A Lie Theoretic Approach to Thompson's Theorems on Singular Values-Diagonal Elements and Some Related Results

Thompson's famous theorems on singular values–diagonalelements of the orbit of an nxn matrix A under the action (1)U(n) U(n) where A is complex, (2) SO(n) SO(n), where A isreal, (3) O(n) O(n) where A is real are fully examined. Coupledwith Kostant's result, the real semi-simple Lie algebra so_n,n yields (2) and hence (3) and the sufficient part (the hardpart) of (1). In other words, the curious subtracted term(s)are well explained. Although the diagonal elements correspondingto (1) do not form a convex set in Cⁿ, the projection of thediagonal elements into Rⁿ (or iRⁿ) is convex and the characterizationof the projection is related to weak majorization. An elementaryproof is given for this hidden convexity result. Equivalentstatements in terms of the Hadamard product are also given.The real simple Lie algebra su_{n, n} shows that such a convexityresult fits into the framework of Kostant's result. Convexityproperties and torus relations are studied. Thompson's resultson the convex hull of matrices (complex or real) with prescribedsingular values, as well as Hermitian matrices (real symmetricmatrices) with prescribed eigenvalues, are generalized in thecontext of Lie theory. Also considered are the real simple Liealgebras so_{p, q} and so_{p, q}, p < q, which yield the rectangularcases. It is proved that the real part and the imaginary partof the diagonal elements of complex symmetric matrices withprescribed singular values are identical to a convex set inRⁿ and the characterization is related to weak majorization.The convex hull of complex symmetric matrices and the convexhull of complex skew symmetric matrices with prescribed singularvalues are given. Some questions are asked.     

Convex Bodies with Homothetic Sections

We prove that if K is a convex body in Eⁿ⁺¹, n2, and p₀ is apoint of K with the property that all n-sections of K throughp₀ are homothetic, then K is a Euclidean ball.     

Fast Solution of Vandermonde-Like Systems Involving Orthogonal Polynomials

Consider the (n + 1) ? (n + 1) Vandermonde-like matrix P=[p_i-1(_j-1)],where the polynomials p_o(x), ..., p_n(x) satisfy a three-termrecurrence relation. We develop algorithms for solving the primaland dual systems, Px = b and P^Ta = f respectively, in O(n²)arithmetic operations and O(n) elements of storage. These algorithmsgeneralize those of Bj?rck & Pereyra which apply to themonomial case p_i(x). When the p_i(x) are the Chebyshev polynomials,the algorithms are shown to be numerically unstable. However,it is found empirically that the addition of just one step ofiterative refinement is, in single precision, enough to makethe algorithms numerically stable.     

Animals, Trees and Renewal Sequences: Corrigendum

IN SECTION 3 of the above we omitted to mention aperiodicity.The period p of the pseudo renewal sequence {a_n: n > 0} isgiven by p = g.c.d. {n > 1: a_n > 0}. We are only concernedwith aperiodic renewal sequences (i.e. where p = 1). As it standsTheorem 3.1 is incorrect and should be restated as: THEOREM 3.1 If a = (a_n: n = 0,1,...) is an aperiodic pseudo-renewalsequence its limit a satisfies g_na^–n > 1 where a^–1 is to be interpreted as; if a = 0.     

The Characterization of the Regularity of the Jacobian Determinant in the Framework of Bessel Potential Spaces on Domains

Let 2 m n. The paper gives necessary and sufficient conditionson the parameters s1, s2, ..., sm, p1, p2, ..., pm such thatthe Jacobian determinant extends to a bounded operator fromHs1p1 x Hs2p2 x ... x Hsmpm into S'. Here all spaces are definedon Rn or on domains Rn. In almost all cases the regularity ofthe Jacobian determinant is calculated exactly.     

