arXiv subjects

Shamgar Gurevich

Publications and source records attributed to Shamgar Gurevich.

At least 19 recordsLinked to original sources

arXiv2021

Ranks for Representations of GL(n) Over Finite Fields, their Agreement, and Positivity of Fourier Transform

In [Frobenius1896] it was shown that many important properties of a finite group could be examined using formulas involving the character ratios of group elements, i.e., the trace of the element acting in a given irreducible representation, divided by the dimension of the representation. In [Gurevich-Howe15] and [Gurevich-Howe17], the current authors introduced the notion of rank of an irreducible representation of a finite classical group. One of the motivations for studying rank was to clarify the nature of character ratios for certain elements in these groups. In fact in the above cited papers, two notions of rank were given. The first is the Fourier theoretic based notion of U-rank of a representation, which comes up when one looks at its restrictions to certain abelian unipotent subgroups. The second is the more algebraic based notion of tensor rank which comes up naturally when one attempts to equip the representation ring of the group with a grading that reflects the central role played by the few "smallest" possible representations of the group. In [Gurevich-Howe17] we conjectured that the two notions of rank mentioned just above agree on a suitable collection called "low rank" representations. In this note we review the development of the theory of rank for the case of the general linear group GL_n over a finite field F_q, and give a proof of the "agreement conjecture" that holds true for sufficiently large q. Our proof is Fourier theoretic in nature, and uses a certain curious positivity property of the Fourier transform of the set of matrices of low enough fixed rank in the vector space of matrices of size m x n over F_q. In order to make the story we are trying to tell clear, we choose in this note to follow a particular example that shows how one might apply the theory of rank to certain counting problems.

math.RT↗

arXiv2021

Harmonic Analysis on GL(n) over Finite Fields

There are many formulas that express interesting properties of a finite group G in terms of sums over its characters. For estimating these sums, one of the most salient quantities to understand is the character ratio trace(π(g)) / dim(π), for an irreducible representation πof G and an element g of G. It turns out [Gurevich-Howe15, Gurevich-Howe17] that for classical groups G over finite fields there are several (compatible) invariants of representations that provide strong information on the character ratios. We call these invariants collectively rank. Rank suggests a new way to organize the representations of classical groups over finite and local fields - a way in which the building blocks are the "smallest" representations. This is in contrast to Harish-Chandra's philosophy of cusp forms that is the main organizational principle since the 60s, and in it the building blocks are the cuspidal representations which are, in some sense, the "LARGEST". The philosophy of cusp forms is well adapted to establishing the Plancherel formula for reductive groups over local fields, and led to Lusztig's classification of the irreducible representations of such groups over finite fields. However, analysis of character ratios might benefit from a different approach. In this note we discuss further the notion of tensor rank for GL_n over a finite field F_q and demonstrate how to get information on representations of a given tensor rank using tools coming from the recently studied eta correspondence, as well as the well known philosophy of cusp forms, mentioned just above. A significant discovery so far is that although the dimensions of the irreducible representations of a given tensor rank vary by quite a lot (they can differ by large powers of q), for certain group elements of interest the character ratios of these irreps are nearly equal to each other.

math.RT↗

arXiv2021

The Pieri Rule for GLn Over Finite Fields

The Pieri rule gives an explicit formula for the decomposition of the tensor product of irreducible representation of the complex general linear group GL(n,C) with a symmetric power of the standard representation on C^n. It is an important and long understood special case of the Littlewood-Richardson rule for decomposing general tensor products of representations of GL(n,C). In our recent work [Gurevich-Howe17, Gurevich-Howe19] on the organization of representations of the general linear group over a finite field F_q using small representations, we used a generalization of the Pieri rule to the context of this latter group. In this note, we demonstrate how to derive the Pieri rule for GL(n,Fq). This is done in two steps; the first, reduces the task to the case of the symmetric group S_n, using the natural relation between the representations of S_n and the spherical principal series representations of GL(n,F_q); while in the second step, inspired by a remark of Nolan Wallach, the rule is obtained for S_n invoking the S_\ell-GL_(n,C)) Schur duality. Along the way, we advertise an approach to the representation theory of the symmetric group which emphasizes the central role played by the dominance order on Young diagrams. The ideas leading to this approach seem to appear first, without proofs, in [Howe-Moy86].

