Ber-driven resource allocation for multimedia communication over downlink OFDMA networks / Tham Mau Luen
Recent years have witnessed the growing pervasiveness of multimedia services due to significant advancements of both wireless access and multimedia compression technologies. On one hand, orthogonal frequency-division multiple access (OFDMA) is the key enabler for high-speed multiuser communications....
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2015
|
| Subjects: | |
| Online Access: | http://studentsrepo.um.edu.my/7575/4/KHA100062.pdf http://studentsrepo.um.edu.my/7575/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Recent years have witnessed the growing pervasiveness of multimedia services due to significant advancements of both wireless access and multimedia compression technologies. On one hand, orthogonal frequency-division multiple access (OFDMA) is the key enabler for high-speed multiuser communications. On the other hand, nonscalable and scalable multimedia coding standards offer superior compression gains as well as strong adaptabilities to channel variations. The inherent error-prone nature of channel environments, the wireless resource scarcity along with the variations in the importance among multimedia packets, however, render the quality-of-service (QoS) provisioning for multimedia communication over wireless networks a challenging task. Leveraging multiple target bit-error-rates (BERs) in concert with resource allocation to optimize the received multimedia quality is largely unexplored. Based on the principle that more important packets should be assigned stricter target BERs, this thesis proposes three BER-driven (BRA) resource allocation methods for transmitting pre-encoded multimedia over downlink OFDMA networks, targeting at generic, nonscalable and scalable bitstreams. For the generic case, each bitstream is associated with a different static target BER. Under this constraint, the impacts on three major classes of resource allocation scheme are analyzed from an optimization perspective. For both nonscalable and scalable cases, the target BER of each packet is dynamically adjusted based on scheduling, subcarrier assignment, bit and power allocation, channel quality, and importance level of that packet. The calculation of packet-importance metric takes into account decoder-based error concealment, where both simple and complex techniques are investigated. An additional constraint of strong decoding dependency is considered for the scalable case. Simulation results show that BRA schemes significantly outperform existing BER-unaware techniques in terms of spectral efficiency, power efficiency and decoded multimedia quality. Further test evaluates the suitability of equal power allocation which is the common assumption in the resource-allocation literature. |
|---|