Read the instructions to get the latest distribution of ANCLES and CLASS from the download page

Introduction

ANCLES stands for ATM Networks Call LEvel Simulator; this acronym stresses that the package is strictly devoted to the simulation of call level dynamics in ATM networks. Whenever a user forwards a connection request, routing and call admission controlfunctions start to work in order to find a convenient path over which transmission resources could be allocated to the incoming connection.

The performance of several routing algorithms and call admission control techniques can be compared on the basis of such indices as:

• the load accepted by the whole network
• the load carried by each link
• the call blocking probability

ANCLES/CLASS interaction

Cell level phenomena are not directly simulated by ANCLES: the experimental evaluation of quality of service indices, such as the cell loss probability and the cell delivery delay, is thus not possible. But ANCLES is able to recognize congested situations of the analized network, and to "take a snapshot" of active connections, thus making the cell level analysis possible by feeding the snapshot to CLASS, a cell level simulator also developed by the Networks Group.

Using ANCLES

ANCLES allows one to define arbitrary topologies, with a great freedom of choice of the number of nodes within the network, the number of ports connected to each node, the traffic characteristics, and so on. The simulation sessions start with the acquisition of the network topology. A suitable formal grammar, dubbed ND - Network Descriptor, has been defined in order to check the correct definition of network topology and simulation parameters, while providing users with an easy, flexible tool to set up complex simulation scenarios. The downloadable ANCLES distribution includes a sample .nd file, describing a parking lot topology.

The results obtained during the ANCLES simulation runs are written to ASCII files. Results include statistics on the accepted load and the call blocking probability, and refer to both the whole network and each link or user.

Written in ANSI C, and using such common tools as FLEX and YACC for its lexical analyzer, ANCLES is completely machine-independent.