Project Name:WHERE - Wireless Hybrid-Enhanced Mobile Radio Estimators 2
Research Programme: EU Seventh Framework Programme, Information and Communication
EC Instrument: STREP Project
Call : FP7-ICT-2009-4
Contract Number: 248894
Man-Power Effort: 684 Person-Months
Total Financial Volume:7.45 Million Euro
EC Funding: 5.26 Million Euro
Duration: 36 months (01.07.2010 – 30.06.2013)
The WHERE-2 project is a successor of the WHERE project and addresses the combination of positioning and communications in order to exploit synergies and to enhance the efficiency of future wireless communications systems. The key objective of WHERE2 is to assess the fundamental synergies between the two worlds of heterogeneous cooperative positioning and communications in the real world under realistic constraints. There are fundamental synergies between the two worlds of communications and navigation/positioning and for this reason the following essential questions arise:
The estimation of the position of mobile terminals (MTs) is the main goal in WHERE2. The positioning algorithms combine measurements from heterogeneous infrastructure and complement them by cooperative measurements between MTs, additional information from inertial sensors, and context information. Based on the performance of the geo-aided positioning strategies (in the sense of accuracy, complexity, overhead of signalling, reliability of the provided information, etc.) the impact on coordinated, cooperative, and cognitive networks is assessed. This is done under realistic scenarios and system parameters following on-going standardization processes. A joint and integrated demonstration using multiple hardware platforms provides a verification of the performance of dedicated cooperative algorithms.
In WHERE-2 Sigint contributes in both the positioning and communication aspects of the project. Regarding positioning Sigint is heavily involved with Radio Channel Modeling for providing realistic channel predictions for positioning purposes (either for the generation of fingerprinting databases or for testing advanced positioning algorithms). More specifically, Sigint’s 3D Ray Tracing simulator has been used to provide radio propagation predictions for the so called “synthetic environment” in order to test advanced “self-learning” techniques developed within the scope the project. These techniques try to reverse engineer the problem and extract the building layout from a set of impulse response measurements. Since the algorithms are currently being tested, Ray Tracing becomes an optimum solution for providing realistic impulse responses. The specific environment has been selected, because its 3DTruEM model was calibrated in the context of WHERE (Phase 1) using in-situ measurements.
In addition, 3DTruEM has been extended in order to provide UWB channel predictions. The so called “Band-Divided Ray Tracing Method” has been implemented and verified through measurements. In this technique, the desired UWB band is divided into a number of narrower bands which can be simulated using conventional Ray Tracing and the final UWB Impulse response is obtained by combining in the frequency domain the individual frequency responses of the various sub-bands. The UWB IR is the result of the inverse Fast Fourier Transform (iFFT) of the combined frequency response.
Sigint also studied the use of 3D Ray Tracing (RT) to construct radiomaps for WLAN Received Signal Strength (RSS) fingerprint-based positioning, in conjunction with calibration techniques to make the overall process device-independent. RSS data collection might be a tedious and time-consuming process and also the measured radiomap accuracy and applicability is subject to potential changes in the wireless environment. Therefore, RT becomes a more attractive and efficient way to generate radiomaps. Moreover, traditional fingerprint-based methods lead to radiomaps which are restricted to the device used to generate the radiomap and fail to provide acceptable performance when different devices are considered. We address both challenges by exploiting 3D RT-generated radiomaps and using linear data transformation to match the characteristics of various devices. We evaluated the technique through 3DTruEM simulations and measurements in a typical indoor environment achieving very good accuracy with various deterministic and probabilistic positioning algorithms such as the K-Nearest Neighbor, the Weighted K-Nearest Neighbor and the RSS-Bayesian Inference Method. The positioning accuracy results are shown the table below.
Sigint also acts as an overall integrator in the project in the demonstration phase. We coordinate the task that integrates the low level heterogeneous positioning platforms and algorithms together with a multi-technology (WiFi, WiMAX, UMTS etc.) communication platform in order to demonstrate the benefit of positioning in communication and also advanced positioning techniques which take advantage of a heterogeneous networking environment (LTE, Zigbee, WiFi, Sensor networks, WiMAX etc.). Sigint’s objective, is to demonstrate through this platform the execution of position-based Vertical Handovers and investigate potential benefits of it. The architecture of this demonstration is shown below: