Highly Accurate and Efficient Wireless Indoor Positioning System

在这个项目中，我们开发了一种新的算法Peytech的室内定位和定位。公司为此开发了硬件和软件。然而，现有的软件在精度和处理时间方面不能满足客户的要求。现有的硬件有可能收集足够的信息，用于此目的，本项目的目标是设计一种新的基于机器学习的算法，以满足本地化的限制和要求。该算法将是先驱的混合方法，利用可用的接收信号强度（RSS）和到达时间（ToA）进行有效定位。 拟议的项目有五个主要步骤：1）在软件中对环境要求进行建模，2）创建一种创新方法，该方法最佳地利用从传感器收集的数据，3）开发一种算法，该算法在期望的处理时间内以所需的精度估计物体位置，4）将该算法部署在公司的软件中，测试并反馈任何可能的修改，（5）最终报告的编写。该项目使客户能够利用他们的生产阶段和他们的仓库与一个可靠的和具有成本效益的跟踪方法。因此，在生产的早期或后期阶段进行有效的本地化和定位过程对制造商有利。这最终将不仅使该行业受益，而且使这些设施的用户受益。参与该项目的HQP将接触到一个实际的工业问题，需要机器学习，数据分析和本地化技术的知识和实施。

In this project we develop a new algorithm for Peytech for indoor localization and positioning. The company has developed its hardware and software for such purpose. However, the existing software is not satisfying the desired customers' requirements when it comes to accuracy and processing time. The existing hardware has the potential to gather enough of information for this purpose and the goal of this project is to design a novel machine learning based algorithm to fulfill the localization constraints and requirements. The algorithm will be the pioneer hybrid approach that takes advantages of both available Received Signal Strength (RSS) and Time of Arrival (ToA) for efficient positioning. The proposed project has five main steps: 1) Modeling the environments requirements in software, 2) Creating an innovative approach that optimally employs the data gathered from the sensors, 3) Development of an algorithm that estimates the object positions with the required accuracy during the desirable processing time, 4) Deploying the algorithm in the company's software, test and feedback for any possible modifications, 5) Final report preparation. This project enables the costumers to utilize their production stages and their warehouse with a reliable and cost effective tracking method. Consequently effective localization and positioning process at the early or ending stages of production benefits the manufacturers. This will eventually benefit not only the industry but also the costumers of these facilities. The HQP involved in this project will be exposed to a practical industrial problem that requires knowledge and implementation of machine learning, data analysis, and localization techniques.