---
title: Tubular Domains over Symmetric Cones
url: https://www.emergentmind.com/topics/tubular-domains-over-symmetric-cones
type: topic
---

# Tubular Domains over Symmetric Cones

Searching arXiv for recent and foundational papers on tube domains over symmetric cones.
Tubular domains over symmetric cones are domains of the form $T(\Omega)=V+i\Omega$ in the complexification of a real vector space or Euclidean Jordan algebra, where $\Omega$ is a cone carrying strong homogeneity and symmetry properties. In the finite-dimensional classical setting, $\Omega$ is an irreducible symmetric cone, equivalently an open, convex, self-dual, homogeneous cone, and the resulting tube domain is a tube-type bounded symmetric domain after Cayley transform. In the Banach-space setting, Chu proved that a tube domain $V\oplus i\Omega$ is biholomorphic to a bounded symmetric domain if and only if $\Omega$ is a normal, linearly homogeneous, Finsler symmetric cone, equivalently if $V$ is a unital JB-algebra in an equivalent norm and $\Omega=\operatorname{int}\{v^2:v\in V\}$ [2006.06499]. Around this geometric core, the subject connects Jordan theory, automorphism groups, invariant metrics, Hardy–Bergman analysis, Toeplitz and Hankel operators, Carleson embeddings, and the geometry of Hermitian symmetric spaces [2210.16213].

## 1. Foundational definitions and Jordan-theoretic structure

A tube domain over a cone is defined by
$$
T(\Omega)=V\oplus i\Omega=\{x+iy:x\in V,\ y\in\Omega\},
$$
where $V$ is a real Banach space or finite-dimensional Euclidean space, and $\Omega\subset V$ is a proper open cone in the Banach setting or an irreducible symmetric cone in the Euclidean Jordan setting [2006.06499]. In the finite-dimensional theory, $V$ is a Euclidean Jordan algebra with identity $e$, rank $r$, dimension $n$, determinant $\Delta$, and often Peirce constant $d$, satisfying
$$
n=r+\frac{d\,r(r-1)}{2}.
$$
The cone $\Omega$ is then the open cone of squares, and may also be described spectrally by principal minors $\Delta_k$ via
$$
\Omega=\{x\in V:\Delta_k(x)>0,\ k=1,\dots,r\}
$$
[2509.22024].

The terminology “tube domain” and “Siegel domain of the first kind” is synonymous in this context [2006.06499]. In complex analysis on such domains, one writes $z=x+iy$ with $y\in\Omega$ and uses the Jordan determinant through the standard shorthand
$$
\delta(z)=\Delta(\operatorname{Im} z)
$$
[2509.22024].

A proper open cone induces an order structure on $V$ by $x\le y$ if $y-x\in\overline{\Omega}$. Every $e\in\Omega$ is an order unit, and the associated order-unit norm is
$$
\|x\|_e=\inf\{\lambda>0:-\lambda e\le x\le \lambda e\}.
$$
For proper cones, these norms are equivalent for different $e\in\Omega$ [2006.06499]. In the Banach-space characterization of symmetric tube domains, normality of the cone is equivalent to equivalence between the ambient norm and the order-unit norms [2006.06499].

In the finite-dimensional Euclidean Jordan setting, generalized powers and principal minors play a central role. For $s=(s_1,\dots,s_r)\in\mathbb C^r$,
$$
\Delta_s(x)=\Delta_1(x)^{s_1-s_2}\Delta_2(x)^{s_2-s_3}\cdots \Delta_r(x)^{s_r},
$$
and these functions drive the explicit kernel and integral formulas used in Bergman theory [1703.07862, 1109.1737].

## 2. Symmetric cones, bounded symmetric domains, and the main characterization

The central structural theorem in the Banach setting states that for a proper open cone $\Omega$ in a real Banach space $V$, the following are equivalent: $T(\Omega)=V\oplus i\Omega$ is biholomorphic to a bounded symmetric domain; $\Omega$ is a normal, linearly homogeneous, Finsler symmetric cone; and $V$ is a unital JB-algebra in an equivalent norm with
$$
\Omega=\operatorname{int}\{v^2:v\in V\}.
$$
This is Chu’s main theorem [2006.06499].