An Euler-Type Version of the Local Steiner Formula for Convex Bodies

The purpose of this note is to establish a new version of thelocal Steiner formula and to give an application to convex bodiesof constant width. This variant of the Steiner formula generalizesresults of Hann [3] and Hug [6], who use much less elementarytechniques than the methods of this paper. In fact, Hann askedfor a simpler proof of these results [4, Problem 2, p. 900].We remark that our formula can be considered as a Euclideananalogue of a spherical result proved in [2, p. 46], and thatour method can also be applied in hyperbolic space. For some remarks on related formulas in certain two-dimensionalMinkowski spaces, see Hann [5, p. 363]. For further information about the notions used below, we referto Schneider's book [9]. Let Kⁿ be the set of all convex bodiesin Euclidean space Rⁿ, that is, the set of all compact, convex,non-empty subsets of Rⁿ. Let S^n–1 be the unit sphere.For KKⁿ, let NorK be the set of all support elements of K, thatis, the pairs (x, u)RⁿxS^n–1 such that x is a boundarypoint of K and u is an outer unit normal vector of K at thepoint x. The support measures (or generalized curvature measures)of K, denoted by ₀(K.), ..., _n–1(K.), are the unique Borelmeasures on RⁿxS^n–1 that are concentrated on NorK andsatisfy [formula] for all integrable functions f:RⁿR; here denotes the Lebesguemeasure on Rⁿ. Equation (1), which is a consequence and a slightgeneralization of Theorem 4.2.1 in Schneider [9], is calledthe local Steiner formula. Our main result is the following.1991 Mathematics Subject Classification 52A20, 52A38, 52A55.     

On the Volume Ratio of Two Convex Bodies

Let K and L be two convex bodies in Rⁿ. The volume ratio vr(K,L) of K and L is defined by vr(K, L = inf(|K|/|T(L)|)^1/n, wherethe infimum is over all affine transformations T of Rⁿ for whichT(L) K. It is shown in this paper that vr(K, L) , where c > 0 is an absolute constant. This isoptimal up to the logarithmic term. 2000 Mathematics SubjectClassification 52A40, 46B07 (primary); 52A21, 52A20 (secondary).     

Lie Powers of Modules for Groups of Prime Order

Let L(V) be the free Lie algebra on a finite-dimensional vectorspace V over a field K, with homogeneous components Lⁿ(V) forn 1. If G is a group and V is a KG-module, the action of Gextends naturally to L(V), and the Lⁿ(V) become finite-dimensionalKG-modules, called the Lie powers of V. In the decompositionproblem, the aim is to identify the isomorphism types of indecomposableKG-modules, with their multiplicities, in unrefinable directdecompositions of the Lie powers. This paper is concerned withthe case where G has prime order p, and K has characteristicp. As is well known, there are p indecomposables, denoted hereby J₁,...,J_p, where J_r has dimension r. A theory is developedwhich provides information about the overall module structureof LV) and gives a recursive method for finding the multiplicitiesof J₁,...,J_p in the Lie powers Lⁿ(V). For example, the theoryyields decompositions of L(V) as a direct sum of modules isomorphiceither to J₁ or to an infinite sum of the form J_r J_{p-1} J_{p-1} ... with r 2. Closed formulae are obtained for the multiplicitiesof J₁,..., J_p in Lⁿ(J_p and Lⁿ(J_{p-1). For r < p-1, the indecomposableswhich occur with non-zero multiplicity in Lⁿ(J_r) are identifiedfor all sufficiently large n. 2000 Mathematical Subject Classification:17B01, 20C20.     

Ganea's Conjecture on Lusternik-Schnirelmann Category

A series of complexes Q_p indexed by all primes p is constructedwith catQ_p=2 and catQ_pxSⁿ=2 for either n2 or n=1 and p=2. Thisdisproves Ganea's conjecture on Lusternik–Schnirelmann(LS) category. 1991 Mathematics Subject Classification 55M30.     

Extremal mean width when covering the 1-skeleton

For a given convex body K in ^d, let D_n be the compact convexset of maximal mean width whose 1-skeleton can be covered byn congruent copies of K. Based on the fact that the mean widthis proportional to the average perimeter of two-dimensionalprojections, it is proved that D_n is close to being a segmentfor large n.     

