Dynamic Bandwidth Quasi- reservation Scheme for Real- time Services in IEEE e Networks Yin Ge,and Geng-Sheng Kuo IEEE Wireless Communications and Networking Conference, 2007.WCNC 2007

Outline Introduction Related Content in IEEE e Standard The Dynamic Bandwidth Quasi- Reservation Performance Simulation Conclusion

Introduction IEEE e standard did not specify any bandwidth reservation scheme that provides quality of service (QoS) support for real-time services Handover real-time services > non-real-time

IEEE e Standard Two-phase activation model Consideration for Handover cannot be guaranteed to get the required bandwidth

Dynamic bandwidth quasi- reservation scheme (DBQRS) based on the handover probability the traffic arrival probability Provide QoS guarantee for real-time multimedia services

The Proposed DBQRS

The probability of MS requiring BS to reserve bandwidth Type 0 trafficType 1 traffic MS 1 MS 2 MS … SF 3 SF 0SF 1 SF 2SF 3 SF 0 SF 2 SF 1

The probability of MS requiring BS to reserve bandwidth handover and traffic arrival are independent

The probability of MS requiring BS to reserve bandwidth Due to the different handover states as well as the diverse QoS requirements of traffics Weighting coefficients α n UGS ERT-VR RT-VR

Handover states Weighting coefficients β k β in : immigrate into the BS ’ s coverage area from the neighbor cells β stay : stay in the BS ’ s coverage area β out : move out of the serving BS ’ s coverage area dynamic bandwidth quasi-reservation for MSs according

Admission Control Policy Implemented a handover MS with active SFs enters the BS ’ s coverage, and a local MS issues new SFs Real-time SFs arrive, non-real-time SFs must release the reserved bandwidth By initiating DSC message to change the state of non-real-time SFs to be inactive

Admission Control Policy (real- time SF)

Admission Control Policy (non- real-time SF)

EX Bandwidth MS … Quasi-Bandwidth reservation B free

Performance Simulation simulation model: OPNET Modeler 7 hexagonal cells 70 local MSs 30 handover MSs per cell evaluation on the central cell weighting coefficients α 1 =1 for UGS, α 2 =0.8 for RT-VR/ERT-VR β in =1 for the incoming handover MSs, β stay =0.5 for the local MSs and β out =0.2 for the outgoing handover MSs

Traffic models

New SF blocking rate & Access delay

Handover SF dropping rate & Bandwidth utilization

Conclusion The proposed scheme not only provides QoS guarantee for real-time services, but also ensures the fairness of admitting handover and new real-time services The proposed scheme achieves low NSBR and HSDR, low access delay for new real-time service as well as high system bandwidth utilization

Thank you!