Open Access

Semi-periodic solutions of difference and differential equations

Boundary Value Problems20122012:141

DOI: 10.1186/1687-2770-2012-141

Received: 1 October 2012

Accepted: 7 November 2012

Published: 28 November 2012

Abstract

The spaces of semi-periodic sequences and functions are examined in the relationship to the closely related notions of almost-periodicity, quasi-periodicity and periodicity. Besides the main theorems, several illustrative examples of this type are supplied. As an application, the existence and uniqueness results are formulated for semi-periodic solutions of quasi-linear difference and differential equations.

MSC:34C15, 34C27, 34K14, 39A10, 42A16, 42A75.

Keywords

semi-periodic sequences semi-periodic functions semi-periodic solutions difference equations differential equations

Introduction

In [1], it is observed that although the set of periodic sequences forms a linear space, its uniform closure is not the space of almost-periodic sequences but of semi-periodic sequences. In fact, the space of semi-periodic sequences was shown there to be Banach.

The whole Sections I.6, I.7 in [2] and Sections II.4, II.5 in [3] are devoted to semi-periodic continuous functions, called there limit periodic functions (cf. also [[4], p.129]). This class was shown there to be identical with the one of uniformly almost-periodic functions with one-term -base and, in case of integral one-term base, it reduces to the one of purely periodic functions. For some more references concerning limit periodic functions, see, e.g., [5, 6]. In fact, limit periodic functions were already considered by Bohr in 1925, as pointed out in [[3], p.113].

In the following section, we define analogously to [1] the class of semi-periodic continuous functions (with values in a Banach space) and show that it is the same as the class of limit periodic functions considered in [2, 3] (see Theorem 1 below). Let us note that many different notions with the same name (i.e., semi-periodic), like functions satisfying Floquet boundary conditions (see, e.g., [7, 8]) or those describing Bloch waves (see, e.g., [7], and the references therein), exist in the literature (cf. also [9, 10]).

Hence, after giving a definition of semi-periodic functions, which is analogous to [1], we prove that the uniform closure of the set of periodic functions is again the one of semi-periodic functions. Unlike in the discrete case, the space of semi-periodic functions is, however, not linear and so not Banach. In order to clarify transparently the position of semi-periodic sequences and functions in the hierarchy of closely related spaces, we decided to illustrate it by means of Venn’s diagrams. Thus, the spaces of almost-periodic, semi-periodic, quasi-periodic and periodic functions and sequences and some of their sums (in the continuous case) are compared in this way. For this, the semi-periodicity is considered by means of the Fourier-Bohr coefficients.

There are even more general interesting classes of almost-periodic functions (for their hierarchy, see, e.g., [11, 12]), but for our needs here only those which are uniformly (Bohr) a.p. will be taken into account. It is well known that uniformly continuous Stepanov a.p. functions are Bohr a.p. (see, e.g., [4, 11]). Another nontraditional characterization of Bohr almost-periodicity was recently done in [13], namely that Stepanov a.p. functions with Stepanov a.p. derivatives are also Bohr a.p.

In order to make applications to difference and differential equations, we still need to define the notion of uniform semi-periodicity and prove that the associated Nemystkii operators map the set of semi-periodic sequences into themselves. This is unfortunately not true in the case of functions. On this basis, we finally give two examples about the existence of semi-periodic solutions in the form of theorems, both in the discrete and in the continuous cases. Although many various sorts of periodic-type solutions were investigated (for their panorama, see [7]), as far as we know, semi-periodic solutions in the sense of definitions below of difference or differential equations have been only considered in [14] and in a certain sense also in [5]. Nevertheless, as pointed out in [14], Johnson [15] and Millionshchikov [16] have already given examples of limit periodic differential equations which admit almost automorphic solutions, but not limit periodic ones.

Before passing to semi-periodic functions in the next section, it will be convenient to mention some facts about semi-periodic sequences.

Hence, denoting as usually by ( N ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq1_HTML.gif the set of (positive) integers and letting E https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq2_HTML.gif to be a Banach space endowed with the norm | | E https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq3_HTML.gif, let us recall the definition of semi-periodic sequences (cf. [1]).

Definition 1 A sequence x ̲ E Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq4_HTML.gif is called semi-periodic (s.p.) if
ε > 0 , T N , n Z , k Z , | x k + n T x k | E ε . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equa_HTML.gif

One can readily check that Definition 1 can be regarded as a discrete version of Definition 2 below for semi-periodic functions. Similarly, the definition of quasi-periodic (q.p.) sequences can be regarded as a discretized (i.e., restricted to ) version of the one for quasi-periodic functions recalled below. A q.p. extending function has the Fourier-Bohr expansion with Mod ( ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq5_HTML.gif to be finitely generated which is also true for q.p. sequences. For more properties and details concerning q.p. functions, see, e.g., [17].

In this light, since the analogy of Theorem 2 below holds for sequences (see Remark 4) and since the discrete (i.e., restricted to ) analogies of Examples 1-3 below can be constructed, one can illustrate the relationship of these classes by means of Venn’s diagram in Figure 1. For more properties about s.p. sequences, see, e.g., [1, 18, 19].
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Fig1_HTML.jpg
Figure 1

Venn’s diagram: discrete case.

