Bell inequality violations under reasonable and under weak hypotheses.

Given a sequence of pairs (pi, p[over ¯]i) of spin-1/2 particles in the singlet state, assume that Alice measures the normalized projections ai of the spins of the pi's along vector a while Bob measures the normalized projections bi of the spins of the p[over ¯]i's along vector b. Then quantum mechanics (QM) lets one evaluate the correlation as -cos(θa-θb) where θv is the angle between the vector v and a reference vector chosen once and for all and in a fixed plane. Assuming classical microscopic realism (CMR) there exist also normalized projection pairs (ai', bi') of the spins of the pairs (pi, p[over ¯]i) along (a', b') so that =-cos(θa'-θb'). Since all projections are in {-1,1}, |+|+|-|≤2 for c, d, e, and f in {a,b,a',b'}. Assuming locality (the impossibility of any effect of an event on another event when said events are spatially separated) beside QM and CMR, Bell's theory lets one deduce various violations of this inequality at some choices of quadruplets Q≡(a,b,a',b'). Our main result is the existence of such Q's where at least one of the above inequalities is violated if one only assumes QM, CMR, and some very mild further hypotheses that only concern the behavior of correlations that appear in these inequalities near special Q's.

Bounding the errors for convex dynamics on one or more polytopes.

We discuss the greedy algorithm for approximating a sequence of inputs in a family of polytopes lying in affine spaces by an output sequence made of vertices of the respective polytopes. More precisely, we consider here the case when the greed of the algorithm is dictated by the Euclidean norms of the successive cumulative errors. This algorithm can be interpreted as a time-dependent dynamical system in the vector space, where the errors live, or as a time-dependent dynamical system in an affine space containing copies of all the original polytopes. This affine space contains the inputs, as well as the inputs modified by adding the respective former errors; it is the evolution of these modified inputs that the dynamical system in affine space describes. Scheduling problems with many polytopes arise naturally, for instance, when the inputs are from a single polytope P, but one imposes the constraint that whenever the input belongs to a codimension n face, the output has to be in the same codimension n face (as when scheduling drivers among participants of a carpool). It has been previously shown that the error is bounded in the case of a single polytope by proving the existence of an arbitrary large convex invariant region for the dynamics in affine space: A region that is simultaneously invariant for several polytopes, each considered separately, was also constructed. It was then shown that there cannot be an invariant region in affine space in the general case of a family of polytopes. Here we prove the existence of an arbitrary large convex invariant set for the dynamics in the vector space in the case when the sizes of the polytopes in the family are bounded and the set of all the outgoing normals to all the faces of all the polytopes is finite. It was also previously known that starting from zero as the initial error set, the error set could not be saturated in finitely many steps in some cases with several polytopes: Contradicting a former conjecture, we show that the same happens for some single quadrilaterals and for a single pentagon with an axial symmetry. The disproof of that conjecture is the new piece of information that leads us to expect, and then to verify, as we recount here, that the proof that the errors are bounded in the general case could be a small step beyond the proof of the same statement for the single polytope case.

Convex dynamics: unavoidable difficulties in bounding some greedy algorithms.

A greedy algorithm for scheduling and digital printing with inputs in a convex polytope, and vertices of this polytope as successive outputs, has recently been proven to be bounded for any convex polytope in any dimension. This boundedness property follows readily from the existence of some invariant region for a dynamical system equivalent to the algorithm, which is what one proves. While the proof, and some constructions of invariant regions that can be made to depend on a single parameter, are reasonably simple for convex polygons in the plane, the proof of boundedness gets quite complicated in dimension three and above. We show here that such complexity is somehow justified by proving that the most natural generalization of the construction that works for polygons does not work in any dimension above two, even if we allow for as many parameters as there are faces. We first prove that some polytopes in dimension greater than two admit no invariant region to which they are combinatorially equivalent. We then modify these examples to get polytopes such that no invariant region can be obtained by pushing out the borders of the half spaces that intersect to form the polytope. We also show that another mechanism prevents some simplices (the simplest polytopes in any dimension) from admitting invariant regions to which they would be similar. By contrast in dimension two, one can always get an invariant region by pushing these borders far enough in some correlated way; for instance, pushing all borders by the same distance builds an invariant region for any polygon if the push is at a distance big enough for that polygon. To motivate the examples that we provide, we discuss briefly the bifurcations of polyhedra associated with pushing half spaces in parallel to themselves. In dimension three, the elementary codimension one bifurcation resembles the unfolding of the elementary degenerate singularity for codimension one foliations on surfaces. As the subject of this paper is new for the communities most interested in Chaos, we take some care in describing various links of our problem to classical issues (in particular linked to Diophantine approximation) as well as to various technological or commercial issues, exemplified, respectively, by digital printing and a problem in scheduling.