math.RT↗

arXiv2021

A look at Representations of SL(2,F_q) through the Lens of Size

How to study a nice function on the real line? The physically motivated Fourier theory technique of harmonic analysis is to expand the function in the basis of exponentials and study the meaningful terms in the expansion. Now, suppose the function lives on a finite non-commutative group G, and is invariant under conjugation. There is a well-known analog of Fourier analysis, using the irreducible characters of G. This can be applied to many functions that express interesting properties of G. To study these functions one wants to know how the different characters contribute to the sum? In this note we describe the G=SL(2,F_q) case of the theory we have been developing in recent years which attempts to give a fairly general answer to the above question for finite classical groups. The irreducible representations of SL(2,F_q) are "well known" for a very long time [Frobenius1896, Jordan1907, Schur1907] and are a prototype example in many introductory courses on the subject. We are happy that we can say something new about them. In particular, it turns out that the representations that were considered as "anomalous" in the "old" point of view (known as the "philosophy of cusp forms") are the building blocks of the current approach.

math.RT↗

arXiv2021

Rank and Duality in Representation Theory

There is both theoretical and numerical evidence that the set of irreducible representations of a reductive group over local or finite fields is naturally partitioned into families according to analytic properties of representations. Examples of such properties are the rate of decay at infinity of "matrix coefficients" in the local field setting, and the order of magnitude of "character ratios" in the finite field situation. In these notes we describe known results, new results, and conjectures in the theory of "size" of representations of classical groups over finite fields, whose ultimate goal is to classify the above mentioned families of representations and accordingly to estimate the relevant analytic properties of each family. Specifically, we treat two main issues: the first is the introduction of a rigorous definition of a notion of size for representations of classical groups, and the second issue is a method to construct and obtain information on each family of representation of a given size. In particular, we propose several compatible notions of size that we call U-RANK, TENSOR RANK and ASYMPTOTIC RANK, and we develop a method called ETA CORRESPONDENCE to construct the families of representation of each given rank. Rank suggests a new way to organize the representations of classical groups over finite and local fields - a way in which the building blocks are the "smallest" representations. This is in contrast to Harish-Chandra's philosophy of cusp forms that is the main organizational principle since the 60s, and in it the building blocks are the cuspidal representations which are, in some sense, the "LARGEST".

math.RT↗

arXiv2016

Small Representations of Finite Classical Groups

Finite group theorists have established many formulas that express interesting properties of a finite group in terms of sums of characters of the group. An obstacle to applying these formulas is lack of control over the dimensions of representations of the group. In particular, the representations of small dimensions tend to contribute the largest terms to these sums, so a systematic knowledge of these small representations could lead to proofs of important conjectures which are currently out of reach. Despite the classification by Lusztig of the irreducible representations of finite groups of Lie type, it seems that this aspect remains obscure. In this note we develop a language which seems to be adequate for the description of the "small" representations of finite classical groups and puts in the forefront the notion of rank of a representation. We describe a method, the "eta correspondence", to construct small representations, and we conjecture that our construction is exhaustive. We also give a strong estimate on the dimension of small representations in terms of their rank. For the sake of clarity, in this note we describe in detail only the case of the finite symplectic groups.

math.RT↗

arXiv2014

Performance Estimates of the Pseudo-Random Method for Radar Detection

A performance of the pseudo-random method for the radar detection is analyzed. The radar sends a pseudo-random sequence of length $N$, and receives echo from $r$ targets. We assume the natural assumptions of uniformity on the channel and of the square root cancellation on the noise. Then for $r \leq N^{1-δ}$, where $δ> 0$, the following holds: (i) the probability of detection goes to one, and (ii) the expected number of false targets goes to zero, as $N$ goes to infinity.