On the other hand, the situation in Figure 1 is much simpler than in Figure 2 for continuous functions, because under the restriction to , the sum of (semi-)periodic sequences remains (semi-)periodic while Stepanov almost-periodic (a.p.) sequences were shown in [20] to coincide with Bohr a.p. sequences.
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Fig2_HTML.jpg
Figure 2

Venn’s diagram: continuous case.

Continuous semi-periodic functions

Let C T 0 ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq6_HTML.gif be the set of continuous T-periodic functions,
Per ( R , E ) : = T > 0 C T 0 ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equb_HTML.gif

be the set of periodic functions and BC 0 ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq7_HTML.gif be the set of continuous bounded functions. The last one is a Banach space with the uniform norm (written https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq8_HTML.gif).

Definition 2 A continuous function f C 0 ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq9_HTML.gif is said to be semi-periodic (s.p.) if
ε > 0 , T > 0 , n Z , t R , | f ( t + n T ) f ( t ) | E ε . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equc_HTML.gif

Such a T will be called an ε-semi-period of f.

Let S ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq10_HTML.gif denote the set of semi-periodic functions.

It is easy to see from the definition that every continuous periodic function is semi-periodic. Moreover, if f is semi-periodic, then f is uniformly (Bohr) almost-periodic (i.e., f AP 0 ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq11_HTML.gif), and so it is bounded. Thus, we can rewrite Definition 2 as follows.

Definition 3 A (bounded) continuous function f C 0 ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq9_HTML.gif is said to be semi-periodic (s.p.) if
ε > 0 , T > 0 , n Z , f ( + n T ) f ( ) ε . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equd_HTML.gif
We have
Per ( R , E ) S ( R , E ) AP 0 ( R , E ) BC 0 ( R , E ) . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Eque_HTML.gif
From this, we can consider S ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq10_HTML.gif as a metric space, when using
d ( f , g ) : = sup t R | f ( t ) g ( t ) | E . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equf_HTML.gif

As we will see later, S ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq10_HTML.gif is not a linear space, but S ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq10_HTML.gif is a complete metric space.

Lemma 1 Let f S ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq12_HTML.gif, ε > 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq13_HTML.gif and T ε https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq14_HTML.gif be an ε-semi-period of f. Then there exists a continuous T ε https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq14_HTML.gif-periodic function φ s.t.
f φ ε . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equg_HTML.gif
Proof Consider a T ε https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq14_HTML.gif-periodic function ψ such that its restriction to [ 0 ; T ε ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq15_HTML.gif is the same as the one of f. For each x R https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq16_HTML.gif, we can write x = t + n T ε https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq17_HTML.gif with n Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq18_HTML.gif and t [ 0 ; T ε ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq19_HTML.gif. Thus, we get
| f ( x ) ψ ( x ) | E = | f ( t + n T ε ) ψ ( t + n T ε ) | E = | f ( t + n T ε ) ψ ( t ) | E = | f ( t + n T ε ) f ( t ) | E ε . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equh_HTML.gif
Since ψ is not necessarily continuous, consider still τ ( 0 ; T ε ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq20_HTML.gif such that, for any t [ T ε τ , T ε ] https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq21_HTML.gif, | f ( t ) f ( T ε ) | E ε https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq22_HTML.gif. Define a T ε https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq14_HTML.gif-periodic continuous function φ which is equal to ψ on [ 0 , T ε τ ] https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq23_HTML.gif and which is linear on [ T ε τ , T ε ] https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq24_HTML.gif. For t [ T ε τ , T ε ] https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq21_HTML.gif, we obtain
| f ( t ) φ ( t ) | E T ε t τ | f ( T ε τ ) f ( t ) | E + t ( T ε τ ) τ | f ( T ε ) f ( t ) | E 2 ε , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equi_HTML.gif
and subsequently
sup x R | φ ( x ) f ( x ) | E 2 ε . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equj_HTML.gif

 □

Remark 1 For E = R https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq25_HTML.gif, unlike for semi-periodic functions in the sense of Definition 2 or Definition 3, in fact the same lemma was already proved in [[3], pp.114-115], but for limit periodic functions. As already pointed out in the foregoing section, these classes will be shown to coincide by Theorem 1 below, whose proof is just based on Lemma 1.

We are ready to give the first theorem.

Theorem 1 S ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq10_HTML.gif is the closure of Per ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq26_HTML.gif in the sup-norm.

Proof Assume firstly that f is s.p. Taking in Lemma 1 ε n = 1 / n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq27_HTML.gif, we obtain a sequence of periodic functions ( φ n ) n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq28_HTML.gif s.t. f φ n ε n 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq29_HTML.gif.

Reversely, assume that f is in the closure of the set of continuous T-periodic functions. Then, for any ε > 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq13_HTML.gif, we can find a periodic φ s.t. | f ( t ) φ ( t ) | E ε https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq30_HTML.gif. Let T be its period. Then, for any t R https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq31_HTML.gif,
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equk_HTML.gif

 □

Remark 2 In view of Theorem 1, one can now also define a semi-periodic function, equivalently w.r.t. Definition 2 and Definition 3, as the uniform limit of a uniformly convergent sequence of continuous purely periodic functions. This was so done, e.g., in [2, 3, 5, 6, 14].

