Information Fusion Strategies for Collaborative Radio SLAM
Joshua Morales and Zak (Zaher) M. Kassas, University of California, Riverside
Navigation systems on today’s vehicles mainly rely on integrating global navigation satellite system (GNSS) receivers with an inertial navigation system (INS). This integration benefits from each systems’ complementary properties: the long-term stability and accuracy of a GNSS navigation solution and the short term accuracy of an INS. However, GNSS signals could become unavailable or unreliable in environments such as deep urban canyons or environments experiencing a malicious attack (e.g., jamming and spoofing). In this case, the integrated navigation system relies solely on the INS, whose errors accumulate and eventually diverge without an aiding correction. Recently, signals of opportunity (SOPs) have been considered to enable navigation whenever GNSS signals become inaccessible or untrustworthy [1, 2]. SOPs can be exploited to aid a vehicle’s INS in the absence of GNSS signals . With an appropriate fusion strategy, a team of collaborating vehicles can share their INS data and signals drawn from SOP transmitters to further reduce their INS errors.
SOPs (e.g., AM/FM radio , cellular , digital television , and iridium ) are abundant in GNSS-challenged environments and available at various geometric configurations, making them attractive aiding sources for a vehicle’s INS whenever GNSS signals become unavailable. However, unlike GNSS where the states of space vehicles (SVs) are readily available, the states of SOP transmitters (positions, clock biases, and clock drifts) may not be known a priori and must be estimated.
Estimating the states of unknown SOP transmitters can be done in one of two ways: (1) a mapping framework, where the states of unknown SOPs are mapped beforehand by receivers that have knowledge of their own positions and clock states or (2) a radio simultaneous localization and mapping framework (radio SLAM). The radio SLAM problem is analogous to the SLAM problem in robotics . Both problems ask if it is possible for an autonomous vehicle (AV) to start at an unknown location in an unknown environment and then incrementally build a map of the environment while simultaneously localizing itself within this map. However, in contrast to the static environmental map of the typical SLAM problem, the SOP signal map is more complex– it is dynamic and stochastic. Specifically, in radio SLAM with pseudorange-only observations, the AV must estimate not only its pose, but also the clock states of both the AV mounted receiver and the SOP transmitters.
In collaborative SLAM (C-SLAM), multiple AVs share their pose estimates and mutual observations on the environment in order to improve the quality of their individual state estimates and to build a larger and more accurate map [9, 10]. In collaborative radio SLAM (radio C-SLAM), multiple collaborating AVs simultaneously estimate their states (attitude, position, velocity, clock bias, and clock drift) and the states of all available unknown radio frequency (RF) transmitters by sharing mutual observations drawn from the transmitters. Specifically, in this work, INS data from each vehicle and pseudorange observations drawn from unknown SOP transmitters will be shared and information strategies that use either the signals’ time-of-arrival (TOA) or time-difference-of-arrival (TDOA) to provide INS corrections in the absence of GNSS signals will be compared and studied.
Collaborative SOP-aided INS was originally demonstrated via a centralized framework in  and was shown to achieve reliable and accurate navigation performance in the absence of GPS signals. In , a distributed framework was presented for aiding a team of AVs’ INSs using pseudoranges drawn from unknown SOP transmitters. However, only processing the SOP pseudorange measurements as TOA was considered. In this work two measurement processing approaches will be studied: (i) TOA and (ii) TDOA.
TOA and TDOA are not only common in radionavigation, but are also widely used for node localization in wireless sensor networks when the states of anchor nodes are known . When the clocks of anchor nodes are synchronized, the estimation performance of TOA and TDOA will be identical . This conclusion does not apply to the scenario considered in this paper, where three levels of complexity are introduced. The first pertains to the nature of radio SLAM, which is the unavailability of the SOPs’ states that must be simultaneously estimated. The second arises when practical systems are considered and is due to the unsynchronization of the AV-mounted receivers’ and SOPs’ clocks. The third is attributed to the AVs’ states becoming correlated after fusing mutual observations made on unknown transmitters. These complexities render information fusion strategies that use TOA or TDOA unequal in performance.