cs.IT↗

arXiv2013

Almost Linear Complexity Methods for Delay-Doppler Channel Estimation

A fundamental task in wireless communication is channel estimation: Compute the channel parameters a signal undergoes while traveling from a transmitter to a receiver. In the case of delay-Doppler channel, i.e., a signal undergoes only delay and Doppler shifts, a widely used method to compute delay-Doppler parameters is the pseudo-random method. It uses a pseudo-random sequence of length N; and, in case of non-trivial relative velocity between transmitter and receiver, its computational complexity is O(N^2logN) arithmetic operations. In [1] the flag method was introduced to provide a faster algorithm for delay-Doppler channel estimation. It uses specially designed flag sequences and its complexity is O(rNlogN) for channels of sparsity r. In these notes, we introduce the incidence and cross methods for channel estimation. They use triple-chirp and double-chirp sequences of length N, correspondingly. These sequences are closely related to chirp sequences widely used in radar systems. The arithmetic complexity of the incidence and cross methods is O(NlogN + r^3), and O(NlogN + r^2), respectively.

cs.IT↗

arXiv2013

The Incidence and Cross Methods for Efficient Radar Detection

The designation of the radar system is to detect the position and velocity of targets around us. The radar transmits a waveform, which is reflected back from the targets, and echo waveform is received. In a commonly used model, the echo is a sum of a superposition of several delay-Doppler shifts of the transmitted waveform, and a noise component. The delay and Doppler parameters encode, respectively, the distances, and relative velocities, between the targets and the radar. Using standard digital-to-analog and sampling techniques, the estimation task of the delay-Doppler parameters, which involves waveforms, is reduced to a problem for complex sequences of finite length N. In these notes we introduce the Incidence and Cross methods for radar detection. One of their advantages, is robustness to inhomogeneous radar scene, i.e., for sensing small targets in the vicinity of large objects. The arithmetic complexity of the incidence and cross methods is O(NlogN + r^3) and O(NlogN + r^2), for r targets, respectively. In the case of noisy environment, these are the fastest radar detection techniques. Both methods employ chirp sequences, which are commonly used by radar systems, and hence are attractive for real world applications.

cs.IT↗

arXiv2012

Delay-Doppler Channel Estimation with Almost Linear Complexity

A fundamental task in wireless communication is Channel Estimation: Compute the channel parameters a signal undergoes while traveling from a transmitter to a receiver. In the case of delay-Doppler channel, a widely used method is the Matched Filter algorithm. It uses a pseudo-random sequence of length N, and, in case of non-trivial relative velocity between transmitter and receiver, its computational complexity is O(N^{2}log(N)). In this paper we introduce a novel approach of designing sequences that allow faster channel estimation. Using group representation techniques we construct sequences, which enable us to introduce a new algorithm, called the flag method, that significantly improves the matched filter algorithm. The flag method finds the channel parameters in O(mNlog(N)) operations, for channel of sparsity m. We discuss applications of the flag method to GPS, radar system, and mobile communication as well.

cs.IT↗

arXiv2011

Computing the Matched Filter in Linear Time

A fundamental problem in wireless communication is the time-frequency shift (TFS) problem: Find the time-frequency shift of a signal in a noisy environment. The shift is the result of time asynchronization of a sender with a receiver, and of non-zero speed of a sender with respect to a receiver. A classical solution of a discrete analog of the TFS problem is called the matched filter algorithm. It uses a pseudo-random waveform S(t) of the length p, and its arithmetic complexity is O(p^{2} \cdot log (p)), using fast Fourier transform. In these notes we introduce a novel approach of designing new waveforms that allow faster matched filter algorithm. We use techniques from group representation theory to design waveforms S(t), which enable us to introduce two fast matched filter (FMF) algorithms, called the flag algorithm, and the cross algorithm. These methods solve the TFS problem in O(p\cdot log (p)) operations. We discuss applications of the algorithms to mobile communication, GPS, and radar.

