| Visible light communication (VLC) has attracted much attention and inspired significant research activities as the rapid development of green source i.e. light emitting diode (LED) and conventional wireless communication can not satisfy the increasing demand of indoor access. LED is viewed as one of the most promising technology area in21th century. Compared with traditional illuminance sources, LED presents mang charming advantages, such as small size, reduced weight, long life, high brightness, steadily improved power efficiency and etc. With the reducing in manufacture cost, LED is expected to dominant the illuminance market. On the other side, over congested radio frequency spectrum and over high energy consumption make it almost impossible that traditional radio frequency (RF) systems independently satisfy the constantly increasing demand for wireless data transmission, such as Mobile TV and CMMB. Utilizing higher frequency source has become a common idea in academy and industry. For above reasones, indoor infrastructure-based visible light communication (VLC) was proposed to provide indoor wirelss coverage. This technique is viewed as the hybrid achievement of conventional optical communication and mature radio frequency (RF) wireless communication. More and more reserchers regard it as a forward technique with unique characteristic.Up to now, some key problems of visible light communication need to be solved, which mainly includes:unsatisfying transmission rate due to the LED modulation bandwidth, lacking basic research-of visible light indoor channel modeling, absence of an effective way to improve indoor visible light signal distribution and etc.With the channel modeling of indoor VLC, the research is focused on dynamic signal quality variance among different receiver locations within indoor envirioment. Moreover, highly adaptive optimization scheme is proposed. The main research content of this dissertation is presented as follows.Firstly, a fast and comparably accurate channel modeling scheme named independent reflecting elements interaction characterization (IREIC) is proposed and researched as existed schemes can not realize efficiency and reliability simultaneously. This scheme classifies the modeling of diffuse link into three stages and makes full use of calculated interaction information between fixed reflection elements and realizes remarkable saving in computation time at a cost of limited permanent memory compared with existed recursive modeling. In addition, the scheme can be used to analysis the received optical power, illuminance, and is not restricted in impulse Fesponse. The proposed scheme not only inherits the advantages of conventional deterministic scheme but also satisfy the actual need of computation efficiency. When surface segmentation accuracy of environment is equal, about34%modeling efficiency improvement can be realized. On the modeling accuracy, with reference to the classical quantitative results of traditional recursive solution, respective variation of direct or reflected signals in the percent of overall impulse response less than1%which is acceptable.Secondly, the research is focused on obvious performance variation of VLC signal power in different locations and optimization scheme with basic algorithm is proposed to realize uniform VLC signal coverage distribution on the communication floor. The performance of VLC system is mainly limited by the characteristics of optical signal and multipath propagation channel with intense signal quality turbulence among various locations. As VLC system has to provide illuminance function as well, the signal turbulence will affect the realization of uniformity illuminance. In this dissertation, a tailored evolutionary algorithm (EA) is proposed to dynamically modify the relative optical intensity of transmitters for obtaining uniform received power&illumination distribution. Simulation results illustrate that within three considered distributed lighting configuration, the optimization performance is not significantly influenced by the variation of LED array pattern. Respective reduction range of PDP is11.9%～14.3%,14.6%～16.1%and16.0%～17.2%. In adaptability of FOV, optimal performance of power fluctuation reductions usually appear as FOV is about50°. Furthermore, the impact to bandwidth and Root Mean Square (RMS) delay spread is eatimated and proved to be negligible or acceptable.Thirdly, this dissertation researches the optimization scheme and comparative algorithms of signal-to-noise (SNR) distribution at the receivers of VLC system. SNR is the key quantitative index to determine the communication performance and the main factor to determine the error bit rate (BER) level. This thesis extends the research direction of power distribution optimization by modifying the mentioned scheme design to adapt to the destination of SNR optimization. Through repeative computer simulation, the optimization performance of this scheme and related impact are quantitative analyzed. The analysis shows that this scheme can always generate uniform SNR&illuminance distribution even in different scenario settings. Furthermore, the advantages of proposed scheme in convergence rate, adaptation and robustness are emphasized by comparative experments. Through further analysis of three configurations, the best optimization performance is obtained as FOV=50°. Respective reduction of the SNR deviation is about25.9%,31.0%and30.7%.Fourthly, this dissertation also researches the impact brought by varable receiver alignment to above optimization schemes. In mentioned deterministic modeling and optimization schemes, the alignment conditions of receivers are simplied for convenience of analysis. In actual cases, the receiver is orientated vertically upwards with certain randomized probability. For quantitative analyzing the performance gain of mentioned optimization algorithm under the tight condition of receivers with randomized direction, this thesis resets the related parameters of receivers and makes repeative experiments through Monte Carlo solutions. Simulations show that when the randomly pointing probability of the receiver is0.8within±15°, optimization scheme can still provide27.2%reduction of SNR deviation for light source configuration of16(4X4) LED array. |
PDF Full Text Request