Research on Spectrum Allocation and Antenna Mode Selection in Wireless Local Area Networks
|Course||Computer Science and Technology|
|Keywords||Dynamic Spectrum Access Spectrum Allocation Path Metric MultipleInput Multiple Output Antenna Mode Selection|
WLAN(Wireless Local Area Networks) has been a major access method of mobileInternet for its high speed,low cost and easiness to be deployed. With the exponentialgrowth of mobile Internet and mobile application, how to improve the efciency of lim-ited spectrum resource becomes a grand challenge. To address this issue, researcherstry to apply the technologies of dynamic spectrum access and MIMO (Multiple InputMultiple Output) into WLAN. This thesis uses the IEEE802.11WLAN as researchobjects and performance optimization as research goal, combining the technologies ofdynamic spectrum access and MIMO, and focuses on solving the issues of spectrumallocation and antenna mode selection from the perspective of network. The main re-search works and conclusions are as follows:1. The problem of spectrum allocation aiming at throughput optimization is stud-ied. A spectrum allocation algorithm based on association control and load is proposed.The proposed algorithm frst solves the near-far efect problem based on clustering the-ory, and then designs a spectrum allocation sub-algorithm based on queue length andsignal noise ratio to solve the load imbalance issue. Trace-driven simulations show thatthe proposed algorithm can efciently improve the system throughput.2. The problem of spectrum allocation aiming at joint optimization throughputand delay is studied. A stable and efcient solution based on Lyapunov optimizationmethod for rate control and spectrum allocation is proposed. Both theoretical analysisand simulation results show that the proposed solution can optimize system throughputand guarantee upper bound of delay simultaneously.3. The problem of path selecting based on spectrum allocation is studied. Anew path metric is proposed, and a new path selecting protocol based on the metricis implemented. Simulation results show that the proposed protocol can improve TCPperformance and reduce protocol overhead.4. The throughput problem of adaptive selection of MIMO antenna mode is stud- ied. Three MIMO antenna mode design guidelines are proposed through large numbersof experiments under diferent environments and scenarios. Based on these guidelines,a simple and efcient MIMO antenna mode selecting algorithm is proposed and imple-mented. Realistic experiment results show that the proposed algorithm can efcientlyimprove the system throughput.5. The energy problem of MIMO antenna mode selecting is studied. An energy-aware MIMO antenna mode selecting algorithm is proposed. Trace-driven simulationsshow that the proposed algorithm can improve the energy efciency of MIMO senderprofoundly without signifcantly afecting throughput.