Generalizations of the Borsuk Theorem
WY DING
DOI: https://doi.org/10.1016/0022-247x(85)90316-6
2017-01-01
Abstract:For any subgroup H of G, we will denote by V, the fixed point subspace of H-action, i.e., I/,= (xEIR~) T,x=x’dh~H}. For xrz(WN, let O(x)= ( T,x 1 g E G} be the G-orbit of 3~. If G is finite, we let l(x) denote the order of O(x). Suppose that !2 is an open bounded G-invariant set in IWN andf: S + [WN is a continuous G-equivariant map. If 0 #f(X?), the Brouwer degree deg(f, 52,O) is well defined. It is known that we can obtain important informations on deg(f, Q, 0) if the G-action on aBN is of some special types. For example, in the case that G = Z, = (0, 1 }, T, = I and T, = -Z, where I is the identity transformation on IWN, the Borsuk theorem asserts that deg(f, 52, 0) is an odd number provided 0 E Q. Recently, efforts have been made to generalize the Borsuk theorem for the study of differential equations with various symmetries. Under the assumption that VG = {0), Marzantowicz [4] obtains the following results. (In the sequel, we always assume that .C2 is an open bounded G-invariant subset of RN.)