Re: SDL-News: Time restriction


Subject: Re: SDL-News: Time restriction
From: Rick Reed TSE (rickreed#tseng.co.uk)
Date: Tue Dec 29 1998 - 20:28:05 GMT


The originator of this message is responsible for its content.
-----From Rick Reed TSE <rickreed#tseng.co.uk> to sdlnews -----

At 18:28 +0000 29/12/98, Maha Boughdadi wrote:
>I would like to know how SDL handles reusability, time constraints, how
>to preserve encapsulation, verification of timing restriction.

1. Reusability

SDL has a number of features that allow components to be defined.
Components can be collected togther in a package and then selectively
extracted from a package. Most re-usable components will have to be defined
as types.

2. Time contraints

Current SDL models observable behaviour. Usually a time constraint is that
a certain action must be performed before some particular time has expired.
Such constraints are a form of invariant, and the behaviour of the system
if the constraint is not met is not considered because it should never
occur.

If the constraints are of this kind then they are not part of the model,
but are a form of additional information. The SDL timer mechanism can be
used to monitor time constraint deadlines. If the timer ever expires, the
constraint is not met.

Further study is being undertaken within the ITU-T of how to more
explicitly link performance issues with the SDL model.

3. Preserving encapsultation

SDL blocks are a natural unit of encapsulation.

4. Verifying Timing restriction

Interpreting this as "verifying that a time deadlines are always met",
obviously requires a model of real resources.

The basic SDL model (more or less) assumes that each process instance may
have its own processor, there may be an insignificant delay in sending a
message from one process to another, and the time for a transition from one
state to another is insignificant. That is, the SDL may be running on an
ideal environment. In reality there may be a monoprocessor of limited
power, or the speed of communication between the components may be limited.

Verify timing of the SDL, therefore requires that delays and restrictions
are added to the SDL model. This can be done in SDL (by introducing extra
processes that delay signals, and extra waiting states in the SDL), or by
using a tool that allows such information to be attributed to transitions
and channels.

Verifying the timing (that is - proving that it is correct) would not be an
easy task, even with sophisticated tools. What is more feasible to to
validate the model - that is, increase confidence that it is correct. Of
course, it is often shown that the model is not valid and some change has
to be made to the model.

--
Rick Reed, TSE Limited
13 Weston House, 18-22 Church Street
Lutterworth Leicestershire LE17 4AW United Kingdom
Tel +44 14 55 55 96 55; Fax +44 14 55 55 96 58
Mob +44 79 70 50 96 50
email: rickreed#tseng.co.uk
http://www.tseng.co.uk   ftp://ftp.tseng.co.uk/tseng/

-----End text from Rick Reed TSE <rickreed#tseng.co.uk> to sdlnews ----- For help, email "majordomo#sdl-forum.org" with the body of your email as: help or (iff this does not answer your question) email: owner-sdlnews#sdl-forum.org



This archive was generated by hypermail 2a23 : Sun Jun 16 2013 - 10:41:40 GMT