Browse by author
Lookup NU author(s): Dr Rajesh Tiwari, Emeritus Professor Satnam Dlay, Dr Hal Strangeways, Dr Kirill Palamartchouk
Full text for this publication is not currently held within this repository. Alternative links are provided below where available.
The electric field of a GNSS signal describes helix of a right hand screw in the direction of transmission, referred to as right hand circular polarisation (RHCP). The polarisation may change from its original RHCP (for direct line of sight) to left hand circular polarised (LHCP) in a multipath environment (change in polarisation due to reflected signals) which will depend on type of reflector and incident angles. The geodetic GNSS antenna is so designed that it accepts only RHCP signal. In a heavily built up area with many skyscraper buildings and particularly when reception takes place in a street canyon, the GPS receiver may not acquire the minimum satellites required for an accurate position determination. A GNSS receiver, signal acquisition is the first step to determine the visible satellites at a particular location. Several methods are used for acquisition such as the serial search method, parallel FFT-based circular correlation, the data folding method and zero padding (ZP). The parallel FFT-based circular correlation method is the most commonly used for acquisition due to its low computational cost. However, in a highly dynamic multipath environment, it fails to acquire a sufficient number of satellites. In order to improve the detection performance, Parallel Dual Window Based Acquisition (PDWBA) is proposed, based on the zero padding method. The PDWBA method uses a parallel overlapping window of received signals in the time domain rather than padding zeros as in the case of the ZP method, and then correlates the signal with the local PRN code generated and then up-samples it according to the PDWBA data points. This basically helps in increasing the coherence integration time and, due to dual window selection, the variance reduces during cross-correlation, similar to the approach in the Welch power spectrum technique. For evaluating the method, we collected real GPS signals in a weak signal environment with high multipath. The experimental equipment was placed in an open space in the middle of a six story trapezoid shaped building; the Merz Court containing the School of Electrical and Electronic Engineering at Newcastle University (54.978 N, 1.615 W), UK. Due to the building environment and the narrow open sky view of the receiver, the received GPS signals were mostly very weak due to a NLOS paths between the receiver and the GPS satellites. A dual polarisation (RHCP/LHCP) active L1/L2 GPS antenna from Antcom Corporation was used for GPS signal reception with two USRPs as RF frontends. An external clock reference at 10MHz, using a rubidium oscillator was applied to one of the USRPs and synchronized to the other using a MIMO cable. The inbuilt oscillator (TXCO) in the USRP for clock synchronization is only generic and not suitable for acquiring GPS signals in a highly dynamic environment. A geodetic choke-ring GNSS antenna was used to collect the GPS signals employing a NovAtel GSV 4004B receiver to enable the comparison between the GPS signal acquisition using the standard geodetic GPS receiver and a dual polarised GPS antenna enabled GPS software receiver. The data for the acquisition test was collected on 8 September, 2014 for a period of one hour (16:30 – 17:30 BST). The visible GPS SVs above 20 degree elevation (using the ephemeris file) over Merz Court (54.978 N, 1.615 W) were estimated and these were in fact PRNs 4, 11, 14, 19, 22, 27, 28 and 32. Our proposed algorithm, DWBA was used in the software receiver for signal acquisition received using the LHCP and RHCP polarisation antenna. Further, results were compared with the existing parallel FFT-based circular correlation and ZP (Zero Padding) methods. For the NovAtel receiver, the parallel FFT-based circular correlation was used in the signal acquisition and, when compared with the our PDWBA method, it was found that this method, together with the dual polarised antenna, manage to tracked the three PRNs 19, 22, and 4 whilst the NovAtel GSV 4004B could only manage to track the two PRNs 19 and 22. Based on the experimental data analysis, t was found that the detection probability and computational load of the new proposed PDWBA method are better method when acquiring weak signals and can be a very useful tool in improving the Time to First Fix (TTFF) for signal reception in a multipath environment.
Author(s): Tiwari R, Ahmad A, Dlay S, Strangeways HJ, Palamartchouk K
Publication type: Conference Proceedings (inc. Abstract)
Publication status: In Press
Conference Name: ION GNSS 2015
Year of Conference: 2015
Acceptance date: 09/04/2015
Publisher: Institute of Navigation