On the Number of Singularities in Generic Deformations of Map Germs

Let f:Cⁿ, 0C^p, 0 be a K-finite map germ, and let i=(i₁, ...,i_k) be a Boardman symbol such that ⁱ has codimension n in thecorresponding jet space J^k(n, p). When its iterated successorshave codimension larger than n, the paper gives a list of situationsin which the number of ⁱ points that appear in a generic deformationof f can be computed algebraically by means of Jacobian idealsof f. This list can be summarised in the following way: f musthave rank n–i₁ and, in addition, in the case p=6, f mustbe a singularity of type ^i₁,i₂.     

The Direct Limits Of The Banach-mazur Compacta

Let 1 p . For each n-dimensional Banach space E = (E, || ·||), we define a norm || · ||_p on E x R as follows: [formula] It is shown that the correspondence (E, || · ||) (Ex R, || · ||_p) defines a topological embedding of oneBanach–Mazur compactum, BM(n), into another, BM(n 1),and hence we obtain a tower of Banach–Mazur compacta:BM(1) BM(2) BM(3) ···. Let BM_p be thedirect limit of this tower. We prove that BM_p is homeomorphicto Q = dir lim Qⁿ, where Q = [0, 1] is the Hilbert cube. 1991Mathematics Subject Classification 46B04, 46B20, 52A21, 57N20,54H15.     

Determination of a Convex Body from Minkowski Sums of its Projections

For a convex body K in R^d and 1 K d – 1, let P_K (K)be the Minkowski sum (average) of all orthogonal projectionsof K onto k-dimensional subspaces of R^d. It is Known that theoperator P_k is injective if kd/2, k=3 for all d, and if k =2, d 14. It is shown that P_2k (K) determines a convex body K among allcentrally symmetric convex bodies and P_2k+1(K) determines aconvex body K among all bodies of constant width. Correspondingstability results are also given. Furthermore, it is shown thatany convex body K is determined by the two sets P_k (K) and P_k'(K) if 1 < k < k'. Concerning the range of P_k , 1 k d–2, it is shown that its closure (in the Hausdorff-metric)does not contain any polytopes other than singletons.     

Optimal Smoothing for Convex Polytopes

It is proved that, given a convex polytope P in Rⁿ, togetherwith a collection of compact convex subsets in the interiorof each facet of P, there exists a smooth convex body arbitrarilyclose to P that coincides with each facet precisely along theprescribed sets, and has positive curvature elsewhere. 2000Mathematics Subject Classification 53A07, 52B11, 53C45.     

Realisations of Non v-Sufficient Jets with value in Rp, p > 1

The purpose of this paper is to show that if a jet cue oJ^r(n,p), n p, p > 1, is not v-sufficient in C^r+1, there existsan infinite sequence (f_i)_iN^* of realisations of o with mutuallynon-homeomorphic germs of varieties . Bochnak and Kuo [2, 5] showed it when p = 1 and thought thatthe same argument slightly modified can be used in the casep 2 [7, p. 225]. But when n p + 2, p > 1, we have to proceeddifferently. Moreover, it is necessary to prove separately theresult when n = p and n = p + 1. About C⁰-sufriciency and p> 1, Brodersen [3, p. 168] showed a similar theorem.     

On Restrictions and Extensions of the Besov and Triebel-Lizorkin Spaces with Respect to Lipschitz Domains

The restrictions B^s_pq() and F^s_pq() of the Besov and Triebel–Lizorkinspaces of tempered distributions B^s_pq(Rⁿ) and F^s_pq(Rⁿ) to Lipschitzdomains Rⁿ are studied. For general values of parameters (sR,p>0, q>0) a ‘universal’ linear bounded extensionoperator from B^s_pq() and F^s_pq() into the corresponding spaceson Rⁿ is constructed. The construction is based on a new variantof the Calderón reproducing formula with kernels supportedin a fixed cone. Explicit characterizations of the elementsof B^s_pq() and F^s_pq() in terms of their values in are also obtained.     

