A LOCAL FAIRNESS ALGORITHM FOR GIGABIT LANS MANS WITH SPATIAL REUSE

Citation
Jsc. Chen et al., A LOCAL FAIRNESS ALGORITHM FOR GIGABIT LANS MANS WITH SPATIAL REUSE, IEEE journal on selected areas in communications, 11(8), 1993, pp. 1183-1192
Citations number
16
Categorie Soggetti
Telecommunications,"Engineering, Eletrical & Electronic
ISSN journal
07338716
Volume
11
Issue
8
Year of publication
1993
Pages
1183 - 1192
Database
ISI
SICI code
0733-8716(1993)11:8<1183:ALFAFG>2.0.ZU;2-A
Abstract
In this paper, we present an algorithm to provide local fairness for r ing and bus networks with spatial bandwidth reuse. Spatial bandwidth r euse can significantly increase the effective throughput delivered by the network and is, therefore, desirable to be implemented in high-spe ed LAN/MAN environments. However, spatial bandwidth reuse can result i n unfair access among nodes in the network and, thus, a fairness algor ithm is needed to regulate the access to the network. A local fairness algorithm views the network as multiplicity of communication resource s as opposed to a global fairness algorithm, which views the network a s a single communication resource. Our algorithm can be applied to any dual ring or bus architecture such as MetaRing [11, [4]. In the dual bus configuration, when transporting ATM cells, the local fairness alg orithm can be implemented using two generic flow control (GFC) bits in the ATM cell header. In the performance study section of our paper, w e will show that this local fairness algorithm can exploit the through put advantage offered by spatial bandwidth reuse better than a global fairness algorithm. This is accomplished because it ensures fair use o f network resources among nodes which are competing for the same subse t of links, while permitting free access to noncongested parts of the network. We will demonstrate the performance advantage of our local fa irness scheme by simulating the system under various traffic scenarios and compare the results to that of the MetaRing SAT-based global fair ness algorithm. Furthermore, we will show that under certain traffic p atterns, the performance of this algorithm achieves the optimal throug hput result predicted by the known Max-Min fairness definition [7].