Refine
Document Type
- Article (10)
- Conference Proceeding (9)
- Book (1)
Keywords
- floorplan (1)
- particle filter (1)
- probalistic sensor fusion (1)
- step detection (1)
- turn detection (1)
Institute
Comparison of 2.4 GHz WiFi FTM- and RSSI-Based Indoor Positioning Methods in Realistic Scenarios
(2020)
In this work we present a method to estimate an indoor position with the help of smartphone sensors and without any knowledge of absolute positioning systems like Wi-Fi signals. Our
system uses particle filtering to solve the recursive state estimation problem of finding the position of a pedestrian. We show how to integrate the information of the previous state into the weight update step and how the observation data can help within the state transition model. High positional accuracy can be achieved by only knowing that the pedestrian makes a foot step or changes her direction together with floor map information.
The Smartphone-Based Offline Indoor Location Competition at IPIN 2016: Analysis and Future Work
(2017)
While connection speeds are increasing slowly, some ISPs mention plans about possible traffic limitations in the near future which would keep internet traffic expensive. In addition, Green IT became more important especially over the last few years. Besides on-demand content like video live streams, HTTP traffic plays an important role. Thinking of textual web content, compression quickly comes to mind as a possibility to reduce traffic. The current HTTP/1.1 standard only provides gzip as an option for content encoding. HTTP/2.0 is under heavy development and numerous new algorithms have been established over the last few years. This paper analyzes HTTP traffic composition on a production server and concludes that about 50% is compressible. It further examines the effectiveness of custom and existing algorithms with
regards to compression ratio, speed, and energy consumption. Our results show that gzip is a sound choice for web traffic but alternatives like LZ4 are faster and provide competitive compression ratios.
With the steadily increasing need and wish to travel, people often have to reach locations they have never been to before. Modern means of transportation, like cars, ships and planes, thus come equipped with onboard navigation systems, assisting with this task, based on the global positioning system (GPS), or a derivative. However, the navigation task is not solely limited to outdoor environments. Reaching the correct gate within an airport, finding a ward in an unknown hospital, or the auditorium within a new university, represent navigation problems as well. With the GPS requiring a direct line of sight towards the sky, it is unavailable for absolute location estimation indoors. Therefore, the question for suitable indoor navigation techniques arises. Besides localization accuracy, additional factors should be met for such a new system to become a success. It should be easy to set up and maintain, limiting required working hours and costs. Likewise, hardware for the users themselves should be cheap, and readily available.
Due to the ubiquity of smartphones, these devices represent a desirable platform for pedestrians, backed by the variety of sensors installed in these devices. Within this work, smartphone-based
pedestrian indoor localization and navigation is discussed in detail. This covers examining the suitability of several available sensors: step-detection using readings from the accelerometer,
relative turn-detection utilizing the turn rates of the gyroscope, absolute heading estimations based on the magnetometer’s indications, and altitude evaluation from the barometer. While all aforementioned sensors do not require any additional infrastructure, thus suitable for all sorts of buildings, they only allow for relative location estimations. Absolute localization can utilize Wi-Fi, as it is supported by all smartphones, and most public buildings already contain the required infrastructure. Due to the behavior of radio signals, the smartphone’s current location can be approximated by examining signal strengths of nearby transmitters. This aspect is often utilized by Wi-Fi fingerprinting, which, however, requires a time consuming setup process. Therefore, an alternative is developed that allows for significantly faster setup times. Additionally,
the building’s 3D floorplan is included, modeling potential pedestrian movements, limiting impossible walks to improve estimation results, and to provide routing towards a desired destination. For this, two spatial floorplan representations are derived and examined. All aforementioned aspects are hereafter combined probabilistically, using recursive density estimation based on the particle filter. This allows for fusioning all sensor observations while respecting their individual uncertainties, and the building’s floorplan as additional constraints.
To summarize, the system described within this work covers probabilistic 3D pedestrian indoor localization, using commodity smartphones, contained sensors, a building’s existing infrastructure
and floorplan, all combined by the particle filter to derive an indoor localization and navigation system that is easy to set up and maintain.