Wi-Fi Based Proximity Detection System Design And Optimization

In this thesis, a feasible Wi-Fi based Proximity Detection (PD) solution wasproposed and successfully developed. This Wi-Fi based PD system can be applied inpractical PD projects and it provides a design reference for further system development.Firstly, some critical parameters and popular application areas of PD wereintroduced. And then several mainstream wireless technologies which may be suitablefor PD solution were discussed in detail. Among all the wireless technologies, Wi-Fi hasmany obvious advantages and very promising for PD applications. There are threemechanisms which can realize PD based on Wi-Fi signals, namely, tight timesynchronization mechanism, adaptive clocking technology-based mechanism, andmulti-antennae-based mechanism. In our work, to simplify the hardware system design,two-antennae-based mechanism was adopted. The whole PD system includes hardwarepart and software part. From the hardware aspect, the two Wi-Fi antennae were placedon the top-left and top-rightof devices’ screen respectively. When another PC isapproaching the testing PC, the two antennae can obtain the relative direction (left orright) of another PC by calculating the RSSI difference between. Through RSSI, we cangot both distance and direction data, and hence realize the fundamental functions of PD.However, when the distance between two devices is too large or the noise is tooheavy in the detectable range, the RSSI values of the antennas aren’t always satisfactory.In order to stabilize the original RSSI curves and make them more precise, Kalmanalgorithm was employed. Kalman algorithm is a popular filter algorithm, which has twobasic operations, one is the Propagation and the other is the Observation. Byalternatively executing the two operations, it can mitigate or even eliminate the randominference and noises, and then restore the real state of system. The whole procedure is“predict-actual measurement-revise”. Herein, Kalman Filter algorithm was used toprocess and optimize the original RSSI data measured by antennae. Comparing theoptimized data with original one, we found that the precision of RSSI values wereimproved and the data were more reliable.Besides, in order to provide proximity interfaces for various different applications (local application, internet browser application, etc.), we also designed the wholesystem, especially the PD service. API was defined and callback code was written onwindows platform. The PD application demo indicated that the designed PD can meetthe requirements for practical applications.