Decentralized dynamic pole assignment with low-order compensators

N. Karcanias Control Engineering Centre, School of Engineering and Mathematical Sciences, City University, Northampton Square London EC1V OHB, UK Email: n.karcanias{at}city.ac.uk Received on June 14, 2006; Accepted on October 2, 2006 The problem of arbitrary pole placement via dynamic decentralizedoutput feedback is studied for minimal systems described bya proper transfer function matrix P(s) R^{m x p}(s) (m = m_i andp = p_i), with McMillan degree n. The family of controllersto be used includes those decentralized controllers with channelswhose ith channel has maximum observability index at most d_i.The method presented here is based on asymptotic linearizationaround a decentralized degenerate compensator of the pole placementmap related to the problem. It is shown that the method worksgenerically when m⁺p > n, where m⁺ = min{d_i(p_i + m_i –1) + m_i}, i = 1, ..., , and the smallest d_i of the compensatorof the ith channel is the integral part of (n – pm_i)/p(p_i+ m_i – 1).     

Binomial Sums and Functions of Exponential Type

Let (a_n)_n0 be a sequence of complex numbers, and, for n0, let A number of results are proved relating the growth of the sequences(b_n) and (c_n) to that of (a_n). For example, given p0, if b_n= O(n^p and for all > 0,then a_n=0 for all n > p. Also, given 0 < p < 1, then for all > 0 if and onlyif . It is further shown that, given rß > 1, if b_n,c_n=O(rßⁿ), then a_n=O(ⁿ),where , thereby proving a conjecture of Chalendar, Kellay and Ransford. The principal ingredientsof the proogs are a Phragmén-Lindelöf theorem forentire functions of exponential type zero, and an estimate forthe expected value of e^(X), where X is a Poisson random variable.2000 Mathematics Subject Classification 05A10 (primary), 30D15,46H05, 60E15 (secondary).     

Geometric Approximation Of The Fibre Of The Freudenthal Suspension

Let Rat_k(CPⁿ) denote the space of based holomorphic maps ofdegree k from the Riemannian sphere S² to the complex projectivespace CPⁿ. The basepoint condition we assume is that f()=[1,..., 1]. Such holomorphic maps are given by rational functions: Rat_k(CPⁿ) ={(p₀(z), ..., p_n(z)):each p_i(z) is a monic, degree-kpolynomial and such that there are no roots common to all p_i(z)}.(1.1) The study of the topology of Rat_k(CPⁿ) originated in [10]. Later,the stable homotopy type of Rat_k(CPⁿ) was described in [3] interms of configuration spaces and Artin's braid groups. LetW(S²ⁿ) denote the homotopy theoretic fibre of the Freudenthalsuspension E:S²ⁿ S²ⁿ⁺¹. Then we have the following sequenceof fibrations: ²S²ⁿ⁺¹ W(S²ⁿ)S²ⁿ S²ⁿ⁺¹. A theorem in [10] tellsus that the inclusion Rat_k(CPⁿ) ²_kCP^{n 2}S²ⁿ⁺¹ is a homotopy equivalenceup to dimension k(2n–1). Thus if we form the direct limitRat(CPⁿ)= lim_k Rat_k(CPⁿ), we have, in particular, that Rat(CPⁿ)is homotopy equivalent to ²S²ⁿ⁺¹. If we take the results of [3] and [10] into account, we naturallyencounter the following problem: how to construct spaces X_k(CPⁿ),which are natural generalizations of Rat_k(CPⁿ), so that X(CPⁿ)approximates W(S²ⁿ). Moreover, we study the stable homotopytype of X_k(CPⁿ). The purpose of this paper is to give an answer to this problem.The results are stated after the following definition. 1991Mathematics Subject Classification 55P35.     

Hormander's Hp Multiplier Theorem for the Heisenberg Group

Let Hⁿ denote the (2n+1)-dimensional Heisenberg group. Givenan operator-valued function M, define the operator T_M with ‘^’ denoting theFourier transform. Hörmander-type sufficient conditionsare determined on M for the H^p-boundedness, p 1, of the operatorT_M on Hⁿ.