Consider an environment comprising multiple AVs and unknown SOPs. Each AV is assumed to have access to GNSS SV pseudoranges, multiple unknown terrestrial SOP pseudoranges, and INS data. Suddenly, GNSS pseudoranges become unavailable, and the AVs continue drawing pseudorange observations from the unknown SOP transmitters. Three radio C-SLAM strategies for navigating in the absence of GNSS signals are presented and compared. The first strategy consists of aiding the AVs’ on-board INSs with TOA measurements from SOPs. In the second strategy, the AVs’ on-board INSs are aided with TDOA measurements taken with reference to an SOP. This strategy differs from traditional TDOA approaches in two main ways: (1) the clock bias term is not completely eliminated from the measurement and (2) the foci of the hyperboloids defined by the TDOA measurements are simultaneously estimated with the hyperboloids’ intersection. The third strategy relies on aiding the AVs’ on-board INSs with TDOA measurements taken with reference to a specific AV-mounted receiver. This strategy is similar to the second strategy, with the added complexity that the foci of the hyperboloids (the positions of the AVs) are not stationary.
The contribution of this paper is to study and compare these three C-SLAM strategies. Specifically, this paper will establish analytically the strategy which guarantees the minimal estimation uncertainty of the AVs’ position states. The conditions under which these strategies produce identical estimates will be derived. Experimental results will demonstrate multiple unmanned aerial vehicles (UAVs) collaboratively aiding their INSs with cellular signals using each of the studied C-SLAM strategies.
 L. Merry, R. Faragher, and S. Schedin, “Comparison of opportunistic signals for localisation,” in Proceedings of IFAC Symposium on Intelligent Autonomous Vehicles, September 2010, pp. 109–114.
 Z. Kassas, “Analysis and synthesis of collaborative opportunistic navigation systems,” Ph.D. dissertation, The University of Texas at Austin, USA, 2014.
 J. Morales, P. Roysdon, and Z. Kassas, “Signals of opportunity aided inertial navigation,” in Proceedings of ION GNSS Conference, September 2016, pp. 1492–1501.
 S. Fang, J. Chen, H. Huang, and T. Lin, “Is FM a RF-based positioning solution in a metropolitan-scale environment? A probabilistic approach with radio measurements analysis,” IEEE Transactions on Broadcasting, vol. 55, no. 3, pp. 577–588, September 2009.
 Z. Kassas, J. Khalife, K. Shamaei, and J. Morales, “I hear, therefore I know where I am: Compensating for GNSS deficiencies with cellular signals.” IEEE Signal Processing Magazine, pp. 111–124, September 2017.
 P. Thevenon, S. Damien, O. Julien, C. Macabiau, M. Bousquet, L. Ries, and S. Corazza, “Positioning using mobile TV based on the DVB-SH standard,” NAVIGATION , Journal of the Institute of Navigation, vol. 58, no. 2, pp. 71–90, 2011.
 K. Pesyna, Z. Kassas, and T. Humphreys, “Constructing a continuous phase time history from TDMA signals for opportunistic navigation,” in Proceedings of IEEE/ION Position Location and Navigation Symposium, April 2012, pp. 1209–1220.
 H. Durrant-Whyte and T. Bailey, “Simultaneous localization and mapping: part I,” IEEE Robotics & Automation Magazine, vol. 13, no. 2, pp. 99–110, June 2006.
 J. Fenwick, P. Newman, and J. Leonard, “Cooperative concurrent mapping and localization,” in Proceedings of IEEE International Conference on Robotics and Automation, May 2002, pp. 1810–1817.
 A. Howard, “Multi-robot simultaneous localization and mapping using particle filters,” The International Journal of Robotics Research, vol. 25, no. 12, pp. 1243–1256, December 2006.
 J. Morales, J. Khalife, and Z. Kassas, “Collaborative autonomous vehicles with signals of opportunity aided inertial navigation systems,” in Proceedings of ION International Technical Meeting Conference, January 2017, 805–818.
 J. Morales and Z. Kassas, “Distributed signals of opportunity aided inertial navigation with intermittent communication,” in Proceedings of ION GNSS Conference, September 2017, accepted.
 N. Patwari, J. Ash, S. Kyperountas, A. Hero, R. Moses, and N. Correal, “Locating the nodes: cooperative localization in wireless sensor networks,” IEEE Signal Processing Magazine, vol. 22, no. 4, pp. 54–69, July 2005.
 D. Shin and T. Sung, “Comparisons of error characteristics between TOA and TDOA positioning,” IEEE Transactions on Aerospace and Electronic Systems, vol. 38, no. 1, pp. 307–311, January 2002.