Some algebraic aspects of quadratic forms over fields of characteristic two (Q5949803)

This is a well written survey article on some of the author's important and more recent results on quadratic and bilinear forms in characteristic \(2\) which he obtained in collaboration with R.~Aravire. Let \(F\) be a field of characteristic \(2\) and \(\Omega_F^n\) be the \(F\)-space of absolute differential \(n\)-forms. There is a homomorphism of additive groups \(\wp : \Omega_F^n\to \Omega_F^n/d\Omega_F^{n-1}\) which acts on generators by mapping \(xdx_{1}/x_{1}\wedge\cdots \wedge dx_{n}/x_{n}\) to \((x^2+x)dx_{1}/x_{1}\wedge\cdots \wedge dx_{n}/x_{n}\bmod d\Omega_F^{n-1}\). The kernel of this map is denoted by \(\nu_F(n)\), its cokernel by \(H^{n+1}(F)\). Let \(WF\) denote the Witt ring of symmetric bilinear forms over \(F\), and \(I^nF\) the \(n\)-th power of the ideal \(IF\) of classes of even-dimensional bilinear forms in \(WF\). Let furthermore \(W_qF\) denote the Witt group of nondegenerate quadratic forms which is also a \(WF\)-module in a natural way, and let \(I^nW_qF=I^nF\cdot W_qF\). A famous theorem by Kato states that there are natural isomorphisms between \(\nu_F(n)\), \(I^nF/I^{n+1}F\), and \(k_nF\) (Milnor's \(K_nF\) modulo \(2\)), and also between \(H^{n+1}F\) and \(I^nW_qF/I^{n+1}W_qF\) (the analogue of the Milnor conjecture in characteristic \(2\)). The author describes the construction of the objects and maps involved in these isomorphisms, and also how the isomorphism between \(H^{n+1}F\) and \(I^nW_qF/I^{n+1}W_qF\) can be extended to a triangle of isomorphisms by including a group \(h_{n+1}F\) which is a certain quotient of \(k_nF\otimes F/\wp F\). This latter result is due to \textit{R. Aravire} and the author [Commun. Algebra 20, No. 4, 1087--1107 (1992; Zbl 0747.11021)]. The survey then focusses on some of their later work concerning the determination of the kernel for the restriction map \(H^nF\to H^nK\) for certain types of field extensions \(K/F\), most notably in the case where \(K\) is the function field of a bilinear Pfister form or a quadratic Pfister form over \(F\), and certain consequences derived from these computations, for example a proof of Knebusch's degree conjecture in characteristic \(2\). These results have been published in a series of papers by \textit{R. Aravire} and the author [Commun. Algebra 27, No. 7, 3473--3477 (1999; Zbl 0933.11018), J. Algebra 259, No. 2, 361--414 (2003; Zbl 1168.11308), Commun. Algebra 31, No. 1, 463--473 (2003; Zbl 1022.11015)].