A Finsler symmetric cone is defined using a compatible tangent norm $v:T\Omega\to[0,\infty)$, a $G(\Omega)$-invariant symmetric-space structure, and point symmetries that are involutive $v$-isometries with isolated fixed points [2006.06499]. This replaces finite-dimensional self-duality by a Banach-space notion adapted to the absence of a canonical positive-definite pairing. In finite dimension, by contrast, symmetric cones are classically open, convex, self-dual, and homogeneous [2210.16213].

The finite-dimensional counterpart is the Koecher–Vinberg picture: if $V$ is finite-dimensional, then $T(\Omega)$ is biholomorphic to a bounded symmetric domain if and only if $\Omega$ is linearly homogeneous and self-dual, equivalently if $V$ is a Euclidean Jordan algebra and
$$
\Omega=\operatorname{int}\{x^2:x\in V\}
$$
[2006.06499]. Chu’s theorem extends this by replacing self-duality with “normal, linearly homogeneous, Finsler symmetric” in Banach spaces [2006.06499].

A frequent misconception is that every homogeneous cone automatically yields a bounded symmetric tube domain. The finite-dimensional data do not support that statement: the cone must be symmetric, meaning both homogeneous and self-dual in the Euclidean Jordan sense [2210.16213]. In the Banach formulation, the corresponding replacement is the stronger package “normal, linearly homogeneous, Finsler symmetric” [2006.06499].

The bounded realization is obtained by a Cayley transform modeled on the Jordan structure. In finite dimensions, for a unital Jordan algebra with identity $e$,
$$
C(z)=(z-e)(z+e)^{-1},
$$
defined where $z+e$ is invertible, maps the tube onto a bounded symmetric domain of tube type [2006.06499]. This transform underlies the equivalence between tube realizations and bounded symmetric domains of tube type.

## 3. Lie-theoretic and automorphism-group descriptions

The linear automorphism group of a cone is
$$
G(\Omega)=\{g\in GL(V):g(\Omega)=\Omega\},
$$
which is a closed Banach Lie subgroup in the Banach setting [2006.06499]. Linearly homogeneous means precisely that $G(\Omega)$ acts transitively on $\Omega$ [2006.06499].

In Chu’s reconstruction of the algebra from the cone, fixing $e\in\Omega$ yields a symmetry $S_e$ and an induced involution on the Banach–Lie algebra of Killing fields, producing the decomposition
$$
\operatorname{Kill}\Omega=\mathfrak k\oplus\mathfrak p,
$$
with
$$
[\mathfrak k,\mathfrak k]\subset\mathfrak k,\qquad
[\mathfrak k,\mathfrak p]\subset\mathfrak p,\qquad
[\mathfrak p,\mathfrak p]\subset\mathfrak k.
$$
Evaluation at $e$ identifies $\mathfrak p\simeq V$, and the Jordan product is reconstructed by
$$
xy=L(x)(y),
$$
where $L(x)\in\mathfrak p$ is the unique element with $L(x)(e)=x$ [2006.06499]. The Jordan identity is then established by operator-theoretic arguments involving complexification and commutator estimates [2006.06499].

In the real finite-dimensional tube-type setting, de Oliveira described the automorphism group of the real tube domain $T_\Omega=\Omega\oplus V_-$ associated with a real semisimple Jordan algebra with Cartan decomposition $V=V_+\oplus V_-$ [1012.1018]. The automorphism group is generated by the cone-preserving linear group, translations along $V_-$, and Jordan inversion:
$$
L=\langle G(\Omega)\cup N_+\cup\{j\}\rangle,
$$
with $j(x)=x^{-1}$ [1012.1018]. This is the real analogue of the classical complex statement that automorphisms of a tube-type domain are generated by translations, linear cone automorphisms, and inversion [1012.1018].