In the following proposition, we look for the link between s.p. sequences and functions. Given a sequence x ̲ = ( x t ) t Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq32_HTML.gif, we set f x ̲ : R E https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq33_HTML.gif, the function s.t. its restriction to is x ̲ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq34_HTML.gif and which is linear on each [ k , k + 1 ] https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq35_HTML.gif, k Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq36_HTML.gif, i.e.,
t Z , f x ̲ ( u ) : = { u } x t + 1 + ( 1 { u } ) x t , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equl_HTML.gif

where { u } https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq37_HTML.gif is the fractional part of u, i.e., { u } [ 0 , 1 ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq38_HTML.gif and u { u } Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq39_HTML.gif.

Proposition 1 Let x ̲ E Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq4_HTML.gif. All the following statements are equivalent:
  1. 1.

    f x ̲ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq40_HTML.gif is s.p. with a semi-period in ,

     
  2. 2.

    there exists a s.p. function with a semi-period in whose restriction to is x ̲ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq34_HTML.gif,

     
  3. 3.

    x ̲ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq34_HTML.gif is s.p.

     
Proof For (1)  (2), take f x ̲ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq40_HTML.gif in (2). For (2)  (3), take T as an ε-semi-period for the function f in (2). Then we have
t Z , | x t + T x t | E = | f ( t + T ) f ( t ) | E f ( + T ) f ε . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equm_HTML.gif
For (3)  (1), given T as an ε-semi-period of x ̲ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq34_HTML.gif, we have for all t Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq41_HTML.gif
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equn_HTML.gif

 □

Let us now consider the Fourier expansion of a semi-periodic function. Recall that every a.p. function has the Fourier-Bohr expansion,
f ( t ) j = 1 a λ j ( f ) e i λ j t , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equo_HTML.gif
where
a λ ( f ) : = M { f ( t ) e i λ t } , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equp_HTML.gif
and
M { g } : = lim l ( 2 l ) 1 l l g ( t ) d t https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equq_HTML.gif

is the mean operator (see, e.g., [3, 4, 11]). It follows from the above formula that f a λ ( f ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq42_HTML.gif is 1-Lipschizian (and so it is continuous) from AP 0 ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq43_HTML.gif to E https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq2_HTML.gif.

Set Λ ( f ) : = { λ , a λ ( f ) 0 } https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq44_HTML.gif and denote by Mod ( f ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq45_HTML.gif the -modulus generated by Λ ( f ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq46_HTML.gif. Recall that an a.p. function is quasi-periodic (q.p.) if Mod ( f ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq45_HTML.gif has a finite -basis, and that T is a period of f if and only if Λ ( f ) 2 π T Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq47_HTML.gif (see, e.g., [4, 17]).

Proposition 2 (for E = R https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq25_HTML.gif, cf. [[2], p.32])

Set Q : = { r n , n N } https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq48_HTML.gif and consider
f ( t ) : = λ θ Q a λ ( f ) e i λ t , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equr_HTML.gif
for a fixed θ > 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq49_HTML.gif, where
λ θ Q | a λ ( f ) | E < + . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equs_HTML.gif

Then f is s.p.

Proof Consider
f N ( t ) = n = 1 N a r n θ ( f ) e i r n θ t . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equt_HTML.gif
Clearly, if r n = p n q n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq50_HTML.gif, then 2 π q n p n θ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq51_HTML.gif is a period of the n th term. The same is obviously true for 2 π q n θ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq52_HTML.gif. Thus, 2 π q 1 q N θ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq53_HTML.gif is a period of f N https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq54_HTML.gif which is so periodic. Moreover,
f f N n N + 1 | a r n θ ( f ) | E 0 , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equu_HTML.gif

which already proves that f is s.p. □

The following result is also, at least for E = R https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq25_HTML.gif, well known (see, e.g., [14], [[3], pp.118-119], and the references therein).

Lemma 2 If f S ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq55_HTML.gif, then there exists θ > 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq49_HTML.gif s.t.
Λ ( f ) θ Q . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equv_HTML.gif

Proof Let us consider λ and μ s.t. a λ ( f ) 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq56_HTML.gif and a μ ( f ) 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq57_HTML.gif and a sequence of periodic functions ( f n ) n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq58_HTML.gif s.t. f n f https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq59_HTML.gif, uniformly. It follows from the continuity that, for sufficiently large N, a λ ( f N ) 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq60_HTML.gif and a μ ( f N ) 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq61_HTML.gif, but since f N https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq54_HTML.gif is periodic, it follows that λ / μ Q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq62_HTML.gif. □

Remark 3
  1. 1.

    This proof also demonstrates that, for a sufficiently large n, the period T n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq63_HTML.gif of f n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq64_HTML.gif satisfies T n θ 2 π Q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq65_HTML.gif.

     
  2. 2.

    It indicates that S ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq66_HTML.gif is not a linear space. For instance, a simple q.p. function t cos ( t ) + cos ( t 2 ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq67_HTML.gif is not s.p. although it is a sum of two s.p. functions. On the other hand, the sum of two a.p. functions is trivially a.p.

