A Cross-Layer QoS Guarantee And Scheduling Algorithm For 802.16 Systems

The next generation wireless metropolitan access network (WMAN) technology IEEE 802.16 d/e has defined a set of specifications to support quality of service (QoS) for heterogeneous services over wireless air interface. At the media access control (MAC) layer, it specifies scheduling services, service flow management, bandwidth request, polling, grant mechanisms and relevant signaling. However, it doesn't define QoS mapping, admission control, traffic adjustment and scheduling algorithms, which are open for developers. At the physical layer, IEEE 802.16 utilizes advanced techniques such as adaptive modulation and coding (AMC) to guarantee bit error rate (BER) of 1*10^-6 . Due to the heterogeneous services over WMAN, as well as the dynamic variation of wireless transmission channel, the end-to-end QoS guarantee over 802.16 systems requires cooperation across layers, and consequently cross-layer design becomes critical in the future.In this article, the author studies the existing QoS guarantee mechanisms and scheduling algorithms with QoS support, and proposes a cross-layer QoS guarantee and scheduling algorithm for 802.16 systems. The proposed scheme includes admission control module, uplink scheduling module, and AMC control module at the base station (BS), as well as traffic adjustment module, downlink scheduling module, and AMC selection module at the subscriber station (SS). The scheduling module adopts a two-layer scheduling algorithm based on priority. At the first layer, deficit fair priority queue (DFPQ) algorithm is utilized to schedule different service classes with fairness. At the second layer, according to the QoS requirements of UGS, rtPS, ertPS, nrtPS and BE services, as well as the current physical channel quality, a priority function (PRF) is assigned to each connection and dynamically updated. Thus, the connection with the highest priority is scheduled each time. AMC selection module periodically measures the wireless channel, updates the state transition matrix of finite-state Markov chain to trace the channel status, and dynamically selects transmission mode (TM), which includes modulation and forward error coding, so as to guarantee packet error rate (PER) performance and improve spectrum efficiency.In the proposed cross-layer QoS architecture, each module collaborates to enhance end-to-end QoS support of WiMAX. AMC control module provides physical layer parameter normalized channel quality factor for scheduling module. Scheduling module assigns priority with fairness for each connection to satisfy diverse QoS requirements. Based on the minimum reserved rate provided by the scheduling algorithm, the token-bucket algorithm in the traffic adjustment module could satisfy the requirement of average bit rate. For real-time services which are sensitive to delay and delay jitter, traffic adjustment could guarantee their delay and delay jitter requirements. In the case of heavy load, scheduling algorithm collaborates with admission control to provide QoS guarantee.With OPNET 11.0 simulator, WiMAX system modules are developed in order to simulate the schemes and algorithms in this article. The simulation result verifies the system performances on diverse QoS support, scheduling fairness, spectrum efficiency, and scalability.