Return to Session E3 Next Abstract

Session E3: All-source Intelligent PNT Method

Exploiting On-Demand 5G signals for User Equipment-Based Navigation
Ali A. Abdallah and Zak (Zaher) M. Kassas, University of California, Irvine
Date/Time: Thursday, Sep. 22, 8:35 a.m.

The fifth-generation (5G) cellular signals is considered to have a major role in the navigation module of automated driving systems (ADSs). Network-based approaches have promised/demonstrated sub-meter-level positioning accuracy using 5G signals [1–3]. Those approaches in particular require the user to be a subscriber in the network in order to utilize the downlink/uplink channels between the 5G bases station (also known as gNodeB (gNB)) and the user equipment (UE), which compromises the user privacy by revealing their accurate location and limits the user to only the gNBs of the network they are subscribed to. In light of that, UE-based approaches have been studied recently and showed meter-level positioning accuracy on ground and aerial vehicles in different environments utilizing sub-6 GHZ existing infrastructure in the United States [4–8]. However, unlike previous cellular systems, 5G applies an ultra-lean transmission policy, which minimizes the transmission of “always- on” signals and limits UE-based opportunistic navigation (OpNav) to only synchronization signals. To give a closer idea about this limitation, as the possible 5G downlink bandwidth ranges between 4.32 to 397.44 MHz, the synchronization signals span a bandwidth that ranges between 3.6 to 57.6 MHz. In other words, only 14.5% of the bandwidth is being exploited for OpNav. Having higher signal bandwidth is highly attractive for signals-based navigation as it enables differentiating multipath from line-of-sight (LOS), resulting in a better accuracy in time-of-arrival estimation).
This paper proposes a UE-based approach; thus, it only considers the 5G downlink signal, which uses orthogonal frequency division multiplexing (OFDM) with cyclic prefix (CP) as a modulation technique. A 5G frame has a duration of 10 ms, which consists of 10 subframes each with a duration of 1 ms. Each subframe breaks down into numerous slots, each of which contains 14 OFDM symbols for a normal CP length. The subcarrier spacing in 5G is flexible and is defined as based on a pre-defined numerology. Each subframe is divided into numerous resource grids, each of which has multiple resource blocks with 12 subcarriers. A resource element is the smallest element of a resource grid that is defined by its symbol and subcarrier number.
The 5G frame contains two synchronization signals that can be exploited for navigation: Primary synchronization signal (PSS) and secondary synchronization signal (SSS). Which are two orthogonal maximal length sequences of length 127. PSS has 3 possible sequences and specifies the sector ID of the gNB, and SSS has 336 possible sequences, which specifies the group identifier of the gNB. Together they provide frame start time and the gNB physical cell ID NCell
ID . The physical broadcast channel (PBCH) demodulation reference signal (DM-RS) is also transmitted in the same symbols as the synchronization signals. Where all together forms what is called as SS/PBCH block. The length of the block is 240 subcarriers.
A state-of-the-art carrier-aided code phase-based 5G software-defined receiver (SDR) has been developed to extract navigation observables from the known “always-on” 5G downlink signals in [5, 7]. The developed SDR exploits the known synchronization signals (SSs) simultaneously and proposes what is a so-called ultimate SS (USS) to utilize the time-domain orthogonality of the 5G downlink signals. The USS is nothing but the 5G frame with a normalized SS/PBCH and zeros everywhere else. The performance of this approach is limited by the USS bandwidth ratio versus the entire downlink bandwidth and the duty factor, which limit the delay and carrier phase estimation accuracies, respectively [9]. For different configurations, the aforementioned metrics range between 14.5%–36% and 0.0104%–5.33%, respectively.
The deterministic approach requires knowing the signal structure, specifically the RSs. To overcome this requirement, a 5G CON framework was proposed in the literature for the goal of exploiting all available RSs including the ones unknown by the UE [10]. The proposed framework succeeded in estimating a periodic RS that was used cognitively to track 5G signals and navigate using them. The questions here is: How much of the available resources does the cognitively-acquired RS capture compared to the deterministic (i.e., USS)? Given that the OFDM frame start time is unknown in the CON framework, the only way to assess the acquired signal is to look at the narrowness of the normalized autocorrelation function (ACF) of both RSs, which gives an estimate of the bandwidth that is being exploited. The results in [10] showed a CON bandwidth ratio of 25% compared to 36% from USS. To exploit all available RSs in the 5G downlink, the CON framework faces major limitations such as
• The acquisition in the CON framework is challenged by the propagation channel fading and stationarity, which limits the coherent processing interval (CPI), i.e., the time interval in which the Doppler, delay, and channel gains are considered constant. Short CPI means less resources to be captured in the cognitively-acquired signal.
• The CON framework requires the UE to be in motion to exploit multiple gNBs transmitting at the same channel. Yet, to do so, the CON framework uses Doppler subspace to differentiate between gNBs; thus, the framework acquires only the most powerful gNB among different gNBs with similar Doppler profile. This results in acquiring less gNBs than the deterministic approach.
• The 5G frame start time remains unknown in CON framework; hence, it is not possible to construct the frame structure of the acquired signal. To this end, pre-filtering and power allocation of different RSs cannot be performed, which significantly changes the fidelity of the acquired signal.
This paper utilizes the best of both deterministic and cognitive approaches and makes the following contributions
• Discuss the 5G downlink signal and present a suitable model for exploiting the entire bandwidth.
• Develops a deterministic-CON (DeCON) framework that aims to maximizing the bandwidth ratio and the duty factor by exploiting other periodic RSs in the 5G downlink signals that are not known by the UE such as: the channel state information RS (CSI-RS); other DM-RSs for the physical downlink control channel (PDCCH) and the physical data shared channel (PDSCH); and the phase tracking RS (PTRS).
• Presents a pre-processing algorithm to suppress noise and interference and maintain equally-distributed power among different RSs.
• Validates experimentally the proposed framework and acquire a so-called ultimate reference signal (URS) which spans the entire 5G bandwidth and achieves significantly more accurate code and carrier phase measurements compared to the state-of-art SDR.
• Demonstrates, to the best of the authors’ knowledge, the first mobile user-based navigation experiment while exploiting the entire 5G downlink bandwidth.
It is worth mentioning that it is guaranteed that the USS is always transmitted in the 5G downlink signal; hence, it is safe to assume it as a prior to cognitively acquire more OFDM resources. Having this prior (i) elongates the CPI, (ii) utilizes the USS subspace to exploit all available gNBs (even gNBs with similar Doppler profile), and (iii) allows preprocessing of the acquired replica to suppress noise and interference and maintain equally-distributed power among different RSs. Figure 2 depicts the block diagram of DeCON.
To assess the potential of the proposed framework, a preliminary study was performed, in which a stationary National Instrument (NI) universal software radio peripheral (USRP-2955 was equipped with a consumer-grade omnidirectional Laird Antenna to receive 5G downlink signals. The bandwidth was set to 10 MHz and the carrier frequency was set to 632.55 MHz, which corresponds to T-Mobile U.S. cellular provider. The collected data was stored on a laptop for off-line processing. The USRP was set to record 5G signals over a period of 300 seconds. The USS is used to detect nearby gNBs as in [7]. After the tracking loops achieved lock to the received signal, DeCON acquired the URS signal for 4 seconds.
To study URS’s spectral efficiency and duty factor of the resulting URS, the number of active subcarriers and symbols was obtained from the preprocessed URS. Assuming that a URS symbol is active if 10 or more subcarriers are active within that symbol results in having 32 active symbols; hence, rT,URS = 22.86% compared to rT,USS = 2.86%. For the bandwidth ratio, Fig. 4 shows that rB,URS = 100% compared to rB,USS = 36% and rB,CON = 25%. The advantage of this increase in bandwidth ratio can be seen in the narrowness of the URS-ACF as shown in Fig. 5, which gives more resolution in time-domain to discriminate the LOS from multipath components.
The paper will extend these preliminary experimental results to improve the framework design and characterize the navigation performance of the proposed navigation framework on both ground and aerial vehicles.