     
Example 1 On the basis of Proposition 2 and Lemma 2, we can easily give the following example of a purely s.p. (i.e., not periodic) function:
f ( t ) = n 1 e i t / n n 2 . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equw_HTML.gif
Moreover, one can readily check that the function f can be obtained as a uniform limit of the sequence ( f N ) N https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq68_HTML.gif, where f N https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq54_HTML.gif is a continuous 2 π N ! https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq69_HTML.gif-periodic function,
f N ( t ) = n = 1 N e i t / n n 2 . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equx_HTML.gif
Theorem 2 Every s.p. function which is also q.p. is in fact periodic:
S ( R , E ) QP 0 ( R , E ) = Per ( R , E ) . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equy_HTML.gif
Proof Let f S ( R , E ) QP 0 ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq70_HTML.gif. Since f is q.p., we can find ω 1 , , ω m https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq71_HTML.gif such that
Λ ( f ) Z ω 1 + + Z ω m . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equz_HTML.gif
Set G 1 : = Z ω 1 + + Z ω m https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq72_HTML.gif. G 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq73_HTML.gif is an additive subgroup of . Since f S ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq12_HTML.gif, we can find θ > 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq49_HTML.gif s.t. Λ ( f ) θ Q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq74_HTML.gif. Set G 2 : = θ Q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq75_HTML.gif. G 2 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq76_HTML.gif is another additive subgroup of , so G = G 1 G 2 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq77_HTML.gif is a subgroup of G 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq73_HTML.gif which contains Λ ( f ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq46_HTML.gif. Since G is a subgroup of G 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq73_HTML.gif, there exist p { 1 , , m } https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq78_HTML.gif and positive -independent real numbers ζ 1 , , ζ p https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq79_HTML.gif s.t.
G = Z ζ 1 + + Z ζ p . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equaa_HTML.gif

Let us show that p = 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq80_HTML.gif. Once we have it, we can conclude that Λ ( f ) ζ 1 Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq81_HTML.gif which proves that 2 π ζ 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq82_HTML.gif is a period of f. Since, for each i, ζ i G G 2 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq83_HTML.gif, we know that, for each i, we can find q i Q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq84_HTML.gif s.t. ζ i = q i θ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq85_HTML.gif. This proves that ζ i / ζ j Q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq86_HTML.gif, for i j https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq87_HTML.gif, which is impossible. □

Remark 4 In view of Proposition 1 and its analogy for q.p. sequences mentioned in the foregoing section, a discrete (i.e., restricted to ) analogy of Theorem 2 holds for sequences.

Example 2 As an example of a function which is almost-periodic (a.p.) but neither quasi-periodic nor a sum of semi-periodic functions, consider
f ( t ) = n 1 e i σ n t n 2 , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equab_HTML.gif
where the σ k https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq88_HTML.gif’s are constructed by induction, say for all k,
σ k + 1 σ 1 Q + + σ k Q . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equac_HTML.gif
We will prove that we cannot find a finite set of numbers θ 1 , , θ q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq89_HTML.gif s.t.
Λ ( f ) θ 1 Q + + θ q Q . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equad_HTML.gif
Firstly, assume this has already been proved. Then if f is a sum of semi-periodic functions  f j https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq90_HTML.gif, say f = j = 1 q f j https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq91_HTML.gif, we could find, according to Lemma 2, for each j a θ j https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq92_HTML.gif s.t. Λ ( f j ) θ j Q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq93_HTML.gif. This implies that
Λ ( f ) Λ ( f 1 ) Λ ( f q ) ( θ 1 Q ) + + ( θ q Q ) , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equae_HTML.gif
which is not true. If f were quasi-periodic, we could find θ 1 , , θ q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq89_HTML.gif s.t.
Λ ( f ) θ 1 Z + + θ q Z θ 1 Q + + θ q Q , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equaf_HTML.gif
which is again wrong. Now, we can make the first part of the proof. So, let us assume
Λ ( f ) θ 1 Q + + θ q Q . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equag_HTML.gif
We have Λ ( f ) = { σ i , i 1 } https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq94_HTML.gif. Thus, for any i 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq95_HTML.gif, we can find ( a i 1 , , a i q ) Q q { 0 } https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq96_HTML.gif s.t.
σ i = j = 1 q a i j θ j . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equah_HTML.gif
Let us now consider the square matrix
A = ( a i j ) 1 i , j q . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equai_HTML.gif

If it is invertible, we can express θ 1 , , θ q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq89_HTML.gif linearly (with rational coefficients) depending on σ 1 , , σ q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq97_HTML.gif. This proves that σ q + 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq98_HTML.gif should be a (rational) linear combination of σ 1 , , σ q https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq97_HTML.gif, which is not true.

Assuming that the matrix is singular, its rows are linearly dependent. So, we can find ( μ 1 , , μ q ) Q q { 0 } https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq99_HTML.gif s.t. i μ i a i j = 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq100_HTML.gif, for each j. Multiplying it by θ j https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq92_HTML.gif and then summing over j, we obtain i μ i σ i = 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq101_HTML.gif which is not possible.