The same Jordan-algebraic mechanisms appear in concrete examples. For symmetric positive-definite matrices, the action is by congruence and inversion is matrix inversion [1012.1018]. For the Lorentz cone, the Jordan inverse has the explicit quadratic-fractional form recorded in the spin-factor model [1012.1018].

A related Lie-theoretic perspective comes from Hermitian symmetric spaces. For a non-compact irreducible Hermitian symmetric space $G/K$ with Iwasawa decomposition $G=NAK$, Geatti and Iannuzzi showed that every $N$-invariant domain corresponds to a tube domain in $H^r$, the product of $r$ upper half-planes, where $r=\dim\mathfrak a$ [2210.16213]. The explicit biholomorphism
$$
\pounds:H^r\to R\exp(\mathfrak a)\cdot eK
$$
realizes an embedded tube geometry inside $G/K$ [2210.16213]. This suggests that tube-domain geometry over cones is not merely a Jordan-algebraic artifact, but also a canonical coordinate model for large classes of symmetric spaces.

## 4. Invariant metrics, Finsler geometry, and convexity phenomena

For a normal cone $\Omega$, Chu defines a $G(\Omega)$-invariant tangent norm by
$$
T(p,v)=\|v\|_p,
$$
where $\|\cdot\|_p$ is the order-unit norm induced by $p\in\Omega$ [2006.06499]. Its integrated distance is Thompson’s metric:
$$
d_T(x,y)=\max\{\log M(x/y),\log M(y/x)\},
$$
with
$$
M(x/y)=\inf\{\lambda>0:x\le \lambda y\}.
$$
On the purely imaginary slice of the tube domain, the Carathéodory distance restricts to Thompson’s metric:
$$
p(ix,iy)=d_T(x,y)
$$
[2006.06499].

This metric identification links order-theoretic geometry on the cone with holomorphic geometry on the tube. Under biholomorphism to a bounded symmetric domain, the invariant Finsler structure corresponds to canonical invariant metrics such as Carathéodory, Bergman, or Kobayashi metrics up to constants [2006.06499]. In finite dimensions or Hilbert-space settings, Riemannian symmetric-space metrics may coexist with this Finsler description [2006.06499].

On the analytic side, Bergman geometry on finite-dimensional tubes is encoded by the weighted Bergman kernel and the Bergman metric. For tube domains over irreducible symmetric cones, Bergman balls admit volume asymptotics
$$
\operatorname{Vol}_\nu(B(z,\rho))\asymp \Delta(\operatorname{Im} z)^{\nu+n/r},
$$
uniformly in $z$ for fixed small $\rho$ [1408.3072]. Korányi-type comparability lemmas control kernels on nearby points in a Bergman ball and are used throughout operator theory on these domains [1408.3072, 1508.05819].

Convexity enters from another direction in the work of Geatti–Iannuzzi. If $D\subset G/K$ is $N$-invariant and corresponds to a tube $T(\Omega)=\mathbb R^r+i\Omega\subset H^r$, then
$$
D\ \text{is Stein}\iff \Omega\ \text{is convex and }C\text{-invariant},
$$
where the cone $C$ depends on whether the ambient Hermitian symmetric space is tube type or non-tube type [2210.16213]. The envelope of holomorphy is obtained by replacing $\Omega$ with its smallest convex, cone-invariant hull [2210.16213]. In the classical translation-invariant case, Bochner’s theorem uses only convexity; the extra cone-invariance reflects the $N$-action and restricted-root geometry [2210.16213].

