Modified limited data rate usage algorithm using Markovian decision process model

Abstract

The dual mode hand-held devices to connect both the wireless local area network (WLAN) and the cellular network, i.e., the universal mobile telecommunication system (UMTS), are becoming popular. Users want to make trade-off between the use of high cost low speed UMTS covering wide area and that of WLAN providing low cost high speed services within limited area on demand basis. The switching from one network to another is referred as the vertical handover (VHO). VHO from UMTS to WLAN is not challenging in terms of switching time as the UMTS still works within the communication range of WLAN. However, VHO from WLAN to UMTS poses strict time constraints as the signal of the previous network become weaker quickly. Many connections drop in the process depending on the velocity of the mobile devices as the procedures do not complete within the time obtained. The success rate of VHO from WLAN to UMTS can be improved significantly by incorporating delaying handoff, i.e., extending the time span in the previous network using limited data rate usage (LDU) method. Unfortunately, the existing LDU method modelled through Markovian decision process did not consider the basic and important assumption that a call can be returned back to its previous state within its transitional time of the VHO process. In this thesis, we have modified the LDU method and model it using Markovian decision process incorporating the abovementioned assumption. Simulation results prove the effectiveness of our proposed modified LDU method in terms of handoff dropping probability and call blocking probability.