Example 3 As an example of a function which is quasi-periodic (q.p.) but not a sum of periodic functions, consider
f ( t ) = n 1 e i t ( 1 + n 2 ) n 2 . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equaj_HTML.gif
Here Λ ( f ) = { 1 + n 2 , n N } https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq102_HTML.gif, thus Mod ( f ) = Z + 2 Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq103_HTML.gif, i.e., f is q.p. Assume that f is a sum of a finite number of periodic functions. Let T 1 , , T k https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq104_HTML.gif be the periods. According to [21], we have
Δ ( T 1 , , T k ) f = 0 , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equak_HTML.gif
where
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equal_HTML.gif
An easy calculation yields
a λ ( Δ ( T 1 , , T k ) f ) = a λ ( f ) j = 1 k ( e i λ T j 1 ) , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equam_HTML.gif
by which
n N , j { 1 , , k } , ( 1 + n 2 ) T j 2 π Z . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equan_HTML.gif
Since is infinite, we can find two different integers m, n with the same T j https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq105_HTML.gif. Thus, there exist two integers k m https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq106_HTML.gif, k n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq107_HTML.gif s.t.
1 + n 2 k n = 2 π T j = 1 + m 2 k m . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equao_HTML.gif
This implies that k m k n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq108_HTML.gif, and we obtain
( k m k n ) = ( m k n n k m ) 2 , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equap_HTML.gif

which is not possible.

Remark 5 We know (see, e.g., [4, 11]) that every almost-periodic (a.p.) f is a uniform limit of a sequence of a finite sum of periodic functions ( f n ) n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq58_HTML.gif. Writing
f = f 0 + n ( f n + 1 f n ) , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equaq_HTML.gif

we can see that every a.p. function can be expressed as a series of periodic functions. Reversely, a uniformly convergent series of periodic functions is a.p.

Summing up the above observations, we can present in Figure 2 Venn’s diagram for continuous functions under our investigation. The classes of almost-periodic, semi-periodic and quasi-periodic functions are in circles, while sums of semi-periodic functions are in the ellipse. Sums of periodic functions are in the intersection of the classes of quasi-periodic functions and sums of semi-periodic functions. In fact, one can check by similar arguments as in the proof of Theorem 2 that a sum of periodic functions is exactly the sum of semi-periodic functions which is quasi-periodic. Periodic functions are, according to Theorem 2, at the same time semi-periodic and quasi-periodic. Purely semi-periodic functions are in the grey strip.

Now, consider the primitives of s.p. functions.

Lemma 3 Assume that f is a.p. and consider F ( t ) : = 0 t f ( s ) d s https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq109_HTML.gif. Assume that there exists φ AP 0 ( R , E ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq110_HTML.gif and a E https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq111_HTML.gif s.t.,
t R , F ( t ) = φ ( t ) + a t . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equar_HTML.gif

Then a = M { f } https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq112_HTML.gif.

Indeed, φ is necessarily differentiable, and integrating the equality f = φ + a https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq113_HTML.gif, we obtain
a = 1 2 l l l f ( s ) d s + O ( 1 l ) , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equas_HTML.gif

because φ is bounded. This already proves Lemma 3. It is well known that M { f } = 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq114_HTML.gif is a necessary and sufficient condition for F to be periodic, provided f is so. It is, however, not sufficient in the case of a.p. functions. For more details, see, e.g., [22]. Despite the approximation by periodic functions, it is also not sufficient in the case of s.p. functions, as demonstrated by the following example.

Example 4 Let us consider the s.p. function
f ( t ) = n 1 cos ( t / n 2 ) n 2 . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equat_HTML.gif
We have a normal convergence, so the series exists and defines a s.p. function for which M { f } = 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq114_HTML.gif. A formal candidate to be its primitive is
F ( t ) = n 1 sin ( t / n 2 ) . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equau_HTML.gif

We have a uniform convergence on each compact set, because | sin ( u ) | | u | https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq115_HTML.gif. Thus, this series also exists and defines a primitive of f. If F were s.p., it should be a.p. which is obviously not true, because the Parseval equality does not apply.

Uniformly semi-periodic functions with respect to a parameter

Definition 4 Let f : R × M R k https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq116_HTML.gif, where M is a subset of R n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq117_HTML.gif. We say that f is uniformly semi-periodic (u.s.p.) if for any compact set K M R n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq118_HTML.gif, we have
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equav_HTML.gif

Since such a function is u.a.p., we know that given a compact subset K of M, f is bounded and uniformly continuous on R × K https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq119_HTML.gif.

Proposition 3 Any u.s.p. function is a uniform limit, on each R × K https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq119_HTML.gif, of a sequence of continuous functions which are periodic w.r.t. their first variables.

