Using object-oriented programming in computational electromagnetic codes

Object-oriented numerical simulation of a passively modelocked fiber laser

“…4,5 The use of OOP for scientific computing enhances code development, maintenance, flexibility, portability and memory efficiency. 2,7 However, object interfaces which usually reside between the abstraction layer and the numerical subroutines hinders program performance as we shall demonstrate. 16 Initially, the laser modeled was programmed in MATLAB.…”

“…2, 3, 6-8 For example, in computational electromagnetics, OOP concepts became vital to developing complex code used for hybrid time-domain numerical analysis. 7 Also, extendable and reusable OOP C++ code utilizing inheritance, polymorphism and encapsulation has been developed for integrated optics modelling. 8 In this work, we present an OOP based simulation of a passively modelocked fiber laser containing a long period fiber grating (LPFG).…”

Introductory review on object oriented paradigm for full-wave microwave CAD

“…2); the reason for this is easily understood by looking at the publication year. An interesting discussion was made in [19], which also used Java. Java seems to be necessary today to take full advantage of the internet and distributed networks.…”

Dynamics of a passively mode-locked fiber laser containing a long-period fiber grating

“…Usually, data is predefined and fu n ctio n calls follow after. However, in O OP b o th are mean in g fu lly integrated to form objects related to the problem dom ain [78]. As a conse quence, O OP enables dynam ic m em ory allocation at any stage o f com puting w ith a specific life tim e, thus, ensuring m axim um generality and enhancing performance [74].…”

“…Usually, data are defined as private to elim inate conflicting definitions [74], w hile most functions are defined as public form ing a class interface [76,77]. The use of O OP for scientific com puting enhances code develop ment, maintenance, fle x ib ility , p o rta b ility and m em ory efficiency [74,78]. However, object interfaces w hich usually reside between the abstraction layer and the num erical subroutines hinders program performance as we shall dem onstrate [80].…”