## 5. Analytic function spaces and Bergman theory on tube domains

Finite-dimensional tube domains over symmetric cones support an explicit Bergman theory governed by the determinant $\Delta$. A standard weighted measure is
$$
dV_\nu(z)=\Delta(\operatorname{Im} z)^{\nu-n/r}\,dx\,dy,
$$
and the weighted Bergman kernel has the form
$$
K_\nu(z,w)=c_\nu\,\Delta\!\left(\frac{z-\overline w}{2i}\right)^{-(\nu+n/r)}
$$
or, in equivalent normalizations used across the literature,
$$
B_\nu(z,w)=d_\nu\,\Delta\!\left(\frac{z-\overline z}{i}\right)^{-\nu-n/r}
$$
[1205.3323, 2509.22024, 1703.07859, 1508.05819]. The corresponding weighted Bergman projection is
$$
(P_\nu f)(z)=\int_{T_\Omega} B_\nu(z,w)f(w)\Delta(\operatorname{Im} w)^{\nu-n/r}\,dv(w)
$$
[2509.22024].

The mixed-norm weighted Bergman spaces are defined by
$$
A_\nu^{p,q}(T_\Omega)=\left\{F\in\mathcal H(T_\Omega):
\left(\int_\Omega\left(\int_V |F(x+iy)|^pdx\right)^{q/p}\Delta(y)^\nu dy\right)^{1/q}<\infty\right\},
$$
with scalar or vector-weight variants depending on the setting [2509.22024, 1703.07862]. In the generalized-power formalism of Békollé–Bonami–Garrigós–Ricci–Peloso–Sehba and related work, vector weights appear through $\Delta_s$ and the measure
$$
dV_s(z)=\Delta_{s-n/r}(\operatorname{Im} z)\,dx\,dy
$$
[1703.07862].

Atomic decomposition and interpolation are particularly important for these spaces. For mixed-norm spaces on tube domains over irreducible symmetric cones, a Whitney decomposition adapted to Bergman balls yields atoms of the form
$$
a_{l,j}(z)=\Delta_{s+n/r}(y_j)K_s(z,z_{l,j}),
$$
and every $f\in A_s^{p,q}$ admits an expansion
$$
f(z)=\sum_{l,j} c_{l,j} a_{l,j}(z)
$$
with sequence norms equivalent to the function norm [1703.07862]. The same paper proves complex interpolation identities such as
$$
[A_{s_0}^{p_0,q_0}(T_\Omega),A_{s_1}^{p_1,q_1}(T_\Omega)]_\theta
=
A_s^{p,q}(T_\Omega),
$$
with the standard convex interpolation relations among $p,q,s$ [1703.07862].

Hardy spaces on tube domains also admit Paley–Wiener representations. For the Hilbert–Hardy space,
$$
\|F\|_{H^2(T_\Omega)}^2=\sup_{y\in\Omega}\int_V |F(x+iy)|^2dx,
$$
and $F\in H^2(T_\Omega)$ if and only if
$$
F(z)=\int_\Omega f(t)e^{i\langle z,t\rangle}dt
$$
for some $f\in L^2(\Omega)$ [1710.11237]. When $2m>n/r-1$, the $m$th Box operator provides an isomorphism
$$
\square^m:H^2(T_\Omega)\to A_{2m}^2(T_\Omega)
$$
[1710.11237]. This reduction is central in derivative-embedding problems.

The Duren–Carleson theorem has also been extended to tube domains over symmetric cones. For $0<p<q<\infty$, the characterization of measures $\mu$ such that $H^p(T_\Omega)$ embeds continuously into $L^q(T_\Omega,d\mu)$ is reduced to testing against standard measures and generalized powers of principal minors [1601.04899]. That work also proves the Hardy–Littlewood-type embedding
$$
H^2(T_\Omega)\hookrightarrow A_{\delta-1}^4(T_\Omega),\qquad \delta=\frac nr,
$$
and derives multiplier theorems from $H^{2m}$ to Bergman spaces [1601.04899].