Proof Let T be given by the definition and consider a T-periodic function φ ( , α ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq120_HTML.gif such that its restriction to [ 0 ; T ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq121_HTML.gif is the same as the one of f ( , α ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq122_HTML.gif. For each x R https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq16_HTML.gif, we can write x = t + n T https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq123_HTML.gif with n Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq18_HTML.gif and t [ 0 ; T ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq124_HTML.gif. Thus, we get
| f ( x , α ) φ ( x , α ) | R k = | f ( t + n T , α ) φ ( t + n T , α ) | R k = | f ( t + n T , α ) φ ( t , α ) | R k = | f ( t + n T , α ) f ( t , α ) | R k ε , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equaw_HTML.gif
uniformly w.r.t. α K https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq125_HTML.gif. Since φ is not necessarily continuous, consider still τ ( 0 ; T ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq126_HTML.gif such that, for any t [ T τ , T ] https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq127_HTML.gif and any α K https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq125_HTML.gif, | f ( t , α ) f ( T , α ) | E ε https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq128_HTML.gif. This is possible, because K is compact. Define a T-periodic continuous function ψ ( , α ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq129_HTML.gif which is equal to φ ( , α ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq120_HTML.gif on [ 0 , T τ ] https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq130_HTML.gif and which is linear on [ T τ , T ] https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq131_HTML.gif. For t [ T τ , T ] https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq127_HTML.gif, we obtain
| f ( t , α ) ψ ( t , α ) | R k T t τ | f ( T τ , α ) f ( t , α ) | R k + t ( T τ ) τ | f ( T , α ) f ( t , α ) | R k 2 ε , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equax_HTML.gif
and subsequently
sup ( x , α ) R × K | ψ ( x , α ) f ( x , α ) | R k 2 ε . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equay_HTML.gif

 □

Remark 6 Assume that f is L-Lipschitzian w.r.t. its second variable. It follows from the proof that so is φ, from which we can deduce the same for ψ. So, a u.s.p. function Lipschitzian w.r.t. its second variable can be approximated uniformly on each R × K https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq119_HTML.gif (K compact) by a sequence of functions which are periodic w.r.t. their first variables and Lipschitzian (with the same constant L) w.r.t. their second variables.

Remark 7 It is possible to define the same for the discrete case and to obtain analogous results. This will be omitted here, because the proofs are quite similar.

Concerning the Nemytskii operator, in the continuous case, it is not true that if f is u.s.p. and ϕ is s.p., then t f ( t , ϕ ( t ) ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq132_HTML.gif is s.p. As an example, take f ( t , x ) = sin ( t ) + x https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq133_HTML.gif and ϕ ( t ) = sin ( π t ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq134_HTML.gif. On the other hand, it is true in the discrete case.

Proposition 4 Assume that f : Z × M R p https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq135_HTML.gif is s.p. and that x ̲ = ( x t ) t https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq136_HTML.gif is s.p. with the range in M R n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq137_HTML.gif. Then the sequence ( f ( t , x t ) ) t Z https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq138_HTML.gif is s.p.

Proof Set K = { x t , t Z } ¯ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq139_HTML.gif. Since x ̲ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq34_HTML.gif is a.p., K is a compact subset of M. So, given ε > 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq13_HTML.gif, we can find η > 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq140_HTML.gif s.t.
sup t Z , | x y | η | f ( t , x ) f ( t , y ) | R p ε . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equaz_HTML.gif
Set η : = min { η , ε } https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq141_HTML.gif. We know that we can find two integers T 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq142_HTML.gif, T 2 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq143_HTML.gif s.t.
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equba_HTML.gif
Let T be a common multiplier of T 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq142_HTML.gif and T 2 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq143_HTML.gif (for instance, T = T 1 T 2 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq144_HTML.gif). The last inequalities remain true, when replacing every T i https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq145_HTML.gif by T. Thus, for any ( t , n ) Z 2 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq146_HTML.gif,
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbb_HTML.gif

 □

For an alternative proof, one can employ the approximation by periodic sequences.

Semi-periodic solutions of difference equations

In this section, we are interested in semi-periodic solutions of the difference equation in  R p https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq147_HTML.gif,
x t + 1 + A x t = f ( t , x t ) , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equ1_HTML.gif
(1)

where A is a real square p × p https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq148_HTML.gif matrix.

Theorem 3 Assuming that A has no eigenvalues with modulus one and that f is u.s.p. and Lipschitzian w.r.t. the second variable with a sufficiently small constant, there exists a unique semi-periodic solution for the difference equation (1).

Proof We know (see, e.g., Proposition 2.2 in [23]) that, for each a.p. sequence ( b t ) t https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq149_HTML.gif with values in R p https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq147_HTML.gif, there exists a unique a.p. solution to
x t + 1 + A x t = b t . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equ2_HTML.gif
(2)

Denoting by AP ( Z , R p ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq150_HTML.gif the Banach space of a.p. sequences (cf. [23]), the linear operator T : AP ( Z , R p ) AP ( Z , R p ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq151_HTML.gif, determined by the left-hand side of (2), is obviously invertible. Since T is continuous satisfying T 1 + A https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq152_HTML.gif, we know from the well-known Banach theorem that T 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq153_HTML.gif must be continuous as well.

Now, consider a s.p. sequence ( q t ) t https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq154_HTML.gif with values in R p https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq147_HTML.gif. We are firstly interested in the a.p. solution to the equation
x t + 1 + A x t = f ( t , q t ) . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equ3_HTML.gif
(3)
By the hypothesis imposed on f and in view of Proposition 4, ( f ( t , q t ) ) t https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq155_HTML.gif is s.p. Therefore, there exists a unique a.p. solution of (3) (see again Proposition 2.2 in [23]). We can now consider T 1 ( ( f ( t , q t ) ) t ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq156_HTML.gif. Since T 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq153_HTML.gif maps the space of periodic sequences into itself, by the unique solvability of (3) in AP ( Z , R n ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq157_HTML.gif and by the continuity of T 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq153_HTML.gif, the mapping
T : ( q t ) t T 1 ( ( f ( t , q t ) ) t ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbc_HTML.gif

maps S ( Z , R p ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq158_HTML.gif into itself. Denote by L the Lipschitz constant to all f ( t , ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq159_HTML.gif. It is easy to see that T 1 L https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq160_HTML.gif is a Lipschitz constant for T https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq161_HTML.gif.

