Font Size: a A A

Dynamic Bandwidth Allocation Strategy With Adaptive Transmission Rates And Performance Analysis

Posted on:2013-02-19Degree:MasterType:Thesis
Country:ChinaCandidate:Q LvFull Text:PDF
GTID:2218330362962876Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
With the rapid development of network technology, there have been many types oftraffics, and different traffics have different QoS requirements. Today, network resourcehas become more and more valuable, how to allocate bandwidth reasonably is a researchfocus. Taking into account both the traffic type and system load, two dynamic bandwidthallocation strategies with adaptive transmission rates are proposed. Performance analysisand system optimization are brought out.Firstly, according to different QoS requirements, traffics are classified into threetypes, namely type Ⅰ, Ⅱ and Ⅲ with decreasing priorities. Under the condition that theQoS requirement of type Ⅰ traffic is satisfied, considering the traffic load, two levels oftransmission rates are set for type Ⅱ traffic, a dynamic bandwidth allocation strategy with2-level adaptive transmission rates, i. e. DBA-2R, is proposed. Based on this strategy,considering the idle period of type Ⅱ traffic, by introducing a zero transmission rate totype Ⅱ traffic, a dynamic bandwidth allocation strategy with3-level adaptivetransmission rates, i. e. DBA-3R, is proposed.Secondly, for the DBA-2R, a queuing model with multiple working vacations isestablished. For the DBA-3R, a queuing model with single working vacation is built. Byusing the method of an embedded Markov chain, the steady-state measures of thequeuing model are derived. Correspondingly, the performance measures in terms of theaverage response time and system utility are given for the two dynamic bandwidthallocation strategies with adaptive transmission rates presented above.Thirdly, based on the two dynamic bandwidth allocation strategies with adaptivetransmission rates, numerical examples are brought out, the tradeoff between averageresponse time and channel utility is investigated. Moreover, the model building andmodel analysis are verified with simulation results.Finally, considering both the average response time and channel utility, theguaranteed bandwidth of type Ⅱ traffic is optimized for DBA-2R, The cycle length of thebandwidth taken by type Ⅲ traffic occupying the remainder bandwidth is also optimized for the DBA-3R.
Keywords/Search Tags:dynamic bandwidth allocation, adaptive transimission rate, traffic type, working vacation, embedded Markov chain, performance measures, parameter optimazation
PDF Full Text Request
Related items