It is proved that if positive definite matrix functions (i.e. matrix spectral densities) S n , n=1,2,… , are convergent in the L 1-norm, ||Sn-S||L1® 0|S_{n}-S|_{L_{1}}to 0, and ò02plogdetSn(eiq) dq®ò02plogdetS(eiq) dqint_{0}^{2pi}log mathop{mathrm{det}}S_{n}(e^{itheta}),dthetatoint_{0}^{2pi}log mathop{mathrm{det}}S(e^{itheta}),dtheta, then the corresponding (canonical) spectral factors are convergent in L 2, ||S+n-S+||L2® 0|S^{+}_{n}-S^{+}|_{L_{2}}to 0. The formulated logarithmic condition is easily seen to be necessary for the latter convergence to take place.
展开▼
机译:证明如果正定矩阵函数(即矩阵谱密度)S n sub>,n = 1,2,…在L 1 sub>范数上收敛,| | | S n sub> -S || L 1 sub> sub>®0 | S_ {n} -S | _ {L_ {1}}至0,和« 0 sub> 2p sup> logdetS n sub>(e iq sup>)dq®ò 0 sub> < sup> 2p sup> logdetS(e iq sup>)dqint_ {0} ^ {2pi} log mathop {mathrm {det}} S_ {n}(e ^ {itheta}),dthetatoint_ {0 } ^ {2pi} log mathop {mathrm {det}} S(e ^ {itheta}),dtheta,则相应的(规范)谱因子在L 2 sub>中收敛,|| S + sup> n sub> -S + sup> || L 2 sub> sub>®0 | S ^ {+} _ {n} -S ^ {+} | _ {L_ {2}}至0。很容易看出,制定的对数条件对于后者的收敛是必要的。
展开▼