Assuming that L < 1 / T 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq162_HTML.gif, the mapping T is a contraction in the Banach space S ( Z , R p ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq158_HTML.gif. So it has a unique fixed point representing the desired s.p. solution of (1). □

Remark 8 Using a triangular form of −A (like Jordan’s one) (see, e.g., [[4], Proposition 6.14 and Remark 6.26]), it is possible to compute explicitly a constant c s.t. T 1 c https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq163_HTML.gif. For such a constant, it is sufficient to assume L < 1 / c https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq164_HTML.gif in order to justify Theorem 3.

Semi-periodic solutions of differential equations

Let us consider the equation
x + A x = f ( t , x ) . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equ4_HTML.gif
(4)
We assume that a real square k × k https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq165_HTML.gif matrix A has an exponential dichotomy property, i.e., that there exist a projection matrix P ( P = P 2 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq166_HTML.gif) and constants C > 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq167_HTML.gif, λ > 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq168_HTML.gif, such that
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbd_HTML.gif

where X is the fundamental matrix of x + A x = 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq169_HTML.gif satisfying X ( 0 ) = I https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq170_HTML.gif, i.e., the unit matrix (see, e.g., [[8], Chapter III.5]). Furthermore, let f : R × R k R k https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq171_HTML.gif be u.s.p. with respect to the variable x.

Setting
C ( A ) : = sup t R | R | G ( t s ) | d s | sup t R R C e λ | t s | d s = 2 C λ , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Eqube_HTML.gif
where
G ( t , s ) : = { e A ( t s ) P , for  t > s , e A ( t s ) P + , for  t < s https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbf_HTML.gif

is the Green function associated to A, and P https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq172_HTML.gif, P + https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq173_HTML.gif stand for the corresponding spectral projections on the invariant subspaces of A, we can formulate the following theorem.

Theorem 4 Assume still that f is L-Lipschitzian w.r.t. the second variable with L < ( λ / 2 C ) 1 / C ( A ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq174_HTML.gif. Then there exists a unique semi-periodic solution of the equation (4).

Proof Let ( f n ) n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq58_HTML.gif be a sequence of periodic functions w.r.t. their first variables s.t. f n f https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq175_HTML.gif, uniformly. We can assume without any loss of generality (see Remark 3) that each f n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq64_HTML.gif is L-Lipschitzian w.r.t. its second variable. Let x n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq176_HTML.gif be the unique bounded (in fact, periodic) solution of the equation
x + A x = f n ( t , x ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbg_HTML.gif

and x ¯ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq177_HTML.gif be the unique bounded solution of (4). Such solutions exist; for more details, see, e.g., [[8], Chapter III.5].

It will be sufficient to show that x n x ¯ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq178_HTML.gif, uniformly.

We have the integral representations (see again, e.g., [[8], Chapter III.5])
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbh_HTML.gif
It can be easily checked that, in view of uniqueness of bounded solutions, the periods T n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq63_HTML.gif of f n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq64_HTML.gif are also periods of x n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq176_HTML.gif. It holds
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbi_HTML.gif
Now, let us prove that there exists a uniform estimate to all x n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq176_HTML.gif. We have
| x n ( t ) | R k R | G ( t s ) | | f n ( s , x n ( s ) ) | R k d s , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbj_HTML.gif
and
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbk_HTML.gif
Thus,
x n C ( A ) ( f ( , 0 ) + f n ( , 0 ) f ( , 0 ) + L x n ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbl_HTML.gif
and, according to C ( A ) L < 1 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq179_HTML.gif, still
x n C ( A ) 1 C ( A ) L ( f ( , 0 ) + f n ( , 0 ) f ( , 0 ) ) = : R , https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbm_HTML.gif
where R is the desired bound. Putting K = B ( 0 , R ) ¯ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq180_HTML.gif, we arrive at
https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbn_HTML.gif
where ε n = sup R × K f n f 0 https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq181_HTML.gif. Thus, we finally get
x n x ¯ C ( A ) 1 C ( A ) L ε n 0 . https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_Equbo_HTML.gif

Since x n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq176_HTML.gif are T n https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq63_HTML.gif-periodic, we conclude that x ¯ https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq177_HTML.gif is semi-periodic. □

Concluding remarks

Remark 9 Because of the right-hand side f ( t , x ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq182_HTML.gif in (4), even in the scalar case, Theorem 4 cannot be deduced from the results in [14], where the scalar equation x + g ( x ) = f ( t ) https://static-content.springer.com/image/art%3A10.1186%2F1687-2770-2012-141/MediaObjects/13661_2012_Article_246_IEq183_HTML.gif was considered.

Remark 10 Since Theorem 3 and Theorem 4 represent only illustrative examples, the obtained existence and uniqueness criteria were tendentiously very simple. More sophisticated situations will be considered by ourselves elsewhere.

