QoS management and performance analysis of WiMAX networks in fixed and highly mobile environments

Driven by the growing demand for high-speed broadband services, Worldwide Interoperability for Microwave Access (WiMAX) technology has emerged as a competitive alternative to wireline broadband access solution. WiMAX technology, considered in this thesis, offers an IP-based framework that provides high data rates at medium and long range with the ability of supporting fixed, nomadic, portable, and mobile access. Moreover, based on the IEEE 802.16 standard, the technology provides a set of built-in QoS mechanisms to support heterogeneous classes of traffic including data, voice and video. The IEEE 802.16 standard, however, leaves open the resource management and scheduling mechanisms, which are crucial components to guarantee QoS performance for these services.

In this thesis, we evaluate the performance of IEEE 802.16 based WiMAX technology in both fixed and highly mobile environments. More particularly, mobile WiMAX is investigated as a vehicular-to-infrastructure (V2I) communication medium since it is expected to play a major role in intelligent transportation systems. The technology is indeed the only mobile broadband technology currently in use.

Moreover, we address in this thesis most of the resource management and scheduling issues that have been left open with the objective of defining an architecture that fulfils the QoS expectations of the five classes of applications addressed by the IEEE 802.16 standard. In fact, after surveying, classifying and comparing different scheduling and admission control mechanisms proposed in this work-in-progress area, we propose two QoS solutions. Both solutions address point-to-multipoint (PMP) 802.16 systems operating in time division duplex mode (TDD) mode.

The first solution includes a hierarchical scheduling algorithm that adapts the DL/UL allocations on a frame-by-frame basis to serve unbalanced traffic. The amounts of these bandwidth grants are set by the connection admission control (CAC) module that adopts a Max-Min fairness approach making efficient and fair use of the available resources. The proposed solution takes into account the link adaptation capability supported by WiMAX and the data rate constraints of the different types of services. The second QoS solution presented in this thesis is a multi-Constraints Scheduling Strategy (mCoSS) that is designed for both OFDM or band-AMC OFDMA air interfaces. Unlike the first QoS solution, mCoSS supposes the use of a predefined DL/UL ratio set by the operator. In addition to data rate constraints, mCoSS offers the advantage of supporting delay constraints of real-time applications and handling bursty traffics. mCoSS is based on a modified dual-bucket traffic shaping mechanism configured on a per-flow basis. This shaping mechanism is combined with a two-rounds scheduling strategy which reflects (i) at the first round, the minimum data rates and latency requirements the BS or MS is committed to provide and (ii) at the second round, the efficiency and fairness of the resources management since the remaining bandwidth is shared in this round using a weighted fair queuing (WFQ) strategy.