LTE OTDOA Positioning Performance Under Interference Conditions

T-Y. Chiou, Y-W. Ting, Y-C. Lin

Abstract:	The purpose of this work is to develop interference mitigation methods and demonstrate the LTE OTDOA positioning performance under interference conditions. In long term evolution (LTE) cellular system, the positioning reference signal (PRS) is transmitted in downlink for executing the method of observed time difference of arrival (OTDOA) positioning. OTDOA is a mobile-assisted positioning scheme, where the time difference measurements are performed in the user equipment (UE) and the UE’s position is calculated by the network with the use of the reported measurements from the UE. In [1], we had developed the signal processing architecture for executing the method of LTE OTDOA and presented the ranging performance of PRS. Hence, in this work, we like to present the final positioning performance. Furthermore, we consider the case of PRS to PRS interference; evaluate the positioning performance degradation due to PRS interference, and show the improvements after applying interference mitigations. In FCC’s E-911 Phase II regulations for handset based positioning methods, it requires that an UE’s location has to be determined to an accuracy of within 50 meters for 67 percent of emergency calls and 150 meters for 95 percent of calls. However, in the scenario with PRS interference, the availability of an accuracy of within 150 meters is degraded down to 92%. It is an evident that PRS interference can cause a performance failure. After applying the proposed mitigation methods, the availability of an accuracy of within 150 meters is improved to 96.3%. Although, the improvement is only 4%, it is a critical recovery in system availability. This paper depicts the modeling in LTE network deployment; PRS signal generating and downlink processing; PRS to PRS interference configuration; interference mitigation methods, and the final performance. The motivation of performing this work comes from the demands of having a cellular phone which fulfills the E911 requirements and also meets the quality of service in location based services (LBSs). Because of the low availability of GNSS signals in urban or indoor environments, people have been seeking alternative positioning methods for years. Given the high availability in indoor environments, the cellular signal is one of the desirable candidates among many terrestrial signals. Moreover, as becoming the dominant cellular network, LTE is also believed to provide reliable positioning results. Furthermore, the 3rd Generation Partnership Project (3GPP) has specified functional and performance requirements for supporting various location services (LCSs) in the standards. Thus, the developing of LTE is not only in favor of communication but also the positioning performance. The positioning methods available in the LTE standard are cell ID; enhanced cell ID; A-GNSS, and OTDOA. When A-GNSS is not available, the OTDOA method provides the most accurate position fix within the aforementioned LTE positioning methods. Therefore, we like to focus on investigating the performance of the OTDOA positioning in LTE. In LTE OTDOA positioning, the PRS signal is dedicated for positioning and for solving the hearability problem. Hence, the 3GPP standard defines PRS to guarantee the quality of positioning measurements and the capability of receiving PRS from sufficient numbers of base stations. PRS is a down link signal that is orthogonal frequency division multiplexing (OFDM). The architecture to process the received OFDM signals for obtaining the timing measurement, equivalent to the ranging measurement, is developed. The system level simulation is also developed to evaluate the final positioning performance. In this paper, we like to discuss the PRS to PRS interference. Due to the limited number of frequency reuse, it is unavoidable that two PRSs will share the same time-frequency resource elements. In consequence, interference occurs between two PRS signals. One of the solutions is to apply muting or time-division multiplexing so that the two PRS signals do not overlap in time domain. In 3GPP standard, the muting configuration is available. However, it may be only applied to mute the serving cell PRS. Thus, PRSs from different neighbor cells interfere with each other. Furthermore, if the network is asynchronous, it is hard to apply muting or time-division multiplexing. Thus, methods to mitigate the interference impacts are proposed. The methods include interference cancellation, adaptive detection threshold, temporal diversity, and quality check. The gain and cost of implementing the above methods are different. We have accomplished a cost effective solution for solving the PRS interference problem. To evaluate the proposed methods, we have developed both link level and system level simulations for LTE network. A network having 57 cells with inter-site distance = 1732 meters is considered (one of the standard scenarios in 3GPP standard). The channel model is Extended Typical Urban (ETU). When the UE is deployed uniformly in the network, each PRS measurement would have 8~9 interference sources since the PRS has the frequency reuse of 6. At each location, the received signal condition is calculated according to the transmission power, channel propagation loss, shadowing, and receiver noise. The signal-to-interference plus noise ratio (SINR) at the receiver antenna is then determined. Given the PRS signal condition from each base station, the link level simulator generates each PRS waveform in time domain to mimic the real condition in the field. The receiver in the link level simulator is actually a software receiver of a LTE modem. The receiver processed the generated PRS signal with steps of FFT, IFFT, coherent sum, non-coherent sum, acquisition, and time of arrival estimation. We first consider a scenario without interference but with Gaussian noise only as the baseline. Then, we include the PRS interference in the received waveform. The degraded positioning performance due to interference is then compared to the baseline performance. After applying the interference mitigation methods, we then evaluate the performance improvements. Given the PRS bandwidth=10 MHz (details of the conditions are depicted in the paper), the LTE OTDOA positioning performance are summarized as follows: Baseline case – Gaussian: 50 m with 86.2% availability, 150 m with 96.8% availability; Interference case: 50 m with 79.3% availability, 150 m with 92.3% availability, and Interference mitigated case: 50 m with 83.7% availability, 150 m with 96.3% availability. In summary, the demands of E911 and LBSs lead to the work of investigating alternatives to GNSS positioning method. The OTDOA positioning in LTE is a potential candidate for indoor location service. A fundamental architecture of signal processing for performance evaluation is the basis for thorough work in the future. From this work, we optimize the process of interference mitigation in the positioning reference signal in LTE. Finally, the positioning performance of the PRS in LTE is demonstrated to meet the E911 requirements in interference conditions. [1] Chiou, T-Y., Ting, Y-W., Lin, Y-C., "Ranging Performance of the Positioning Reference Signal in LTE Cellular System," Proceedings of the 25th International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS 2012), Nashville, TN, September 2012.
Published in:	Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013) September 16 - 20, 2013 Nashville Convention Center, Nashville, Tennessee Nashville, TN
Pages:	473 - 478
Cite this article:	Chiou, T-Y., Ting, Y-W., Lin, Y-C., "LTE OTDOA Positioning Performance Under Interference Conditions," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 473-478.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In