Font Size: a A A

Research On Ethernet Passive Optical Network Based On Ant Colony Algorithms For Bandwidth Distribution In Uplink Direction

Posted on:2011-10-18Degree:MasterType:Thesis
Country:ChinaCandidate:Y S MaFull Text:PDF
GTID:2178360302494706Subject:Detection Technology and Automation
Abstract/Summary:PDF Full Text Request
The upstream bandwidth allocation algorithm of Ethernet Passive Optical Network (EPON) which is the most promising next-generation broadband access network has been a research focus in this area. Efficient bandwidth allocation algorithm requires OLT to distribute bandwidth fairly, ensure QoS of different services and high bandwidth utilization. In this paper, in order to make the cost of the system to reach the minimum, ant colony algorithm is used to distribute reasonable service rates for ONU in system based on working vacation mechanism.First, the paper describes the development of EPON at home and abroad and its main characteristics, focused on the upstream bandwidth allocation algorithm for EPON research progress; the ant colony algorithm with other bionic algorithm were compared to arrive at the advantage and disadvantages of ant colony algorithm.The paper analysis the working principle of the EPON system; decompose the ONU node through decomposition method and analysis the ONU function; proposed a bandwidth allocation program based on working vacation polling mechanism.Then, the paper introduce ant colony algorithm, description the parameters of algorithm and how to configure the parameters; given adaptive pheromone-based ant colony algorithm volatile factors, and compare with basic ant colony algorithm in the same parameters configured by solving TSP problem, obtained that the improved ant colony algorithm can obtain better results.Finally, build the objective function of system cost; describe the ant colony algorithm flow for solving the lesser cost of system; give the result of experiment used basic ant colony algorithm and adaptive pheromone-based ant colony algorithm volatile factors respectively, and compare the two kinds of ant colony algorithm; observe the service rate of the other queues by changing the arrival rate of one, analysis the performance of system.
Keywords/Search Tags:Ethernet Passive Optical Network, Ant colony algorithm, Working vacation queuing, System cost, Serving rate optimize
PDF Full Text Request
Related items