𝔖 Bobbio Scriptorium
✦   LIBER   ✦

RAZAN: a high-performance switch architecture for ATM networks

✍ Scribed by Mostafa Abd-El-Barr; Khalid Al-Tawil; Habib Youssef; Talha Al-Jarad


Publisher
John Wiley and Sons
Year
1998
Tongue
English
Weight
213 KB
Volume
11
Category
Article
ISSN
1074-5351

No coin nor oath required. For personal study only.

✦ Synopsis


In this paper a high-performance packet switch architecture based on the improved logical neighbourhood (ILN) interconnection network, called RAZAN, is presented. RAZAN is an N Γ— N multistage interconnection network (MIN) which consists of n stages, where n = log 2 N, of switching elements. Each stage consists of a column of N switching elements and (n + 1) Γ— N links. Each switch has n + 1 inputs and n + 1 outputs. Every switching element j is connected to those n + 1 neighbouring switches in the next stage whose binary addresses differ by at most 1 bit from the binary address of switch j. The performance of RAZAN is evaluated both analytically and via simulation under uniform traffic load. The analytical and simulation performance evaluation results are compared. The performance of RAZAN is compared with a number of existing ATM switch architectures such as Benes, parallel banyan and Tagle networks. It is shown that RAZAN exhibits better performance in terms of both the rate of cell loss and throughput. This advantage of RAZAN over existing ATM switch architectures has been achieved at the expense of a moderate increase in switch complexity. In addition, an important characteristic which RAZAN possesses and which distinguishes it further from most existing ATM switch architectures is its ability to achieve very high throughput (higher than 80 per cent) in the presence of faulty switches and/or links. In this paper the fault tolerance characteristics of RAZAN are presented. However, space constraints do not allow us to present a detailed analysis of the fault tolerance and reliability features of RAZAN. These aspects are elaborated in a separate publication.


πŸ“œ SIMILAR VOLUMES


A fast parallel-tree switch architecture
✍ Mayez Al-Mouhamed; Habib Youssef; Wasif Hasan πŸ“‚ Article πŸ“… 1998 πŸ› John Wiley and Sons 🌐 English βš– 288 KB πŸ‘ 1 views

In this paper we present a novel fast packet switch architecture based on Banyan interconnection networks, called parallel-tree Banyan switch fabric (PTBSF). It consists of parallel Banyans (multiple outlets) arranged in a tree topology. The packets enter at the topmost Banyan. Internal conflicts ar

A performance analysis scheme for intell
✍ C. Ben Ahmed; N. Boudriga; M. S. Obaidat πŸ“‚ Article πŸ“… 1999 πŸ› John Wiley and Sons 🌐 English βš– 272 KB πŸ‘ 2 views

An intelligent network (IN) is a concept characterized by the distribution of network intelligence and capabilities wherever required within a telecommunications network. It is considered as a platform de"ning an environment that is rich in network capabilities and service functionality. IN techniqu

A non-blocking architecture for a wavele
✍ Takeshi Nagata; Kaoru Sezaki; Yasuhiko Yasuda πŸ“‚ Article πŸ“… 1998 πŸ› John Wiley and Sons 🌐 English βš– 218 KB πŸ‘ 1 views

The authors previously derived the condition for a non-blocking architecture in a wavelength division multiplexing (WDM) photonic switching network, i.e., the generalized parallel delta network (GPDN) . GPDN is configured as a multi-stage link connection of multiple unit switches, where the hardware

A class of centralized high-performance
✍ Georgios I. Papadimitriou; Andreas S. Pomportsis πŸ“‚ Article πŸ“… 1999 πŸ› John Wiley and Sons 🌐 English βš– 155 KB πŸ‘ 1 views

A new class of protocols for WDM star networks is introduced. According to the proposed protocols, two or more stations share each wavelength and transmit their packets in a random access fashion. An array of tunable electro-optic "lters or tunable wavelength converters, placed at the network hub, a

A low power dissipation architecture of
✍ Yoshitaka Tsunekawa; Michiru Iwawaki; Mamoru Miura πŸ“‚ Article πŸ“… 2000 πŸ› John Wiley and Sons 🌐 English βš– 322 KB πŸ‘ 1 views

In this paper, a low power dissipation architecture of high-performance multiprocessor is proposed for statespace digital filters using block-state realization in order to realize high-accuracy, high-speed process with reduced hardware previously proposed by the authors. Distributed arithmetic is ap