Advanced elements for

appliancepartΤεχνίτη Νοημοσύνη και Ρομποτική

19 Οκτ 2013 (πριν από 4 χρόνια και 22 μέρες)

115 εμφανίσεις

Advanced elements for
transforming networks


VLADISLAV SKORPIL


Department of Telekommunications,

Brno University of Technology,

Purkynova 118, 612 00 Brno,

CZECH REPUBLIC



Synopsis


New generation transforming communication networks are
based on convergence of classical telecommunication
networks into computer networks.


This new generation network needs new advanced network
elements (NE).


Classical sequential data processing is constrained by the
speed of central processing unit.


The aim of the paper is to seek an alternative way of
increasing the performance of systems by parallel data
processing whose activities can be realised using the neural
networks.


This paper recommends NE based on active elements,
which process the transmitted data units
.

Introduction


Neural networks are very suitable for controlling
of future NE.


Optimising of Communication Systems Design,
using neural networks involves the research,
simulation and hardware implementation of
selected problems of the fast evolving converged
platform of classical communication and advanced
data services.


Neural networks are used for advanced research
solution. We have primary usable results with
Kohenen neural network ant with
Back/propagation learning algorithm.

II. APPROACH TO STUDY



The switching
-
over is the basic function of
the active network elements that work
above physical layer of OSI model.


The function of switch is to transport fastest
the inputs to (output) target.


The speed can be limited by blocking when
data flow from two or more inputs are to be
directed to one output.


Objective


application of Genetic algorithm (GA) for
design of
Neural N
etwork

(NN)


control of communication Network Element
(NE) by NN


seek an alternative way of increasing the
performance of
NE
by parallel data
processing




classical version of Genetic algorithm uses
three genetic operators


reproduction,
crossover and mutation


Radial Basis Function Network

(
RBFN
)

is
here used


it is a
type of single
-
direction multilayer
network

Design of a General Schematic of
the Genetic Algorithm


Generating of initial population


Ageing


Mutation


Calculation


Sorting of upward population


Crossing


Finalization




III. THE BASIC MODEL OF THE
NETWORK ELEMENT



modern possibility, how to change classical
sequential control of
N
etwork
E
lements
NE
to
control using of neural networks


design a simulat
ion of NE
, containing in the
process of control of switching area artificial
neural network

with GA


NE

switches single data units making provision
for priority

Fig. 1
Model of the switch with artificial
neural network

The basic scheme of the elemen
t



We think over the single
-
stage switching
area, which has three inputs and three
outputs, it is switch on the Fig.
2



The switching area is realized on the cross
-
bar switch, i.e. in the described case the
switching area with 9 switching points.


We can connect arbitrary input to arbitrary
output.

Fig.2 Switch

Fig.3
Switching area

Fig.4 Switching area with addressing

Fig.5 Frame structure

Fig.6 Switching area controlled by control matrix

4. Conclusion


crushing majority to learn neural network for
diagnostic of one object completely on 100%
with
GA


time
of learning is
shorter than
for
classical
methods


the results and the learning time highly depends on
G
A
parameters setting


t
he best results were obtained by G
A

using the D
-
operator and not using sexual reproduction.


it

is shown modern possibility, how to change
classical sequential control of network elements to
control using of neural networks




The convergence of classical
telecommunication networks and data
networks is the first step in designing
universal broadband integrated networks for
different types of user services, including
videoconference applications or multimedia
services and unified network management.


The integrated network must be able to
guarantee different transport parameters for
different services.


The problem is in the network elements,
which must guarantee the required
parameters and also offer a sufficiently
broad bandwidth, all this at a reasonable
price.


In the area of the development of high
-
speed networks the possibilities of
increasing the throughput and effectiveness
of active elements are sought.