Adaptive seamless association handoff for guaranteeing real-time traffic in WiMAX mobile networks

• Main Authors: Ren-Hung Hwang, Ben-Jye Chang, Yan-Min Lin, Ying-Hsin Liang
• Format: Conference Proceeding
• Language: English
• Subjects: adaptive RSS prediction; association delay; Association handoff; Bandwidth; Delay; Distance measurement; load balancing; Load management; Mobile communication; Mobile computing; WiMAX
• ISSN: 1530-1346; 2642-7389
• DOI: 10.1109/ISCC.2010.5546730

The IEEE 802.16e standard is proposed for mobile nodes to access high data rate and long transmission radius in WiMAX. For supporting contention-free-based initial ranging and early performing the initial ranging during handoff, IEEE 802.16e specifies the association handoff mechanism in the MAC layer. However, an MS is difficult to determine the scan duration in the association handoff mechanism because it may lose the transmission opportunity or the response status of the received ranging response (RNG-RSP) message. Although the MS can set a longer scan duration to complete the initial ranging procedure for the association handoff, a long duration significantly degrades the QoS parameters of delay and delay jitter of real-time service flows. This paper thus proposes an efficient adaptive load-balancing association handoff approach (namely ALA) that consists of two phases: (1) the Adaptive Association Handoff phase (AAH) and (2) the Predictive-direction-based load balancing phase (PDLB), to overcome above mentioned problems. AAH proposes an adaptive re-association to reduce lost synchronizations, and thus improve the grade of service. PDLB adopts a prediction scheme to accurately predict the moving direction of mobile nodes, and thus achieve the reduction of the number of candidate target BSs in order to minimize the handoff delay and to balance traffic loads. Numerical results demonstrate that the proposed ALA approach obviously outperforms the IEEE 802.16e protocol and other approaches in average handoff delay, number of handoffs, dropping probability, grade of service, network utilization, and number of lost synchronizations.