Remark 11 Analogously as in [24, 25], where almost-periodic solutions were under consideration, it would be interesting to obtain similar results concerning semi-periodic solutions of monotone systems or those treated by means of variational methods.

Declarations

Acknowledgements

The first author was supported by the project A-Math-Net Applied Mathematics Knowledge Transfer Network No CZ.1.07/2.4.00/17.0100.

Authors’ Affiliations

(1)
Department of Mathematical Analysis, Faculty of Science, Palacký University
(2)
Centre PMF, Laboratoire SAMM, Université Paris I Panthéon-Sorbonne

References

  1. Berg JD, Wilansky A: Periodic, almost-periodic, and semiperiodic sequences. Mich. Math. J. 1962, 9: 363-368.MathSciNetView Article
  2. Besicovitch AS: Almost Periodic Functions. Dover, New York; 1954.
  3. Levitan BM: Almost-Periodic Functions. GITTL, Moscow; 1953. in Russian
  4. Corduneanu C: Almost Periodic Oscillations and Waves. Springer, Berlin; 2009.View Article
  5. Bell H, Meyer KR: Limit periodic functions, adding machines, and solenoids. J. Dyn. Differ. Equ. 1995, 7: 409-422. 10.1007/BF02219369MathSciNetView Article
  6. Schwarz W, Spilker J: Arithmetical Functions. Cambridge University Press, Cambridge; 1994.View Article
  7. Andres J: Periodic-type solutions of differential inclusions. 8. In Advances in Mathematical Research. Edited by: Baswell AR. Nova Sciences Publishers, New York; 2009:295-353.
  8. Andres J, Górniewicz L: Topological Fixed Point Principles for Boundary Value Problems. Kluwer Academic, Dordrecht; 2003.View Article
  9. Ichihara N, Ishii H: Asymptotic solutions of Hamilton-Jacobi equations with semi-periodic Hamiltonians. Commun. Partial Differ. Equ. 2008, 33: 784-807. 10.1080/03605300701257427MathSciNetView Article
  10. Lorenz EN: Noisy periodicity and reverse bifurcation. Annals of the New York Academy of Sciences 357. Nonlinear Dynamics 1980, 282-291.
  11. Andres J, Bersani AM, Grande RF: Hierarchy of almost-periodic function spaces. Rend. Mat. Appl. 2006, 26: 121-188.MathSciNet
  12. Corduneanu C: A scale of almost periodic function spaces. Differ. Integral Equ. 2011, 24: 1-28.MathSciNet
  13. Andres J, Pennequin D: On the nonexistence of purely Stepanov almost-periodic solutions of ordinary differential equations. Proc. Am. Math. Soc. 2012, 140: 2825-2834. 10.1090/S0002-9939-2012-11154-2MathSciNetView Article
  14. Alonso AI, Obaya R, Ortega R: Differential equations with limit-periodic forcings. Proc. Am. Math. Soc. 2002, 131: 851-857.MathSciNetView Article
  15. Johnson RA: On almost-periodic linear differential systems of Milionshchikov and Vinograd. J. Math. Anal. Appl. 1982, 85: 452-460. 10.1016/0022-247X(82)90011-7MathSciNetView Article
  16. Millionshchikov VM: Proof of the existence of irregular systems of linear differential equations with almost periodic coefficients. Differ. Equ. 1968, 4: 203-205.
  17. Blot J, Pennequin D: Spaces of quasi-periodic functions and oscillations in differential equations. Acta Appl. Math. 2001, 65: 83-113. 10.1023/A:1010631520978MathSciNetView Article
  18. Goes G: Fourier-Stieltjes transforms of discrete measures; periodic and semiperiodic functions. Math. Ann. 1967, 174: 148-156. 10.1007/BF01360815MathSciNetView Article
  19. Jiménez MN: Multipliers on the space of semiperiodic sequences. Trans. Am. Math. Soc. 1985, 291: 801-811.View Article
  20. Andres J, Pennequin D: On Stepanov almost-periodic oscillations and their discretizations. J. Differ. Equ. Appl. 2012, 18: 1665-1682. 10.1080/10236198.2011.587813MathSciNetView Article
  21. Mortola S, Peirone R: The sum of periodic functions. Boll. Unione Mat. Ital. 1999, 8: 393-396.MathSciNet
  22. Andres J, Bednařík D, Pastor K: On the notion of derivo-periodicity. J. Math. Anal. Appl. 2005, 303: 405-417. 10.1016/j.jmaa.2004.08.020MathSciNetView Article
  23. Pennequin D: Existence of almost periodic solutions of discrete time equations. Discrete Contin. Dyn. Syst. 2001, 7: 51-60.MathSciNetView Article
  24. Blot J, Cieutat P, Mawhin J: Almost periodic oscillations of monotone second-order systems. Adv. Differ. Equ. 1997, 2: 693-714.MathSciNet
  25. Mawhin J: Bounded and almost periodic solutions of nonlinear differential equations: variational vs nonvariational approach. Res. Notes Math. 410. In Calculus of Variations and Differential Equations. Edited by: Ioffe A, Reich S, Shafrir I. Chapman & Hall/CRC, Boca Raton; 1999:167-184.

Copyright

© Andres and Pennequin; licensee Springer 2012

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://​creativecommons.​org/​licenses/​by/​2.​0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.