Research Article
Connection-Level QoS Provisioning in Multiple Transmission Technology-based OFDM System
@INPROCEEDINGS{10.1109/WIOPT.2006.1666449, author={Youngkyu Choi and Sunghyun Choi and Sung-Pil Hong}, title={Connection-Level QoS Provisioning in Multiple Transmission Technology-based OFDM System}, proceedings={4th International ICST Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks}, publisher={IEEE}, proceedings_a={WIOPT}, year={2006}, month={8}, keywords={}, doi={10.1109/WIOPT.2006.1666449} }- Youngkyu Choi
Sunghyun Choi
Sung-Pil Hong
Year: 2006
Connection-Level QoS Provisioning in Multiple Transmission Technology-based OFDM System
WIOPT
IEEE
DOI: 10.1109/WIOPT.2006.1666449
Abstract
To provide heterogeneous quality-of-service (QoS) to various applications, the bandwidth of orthogonal frequency division multiplexing (OFDM) is divided into multiple number of sub-bands, which employ a different set of transmission technologies, such as multiple-input multiple-output (MIMO), packet scheduling, adaptive modulation and coding (AMC), and hybrid automatic repeat request (HARQ) depending on the traffic type. According to this concept, DiffSeg system, where the two-dimensional resource domain of OFDM is filled with four different resource units, called segments, and the occupancy ratio of each type of segment is determined by the segment map, has been recently proposed in [5][6]. In this paper, we introduce a concept of segment diversion to utilize the radio resource more efficiently considering the status of QoS. We then propose a system optimization model, which finds an optimal segment map and diversion ratios of real-time (RT) connections in order to maximize the residual resource for non-real-time (NRT) traffic while satisfying the minimum QoS requirement of RT traffic. We formulate an optimization problem using mixed integer programming, and then develop two computationally efficient algorithms: simplex-based heuristic and Maximum Diversion Rule. Especially, Maximum Diversion Rule is shown to achieve a near-optimal solution with dramatically less complexity.