cs.IT↗

arXiv2011

The categorical Weil representation

In a previous work the authors gave a conceptual explanation for the linearity of the Weil representation over a finite field k of odd characteristic: There exists a canonical system of intertwining operators between the Lagrangian models of the Heisenberg representation. This defines a canonical vector space H(V) associated with a symplectic vector space V over k. In this paper we prove a general theorem about idempotents in categories, and we use it to solve the sign problem, formulated by Bernstein and Deligne, on the compatibility between the associativity constraint and the convolution structure of the l-adic sheaf of canonical intertwining kernels. This sheaf governs---via the sheaf-to-function correspondence---the function theoretic system of intertwiners. As an application we define a canonical category C(V) associated with the symplectic vector space variety V, and we obtain the canonical model of the categorical Weil representation.

math.RT↗

arXiv2010

Bounds on certain Higher-Dimensional Exponential Sums via the Self-Reducibility of the Weil Representation

We describe a new method to bound certain higher-dimensional exponential sums which are associated with tori in symplectic groups over finite fields. Our method is based on the self-reducibility property of the Weil representation. As a result, we obtain a sharp form of the Hecke quantum unique ergodicity theorem for generic linear symplectomorphisms of the 2N-dimensional torus.

math.RT↗

arXiv2009

Quantization of symplectic vector spaces over finite fields

In this paper, we construct a quantization functor, associating a complex vector space H(V) to a finite dimensional symplectic vector space V over a finite field of odd characteristic. As a result, we obtain a canonical model for the Weil representation of the symplectic group Sp(V). The main new technical result is a proof of a stronger form of the Stone-von Neumann property for the Heisenberg group. Our result answers, for the case of the Heisenberg group, a question of Kazhdan about the possible existence of a canonical vector space attached to a coadjoint orbit of a general unipotent group over finite field.

math.RT↗

arXiv2009

Notes on the Self-Reducibility of the Weil Representation and Higher-Dimensional Quantum Chaos

In these notes we discuss the "self-reducibility property" of the Weil representation. We explain how to use this property to obtain sharp estimates of certain higher-dimensional exponential sums which originate from the theory of quantum chaos. As a result, we obtain the Hecke quantum unique ergodicity theorem for generic linear symplectomorphism $A$ of the torus $T^{2N}=R^{2N}/Z^{2N}.

math-ph↗

arXiv2009

Notes on quantization of symplectic vector spaces over finite fields

In these notes we construct a quantization functor, associating an Hilbert space H(V) to a finite dimensional symplectic vector space V over a finite field F_q. As a result, we obtain a canonical model for the Weil representation of the symplectic group Sp(V). The main technical result is a proof of a stronger form of the Stone-von Neumann theorem for the Heisenberg group over F_q. Our result answers, for the case of the Heisenberg group, a question of Kazhdan about the possible existence of a canonical Hilbert space attached to a coadjoint orbit of a general unipotent group over F_q.

math-ph↗

arXiv2009

Statistical RIP and Semi-Circle Distribution of Incoherent Dictionaries

In this paper we formulate and prove a statistical version of the Candes-Tao restricted isometry property (SRIP for short) which holds in general for any incoherent dictionary which is a disjoint union of orthonormal bases. In addition, we prove that, under appropriate normalization, the eigenvalues of the associated Gram matrix fluctuate around 1 according to the Wigner semicircle distribution. The result is then applied to various dictionaries that arise naturally in the setting of finite harmonic analysis, giving, in particular, a better understanding on a remark of Applebaum-Howard-Searle-Calderbank concerning RIP for the Heisenberg dictionary of chirp like functions.

cs.IT↗

arXiv2009

Incoherent dictionaries and the statistical restricted isometry property

In this article we present a statistical version of the Candes-Tao restricted isometry property (SRIP for short) which holds in general for any incoherent dictionary which is a disjoint union of orthonormal bases. In addition, under appropriate normalization, the eigenvalues of the associated Gram matrix fluctuate around 1 according to the Wigner semicircle distribution. The result is then applied to various dictionaries that arise naturally in the setting of finite harmonic analysis, giving, in particular, a better understanding on a remark of Applebaum-Howard-Searle-Calderbank concerning RIP for the Heisenberg dictionary of chirp like functions.

cs.IT↗