Professional Documents
Culture Documents
ISSN No:-2456-2165
Abstract:- Software complexity crisis becomes an types of errors, and it can also be used as input for model
impediment to further development of software. checking [1].
Specifically, in order to manage increasingly complex
and massive software systems, researchers involve in To continue the research [2][3], we proposed an
building systems with autonomy. The real-time reactive approach to model reactive autonomic system, component
systems with autonomic behaviors could be more self- group and components in this paper. The rest part of the
managed and more adaptive to the environment. paper is presented as follows: Section 2 sketches the work
However, formations of some of such systems are not related to the research. Section 3 introduces the background
formalized, which may lead systems to be error-prone. knowledge needed to understand the paper. Section 4
In this research, we proposed a formal way to describe introduces formation of reactive autonomic system and its
formations of reactive autonomic systems framework. categorical models. Section 5 describes formation of
Firstly, we introduce how to from reactive autonomic reactive autonomic component group and its categorical
system, components group, and component, then we models. Section 6 explains formation of reactive autonomic
focused on categorizing the formations. To do so, the component and its categorical models. In the last,
basis of reactive autonomic systems can be built with conclusion and future work are provided in Section 7.
correct by construction.
II. RELATED WORK
Keywords:- Reactive Autonomic System; Category Theory;
Formation This section introduces related research work to the
paper.
I. INTRODUCTION
A. Real-Time Reactive Systems
Software complexity crisis has been deemed as one of In research [4], Caporuscio specifies that there is a
the major obstacles to the progress in software industry, trend to switch assembling components into systems to
since the management of computing systems with composing autonomous systems into systems-of-systems
complexity becomes difficult for IT Practitioners. We need dynamically, as these kinds of system usually run in highly
to select a target system that can get benefit from applying dynamic environments. In research [5], an approach to
the autonomic computing paradigm. Usually, real-time solve the problem of synchronous programs that cannot be
reactive systems are considered as one of the most complex executed in a time-triggered or event-triggered execution
systems; the complexity involved comes from their real- loop easily. By using dynamic tickets method, semantic
time as well as reactive characteristics: 1) concurrency is timing abstraction of the synchronous approach can be
involved; 2) timing requirements are strict; 3) reliability is a reconciled that can help to give the application fine-grained
must; 4) software and hardware components are involved; control over its real-time behavior.
5) it becomes more and more intelligent and heterogeneous.
So, we need to add autonomic features into real-time B. Formal methods
reactive systems by using Reactive Autonomic Systems Formal method can be used to assess risks at the early
Framework (RASF), which helps to specify, model and stages of developing security-critical real-time systems in
develop the Reactive Autonomic Systems (RAS). By research [6], where a formal model named Object-Message-
adding autonomic behaviors, real-time reactive systems Role (OMR) with Z notation is proposed to specify
become more self-managed and adaptive; the RAS can functional and security aspects of systems. Research [7]
improve and simplify users’ experiences. By using input defined a model using formal method to evaluate the
data and checking results only, it is difficult to find the quality of system architecture. By formalizing
competitive conditions in the real-time reactive system, characteristics of system architecture, good quality
because some errors may only occur when the process attributes can be quantified and can be used in may system
sends or receives data at a specific time. In order to detect architecture tools. In research [8], probabilities are
these errors through testing, every state combination of combined with a formal approach to develop safety-critical
processes must be checked, which may result in an systems. In this approach, model-driven engineering is used
exponential number of states [1]. Formal methods have for reactive systems and a tool-set reactive blocks extend
been proven to be the way to ensure the correctness of the support of modeling and verification of behaviors in
operation in complex interactive systems, because formal real-time systems. In paper [9], instead of reviewing
specifications are proved that it can help to check specific requirements manually, a formal scenario-based method
RACG-Formation
RACS-Type1 RACS-Type2
Comm-Type5
Fig 3:- Evolution for Self-Configuration in RAS1
A. Forming a RACG
After receiving the task of forming a RACG from
RACGM, RACS starts to create RAOL and establish
corresponding connections among them based on the RACG1 RACG2
composition rules and communication protocols specified
by index category RACG-Formation. Fig.4 illustrates an RACS1 of RACS2 of
RACS-Type1 RACS-Type2
example of forming the categories RACG1 and RACG2
from their index category RACG-Formation.
Comm4 of Type4
Comm3 of Type5
Comm4 of Type6
Comm4 of Type4
Comm3 of Type5
Comm4 of Type6
Take-over RAO
can be specified as the categories in which objects are those
sequences (<InitializeRAOL, Heartbeat, InitializeRAO,
Heartbeat>, <ValidateRAOL, Conform, ValidateRAO,
RAC1-0
NotConform>), and morphisms are the equivalence RAOL1