DocumentCode :
3279782
Title :
Hybrid Markov Models Used for Path Prediction
Author :
Yu, Xue-gang ; Liu, Yan-heng ; Wei, Da ; Ting, Min
Author_Institution :
Coll. of Comput. Sci. & Technol., Jilin Univ., Changchun
fYear :
2006
fDate :
9-11 Oct. 2006
Firstpage :
374
Lastpage :
379
Abstract :
Path prediction is an important issue in QoS of wireless networks. The paper points out problems in some existed path prediction schemes, especially the state space expansion problem in order-k Markov predictor. And it firstly proposes a step-k Markov model and validates its feasibility. Secondly, a hybrid Markov predictor model and its improved models are put forward based on the step-k Markov model. Because of the order-2 Markov model´s best performance in order-k Markov models, the Hybrid Markov model takes the order-2 Markov model as its target. The state space´s complexity of the Hybrid Markov Model is 0(N) while the order-2 Markov model is O(N2). And the memory demand of the hybrid Markov model is O(N2) while Order-2 Markov model is O(N3). Finally, it is proved that the hybrid Markov predictor can get close performance with order-2 Markov predictor at much lower expense by conditional entropy analysis and user mobility data analysis. Also it can alleviate the zero probability problem in order-k Markov model to some extent. The hybrid Markov predictor is more practical than order-k Markov predictors under WLAN.
Keywords :
Markov processes; quality of service; wireless LAN; hybrid Markov models; order-k Markov predictor; path prediction; quality of service; state space expansion; step-k Markov model; wireless LAN; wireless networks; Accuracy; Computer science; Educational institutions; Equations; Predictive models; Random variables; Space technology; State-space methods; Wireless LAN; Wireless networks; EM Algorithm; Hybrid; Markov Model; State Space Expansion;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Communications and Networks, 2006. ICCCN 2006. Proceedings.15th International Conference on
Conference_Location :
Arlington, VA
ISSN :
1095-2055
Print_ISBN :
1-4244-0572-6
Type :
conf
DOI :
10.1109/ICCCN.2006.286304
Filename :
4067685
Link To Document :
بازگشت