infty)\) and every \(j=1,2,3\) the operator \[ K_j:\mathbf{L}^p(aMdv)\rightarrow \mathbf{L}^p(aMdv) \] is compact.