References
[1] N. Garcia, H. Wymeersch, E. Larsson, A. Haimovich, and M. Coulon, “Direct localization for massive MIMO,” IEEE Transactions on Signal Processing, vol. 65, no. 10, pp. 2475–2487, May 2017.
[2] M. Koivisto, M. Costa, J. Werner, K. Heiska, J. Talvitie, K. Leppanen, V. Koivunen, and M. Valkama, “Joint device positioning and clock synchronization in 5G ultra-dense networks,” IEEE Transactions on Wireless Communications, vol. 16, no. 5, pp. 2866–2881, May 2017.
[3] Qualcomm, “Demonstrating advanced 5g innovations [video],” https://www.qualcomm.com/news/onq/2021/06/27/demonstrating-advanced-5g-innovations, June 2021.
[4] A. Abdallah, J. Khalife, and Z. Kassas, “Experimental characterization of received 5G signals carrier-to-noise ratio in indoor and urban environments,” in Proceedings of IEEE Vehicular Technology Conference, April 2021, pp. 1–5.
[5] A. Abdallah and Z. Kassas, “Opportunistic navigation using Sub-6 GHz 5G downlink signals: A case study on a ground vehicle platform,” in Proceedings of IEEE European Conference on Antennas and Propagation, 2021, submitted.
[6] Z. Kassas, A. Abdallah, and M. Orabi, “Carpe signum: seize the signal – opportunistic navigation with 5G,” Inside GNSS Magazine, vol. 16, no. 1, pp. 52–57, 2021.
[7] A. Abdallah and Z. Kassas, “UAV navigation with 5G carrier phase measurements,” in Proceedings of ION GNSS Conference, September 2021, pp. 3294–3306.
[8] L. Chen, X. Zhou, F. Chen, L. Yang, and R. Chen, “Carrier phase ranging for indoor positioning with 5G NR signals,” IEEE Internet of Things Journal, 2021, early Access.
[9] A. Graff, W. Blount, P. Iannucci, J. Andrews, and T. Humphreys, “Analysis of OFDM signals for ranging and communications,” in Proceedings of ION GNSS Conference, September 2021, pp. 2910–2924.
[10] M. Neinavaie, J. Khalife, and Z. Kassas, “Cognitive opportunistic navigation in private networks with 5G signals and beyond,” IEEE Journal of Selected



Return to Session E3 Next Abstract