## 6. Operator theory, Carleson measures, and current analytic directions

Operator theory on tube domains over symmetric cones has developed around Bergman projections, Toeplitz and Hankel operators, Cesàro-type operators, and Carleson embeddings. On weighted Bergman spaces $A_\nu^p(T_\Omega)$, a positive Borel measure $\mu$ is a $(q,p)$-Carleson measure when
$$
\int_{T_\Omega}|f(z)|^q\,d\mu(z)\le C\|f\|_{A_\nu^p}^q.
$$
For $1\le p\le q<\infty$, Nana–Sehba proved that this embedding is bounded if and only if
$$
\mu(B(z,\rho))\le C'\Delta(\operatorname{Im} z)^{\nu+n/r}
$$
for Bergman balls $B(z,\rho)$ [1408.3072]. For the case $q<p$, the criterion becomes an $L^s$ condition on the normalized averaging function
$$
\widehat\mu_\rho(z)=\frac{\mu(B(z,\rho))}{\Delta(\operatorname{Im} z)^{\nu+n/r}},
\qquad s=\frac{p}{p-q},
$$
together with boundedness of a suitable Bergman projection in the necessity direction [1408.3072].

These Carleson conditions feed directly into Toeplitz and Cesàro-type Schatten theory. For a positive measure $\mu$, the Toeplitz operator
$$
(T_\mu f)(z)=\int_{T_\Omega} K_\nu(z,w)f(w)\,d\mu(w)
$$
belongs to the Schatten class $\mathcal S_p(A_\nu^2)$ if and only if any of the following equivalent conditions holds: the lattice averages
$$
\sum_j\left(\frac{\mu(B(z_j,\rho))}{\Delta(\operatorname{Im} z_j)^{\nu+n/r}}\right)^p<\infty,
$$
the average function $\widehat\mu_\rho$ belongs to $L^p(T_\Omega,d\lambda)$, or the Berezin transform belongs to the same space [1408.3072]. Sehba extended such results to small exponents $0<p<1$, obtaining equivalence with Berezin-transform criteria under the sharp determinant-integrability threshold
$$
p(\nu+2)>2n/r-1
$$
[1710.02020].

Cesàro-type operators are defined through the Box operator by solving
$$
\square^m F=f\cdot \square^m g
$$
and taking the class of $F$ in a quotient space of holomorphic functions [1408.3072]. For $p\ge2$, such an operator lies in $\mathcal S_p(A_\nu^2)$ if and only if
$$
g\in B_{-n/r}^p(T_\Omega),
$$
equivalently
$$
\int_{T_\Omega} |\square^m g(z)|^p \Delta(\operatorname{Im} z)^{mp-n/r}dV(z)<\infty
$$
[1408.3072].

Toeplitz and Hankel operators from Bergman spaces to analytic Besov spaces have also been characterized. For instance, boundedness of $T_\mu:A_\alpha^p\to B_\beta^q$ is equivalent, in the no-loss regime $1<p\le q<\infty$, to the Bergman-ball estimate
$$
\mu(B_\delta(z))\le C\,\Delta(\operatorname{Im} z)^{\alpha+n/r}
$$
[1508.05819]. The same work characterizes Hankel operators in terms of symbol membership in appropriate analytic Besov spaces and derives weak factorization theorems for Bergman spaces [1508.05819].

A parallel line concerns mixed-norm and off-diagonal boundedness of Bergman-type operators. For the family
$$
T_{\alpha,\beta,\gamma}f(z)=\Delta(\operatorname{Im} z)^{-\alpha}\int_{T_\Omega} B_\gamma(z,w)f(w)\Delta(\operatorname{Im} w)^{-\beta}\,dV(w),
$$
necessary and sufficient conditions were obtained for boundedness between mixed-norm spaces $L_\nu^{p,q}(T_\Omega)\to L_\mu^{p,s}(T_\Omega)$, including explicit homogeneity and Schur-type inequalities relating $\alpha,\beta,\gamma,\mu,\nu,p,q,s$ [1709.03909]. These results provide a complete off-diagonal theory for a large class of Bergman-type operators on tubes over symmetric cones [1709.03909].

