SciELO - Scientific Electronic Library Online

 
vol.26 issue2wPOI: Weather-Aware POI Recommendation EngineRide Sharing Using Dynamic Rebalancing with PSO Clustring: A Case Study of NYC author indexsubject indexsearch form
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

  • Have no similar articlesSimilars in SciELO

Share


Computación y Sistemas

On-line version ISSN 2007-9737Print version ISSN 1405-5546

Abstract

BISWAS, Abhijit  and  HUSSAIN, Anwar. Design and Analysis of a New Reduced Switch Scalable MIN Fat-Tree Topology. Comp. y Sist. [online]. 2022, vol.26, n.2, pp.949-962.  Epub Mar 10, 2023. ISSN 2007-9737.  https://doi.org/10.13053/cys-26-2-3902.

This paper presents a reduced switch scalable MIN Fat-Tree with new inter-router connections. Unlike the conventional Fat-Tree, consecutive switches are connected with new bidirectional links for faster intergroup communication. In case of traditional Fat-Tree, a packet has to travel upward upto the level called summit from where the downward path to the client is available, most often the summit is at the highest router level of the MIN Fat-Tree. The proposed topology aims at lowering the summit by eliminating the entire topmost switch level, thereby reducing the size of the network significantly without considerable degradation in the network performance and maintaining the scalable property of the MIN Fat-Tree. The results indicate that the proposed network can not only reduce the delay but also reduce the number of switches to a great extent. It was found that around 33.33% of switches were reduced in the proposed network for eight clients and 14.28% for 128 clients.

Keywords : Network-on-chip; Fat-Tree; switch; topology; end-to-end delay; hop count; routing algorithm.

        · text in English