The Research Of Indoor Positioning Method Based On Power Line

The known indoor positioning methods, based TOA, TDOA, AOA, RSSI, havebeen studied for many years. Some difficulties have not been able to overcome, butpertinent researches have been advanced. Because of some factors, such as cost, thescope of application and portability, existing indoor positioning technology can not meetthe needs of pervasive computing applications. As a new discipline, power linehigh-speed data communication technology is under development. It’s significative totry to combine with indoor positioning as an emerging field.Based on the existing power line communication technology, this researcherproposed power line-based indoor positioning technology, and made in-depth andsystematic research. Meanwhile, discussed the method to create a channel model and toimprove positioning accuracy. The model was created through a combination ofexperiments and experience. And solutions were proposed through artificial intelligence.It’s proved the feasibility of the program by algorithms compare and experimentalanalysis. The paper’s main work and innovations are as follows:1. Self-made signal transmitter and receiver, adopted power lines as the media, solvedthe problem of the realization of indoor positioning system based on power line. In theface, added signal isolation equipment and processing equipment, and found the rightfrequency radiation signal. The principle and process of the experiment was describedin detail. Recorded the signal strength of alignment position which was received by thereceiver, and create a database.2. Described the transformation from the free space propagation model to thelogarithm of the path attenuation model. According to RSS, discussed the logarithm ofthe path attenuation model and pointed out that the factor of the model parameters hasstrong dependence on the physical environment. In the result, the parameters weredifficulty to be determined. Therefore, it’s failed to put forward for the experimentalmodel, but generally analysis and discussion were carried out.3. Simply analyzed the collected signal strength data to observe their spatialcorrelation. And analyzed positioning error by traditional fingerprint localization method. The research of positioning error made the realization of the power line indoorpositioning possible. The data was processed by common KNN algorithm. Andobserved its positioning effect.4. Considering the problem of positioning accuracy for the signal strength, it’sproposed artificial neural network algorithm, including BP algorithm, RBF algorithm,CPN algorithm and PNN algorithm, to train massive signal strength data and establish amapping relationship to achieve the classification. Then input forecast data validationfor positioning. The experimental results showed that BP and CPN algorithm were lesseffective than the RBP and PNN algorithms in the matter of classification results.5. Started with the support vector machine and classified and predict the sample signalstrength data. The experimental simulation took advantage of the least squares supportvector machine algorithm and optimized the kernel function. Finally, the classifiedprediction was most effective. It’s showed that the positioning accuracy was improved.