Recent work has emphasized product domains and multifunctional operators. Sehba and coauthors introduced analytic mixed-norm spaces on products $T_\Omega^m$, operators such as
$$
T_{\vec\beta}f(\vec z)=\int_{T_\Omega^m}
\frac{f(w_1,\dots,w_m)\prod_{j=1}^m\Delta(v_j)^{\beta_j-n/r}\,dv(w_j)}
{\prod_{j=1}^m \Delta^{\beta_j+n/r}\!\left(\frac{z_j-\overline w_j}{i}\right)},
$$
and proved boundedness from measurable mixed-norm spaces into analytic ones under large-parameter hypotheses [2509.22024]. The same paper formulates new sharp decomposition theorems in $A_\alpha^1(T_\Omega)$, based on special integral representations [2509.22024].

The analytic literature also contains open problems. Among them are extending boundedness of Bergman projections on mixed-norm spaces for all $p,q\ge1$, enlarging the $L^p$ boundedness range of Bergman projectors on higher-rank cones, developing Herz-type analytic spaces on tubes and product tubes, and transferring sharp decomposition theorems from $A_\alpha^1$ to the full $A_\alpha^p$ and $A_\alpha^{p,q}$ scales [1703.07859, 2509.22024, 2511.10270].

## 7. Representative examples and broader geometric significance

Two examples recur throughout the theory.

| Example | Cone and determinant | Tube-domain interpretation |
|---|---|---|
| Lorentz cone | $\Lambda_n=\{y\in\mathbb R^n:y_1^2-y_2^2-\cdots-y_n^2>0,\ y_1>0\}$, $\Delta(y)=y_1^2-\cdots-y_n^2$ | Tube-type domain; rank $r=2$ [2509.22024] |
| Positive-definite symmetric matrices | $\Omega=\mathrm{SPD}_n$ or $\mathrm{Sym}(r,\mathbb R)_{++}$, $\Delta(y)=\det(y)$ | Siegel upper half-space of symmetric complex matrices [2006.06499, 2509.22024] |

For the Lorentz cone, the associated tube domain is the classical tube-type realization related to the spin factor, and the bounded realization is the ball of the corresponding spin factor [2006.06499]. For positive-definite symmetric matrices, the tube is the Siegel upper half-space of symmetric complex matrices and is Cayley-equivalent to the bounded Cartan domain of type III [2006.06499].

The theory also extends beyond finite dimensions. Infinite-dimensional spin factors $H\oplus\mathbb R$ with JB-norm $\|a\oplus\alpha\|_s=\|a\|+|\alpha|$ have cone
$$
\Omega=\{\alpha>\|a\|\},
$$
and their tube domains are bounded symmetric via the JB-framework [2006.06499]. Direct sums of finite-dimensional formally real Jordan algebras and spin factors likewise produce normal, linearly homogeneous Finsler symmetric cones [2006.06499].

Tubular domains over symmetric cones also appear in algebraic geometry. Catanese and Franciosi’s semispecial and slope-zero tensors characterize varieties whose universal covers are the polydisk or a bounded symmetric domain of tube type [1011.6544]. In that context, factorization of a holonomy-invariant tangential polynomial into powers of generic norms determines the tube-type bounded symmetric domain via the ranks and dimensions of the corresponding Jordan factors [1011.6544]. This suggests a precise bridge between intrinsic tensorial data on compact varieties and the Jordan-theoretic classification of tube domains.

More broadly, the subject unifies several themes. From Jordan theory it inherits cones of squares, determinants, and structure groups; from symmetric-space theory it inherits point symmetries, Cayley transforms, and automorphism groups; from several complex variables it inherits Bergman kernels, invariant metrics, and Carleson embeddings; and from operator theory it inherits Toeplitz, Hankel, Cesàro, and Schatten-class questions. The common structural principle is that the determinant $\Delta$, the rank parameter $n/r$, and the symmetry of the cone control both geometry and analysis on $T(\Omega)$ [2006.06499, 2509.22024].

Source: https://www.emergentmind.com/topics/tubular-domains-over-symmetric-cones