On the geometry of master-slave synchronization.

In 1990, Pecora and Carroll reported the observation that one can synchronize the orbits of two identical dynamical systems, which may be chaotic, by feeding state variables of one of them to the other one with no feedback, a phenomenon often called master-slave synchronization. We report here some results on the theory of master-slave synchronization for maps and flows, which are all inspired by a similar geometric and coordinate independent point of view to the one introduced in master-slave synchronization by Tresser, Worfolk, and Bass. Our results are variations on the theme that projection often can compensate for expansion.(c) 2002 American Institute of Physics.

Master-slave synchronization and the Lorenz equations.

Since the seminal remark by Pecora and Carroll [Phys. Rev. Lett. 64, 821 (1990)] that one can synchronize chaotic systems, the main example in the related literature has been the Lorenz equations. Yet this literature contains a mixture of true and false, and of justified and unsubstantiated claims about the synchronization properties of the Lorenz equations. In this note we clarify some of the confusion. (c) 1997 American Institute of Physics.

Piecewise linear models for the quasiperiodic transition to chaos.

We formulate and study analytically and computationally two families of piecewise linear degree one circle maps. These families offer the rare advantage of being non-trivial but essentially solvable models for the phenomenon of mode locking and the quasiperiodic transition to chaos. For instance, for these families, we obtain complete solutions to several questions still largely unanswered for families of smooth circle maps. Our main results describe (1) the sets of maps in these families having some prescribed rotation interval; (2) the boundaries between zero and positive topological entropy and between zero length and non-zero length rotation interval; and (3) the structure and bifurcations of the attractors in one of these families. We discuss the interpretation of these maps as low-order spline approximations to the classic "sine-circle" map and examine more generally the implications of our results for the case of smooth circle maps. We also mention a possible connection to recent experiments on models of a driven Josephson junction. (c) 1996 American Institute of Physics.

Master-slave synchronization from the point of view of global dynamics.

We present a mathematical framework for the theory of a synchronization phenomenon for dynamical systems discovered by Pecora and Carroll [Phys. Rev. Lett. 64, 821-824 (1990)]. From this perspective, we can synchronize, using a single coordinate, an open dense set of linear systems. We use our insights to synchronize nonlinear systems which were not previously recognized as being synchronizable. (c) 1995 American Institute of Physics.

A decoding problem in dynamics and in number theory.

Given a homeomorphism f of the circle, any splitting of this circle in two semiopen arcs induces a coding process for the orbits of f, which can be determined by recording the successive arcs visited by the orbit. The problem of describing these codes has a two hundred year history (that we briefly recall) in the particular case when the arcs are limited by a point and its image; in modern language, it is the kneading theory of such maps, and as such is relevant for our understanding of dynamical problems involving oscillations. This paper deals with questions attached to the general case, a problem considered by many mathematicians in the 50's and 60's in the case where f is a rotation, and which has recently found some applications in physiology. We show that, except for trivial cases, any code determines the rotation number, up to the orientation, of the homeomorphism which generates it. In the case the code is periodic, we can also determine whether or not it can be generated in this way. An equivalent problem in arithmetic consists of finding +/-p, knowing a collection of classes in Z/qZ of the form {m,m+p,.,m+(k-1)p}, where 2

An approach to renormalization on the n-torus.

The coding theory of rotations (by inspecting closely their relation to flows) and the continued fractions algorithm (by considering even two-coloring of the integers with a given proportion of, say, blue and red) are revisited. Then, even n-coloring of the integers is defined. This allows one to code rotations on the (n-1)-torus by considering linear flows on the n-torus and yields a simple geometric approach to renormalization on tori by first return maps on the coding regions.