#### DMCA

## Possibilities and fundamental limitations of positioning using wireless communication networks measurements (2005)

Venue: | University of Bath, U.K |

Citations: | 9 - 5 self |

### Citations

1733 |
A novel approach to non-linear and non-gaussian bayesian state estimation
- Gordon, Salmon, et al.
- 1993
(Show Context)
Citation Context ...e. In case of quite non-linear measurement equation h(pt), or very non-linear error distribution, the position estimate using the EKF is far from the CRLB. The computer intensive particle filter (PF) =-=[34, 35]-=- has been proposed for high-performance positioning [9]. The advantage is that all currently available information described in Section 2 is easily incorporated, including power attenuation and street... |

305 | Relative Location Estimation in Wireless Sensor Networks‖,
- Patwari, Hero
- 2003
(Show Context)
Citation Context ... Matrix (FIM) provides a fundamental estimation limit referred to as the Cramer-Rao Lower Bound (CRLB) [12]. This bound has been analyzed thoroughly in the literature, primarily for AOA, TOA and TDOA =-=[13, 14, 15, 16, 17, 18, 19, 20]-=-, but also for RSS [21, 22] and with specific attention to the impact from non-line-of-sight [23, 24]. The Fisher Information Matrix J(p) is defined as � �T d log pE(y − h(p)) d log pE(y − h(p)) J(p) ... |

258 | Adaptive Filtering and Change Detection.
- Gustafsson
- 2000
(Show Context)
Citation Context ...using merging the states as in Interactive Multiple Models (IMM), pruning as in Adaptive Forgetting through Multiple Models (AFMM) or motion parameter estimation as in Hidden Markov Models (HMM), see =-=[37]-=- for an overview. A performance lower bound when the number of states increases is the CRLB to the right in Figure 8 as a function of true mobility. 5. CONCLUSIONS Location in wireless networks is of ... |

219 | Particle filters for positioning, navigation, and tracking.
- Gustafsson, Gunnarsson, et al.
- 2002
(Show Context)
Citation Context ...and the remaining shadow fading component has typically std(e i t) ≈ 3 dB depending on the environment and spatial map resolution. The map can also be a commercial street map for automotive terminals =-=[9]-=- as examplified in Figure 1. A support to GPS for sea navigation was proposed in [10] using sonar depth measurements and a depth map. • Inertial sensors: yt = hINS( ˙pt, ¨pt)+et describe the terminal ... |

128 | Overview of radiolocation in CDMA cellular systems,”
- Stuber
- 1998
(Show Context)
Citation Context ...gle of arrival, (AOA) between MS and RP, or related to relative distances, such as the time of arrival (TOA), time difference of arrival (TDOA) and received signal strength (RSS). The survey articles =-=[1, 2, 3]-=- provide further information about positioning in wireless networks and associated standardization. Indoor positioning is discussed in[4], which includes its own specific challenges. While yellow page... |

127 |
Indoor Geolocation Science and Technology,
- Pahlavan, Li, et al.
- 2002
(Show Context)
Citation Context ... and received signal strength (RSS). The survey articles [1, 2, 3] provide further information about positioning in wireless networks and associated standardization. Indoor positioning is discussed in=-=[4]-=-, which includes its own specific challenges. While yellow pages services require only crude position estimates (such as cell identity), more advanced services like mobile gaming could benefit from ac... |

89 |
Standardization of mobile phone positioning for 3g systems,
- Zhao
- 2002
(Show Context)
Citation Context ...gle of arrival, (AOA) between MS and RP, or related to relative distances, such as the time of arrival (TOA), time difference of arrival (TDOA) and received signal strength (RSS). The survey articles =-=[1, 2, 3]-=- provide further information about positioning in wireless networks and associated standardization. Indoor positioning is discussed in[4], which includes its own specific challenges. While yellow page... |

89 | Positioning GSM telephones;
- Drane, Macnaughtan, et al.
- 1998
(Show Context)
Citation Context ...gle of arrival, (AOA) between MS and RP, or related to relative distances, such as the time of arrival (TOA), time difference of arrival (TDOA) and received signal strength (RSS). The survey articles =-=[1, 2, 3]-=- provide further information about positioning in wireless networks and associated standardization. Indoor positioning is discussed in[4], which includes its own specific challenges. While yellow page... |

63 |
Simple solutions for hyperbolic and related position fixes,”
- Fang
- 1990
(Show Context)
Citation Context ...eiver in Figure 2. If the transmitter is unsynchronized, each pair of receivers can compute a time-difference of arrival TDOA. This leads to a hyperbolic function where the transmitter can be located =-=[11]-=-. Four receivers can compute six such hyperbolic functions, which intersect in one unique point, see Figure 2. Y 3 2 1 0 −1 −2 Receiver locations and TOA circles −3 −3 −2 −1 0 X 1 2 3 Y 3 2 1 0 −1 −2 ... |

41 | Positioning using time-difference of arrival measurements,”
- Gustafsson, Gunnarsson
- 2003
(Show Context)
Citation Context ...arch methods for non-linear optimization are summarized in Table 3. Example 4 below illustrates typical performances. For further illustrations on computing the ML estimate for TDOA measurements, see =-=[25, 26]-=-. Example 4 Consider again the TDOA estimation problem in Example 1. The least squares loss functions in Figure 2 reveals that the non-linear equations have no local minima, so both a gradient and a N... |

35 |
Wireless position location: fundamentals, implementation strategies, and sources of error,”
- Krizman, Biedka, et al.
- 1997
(Show Context)
Citation Context ...raight-forward to introduce further measurements. This brief section exemplifies the measurements to illustrate 2sthe ideas. Details regarding practical implementations and standards are adopted from =-=[2, 3, 6, 7]-=-. Denote the two-dimensional mobile position at time t by pt = (Xt, Yt) T , and the known base station positions by p i = (Xi, Yi) T . The generic measurements relative reference point i, y i t are su... |

20 |
A maximum likelihood approach for estimating DS-CDMA multipath fading channels
- Ström, Malmsten
- 2000
(Show Context)
Citation Context ... Matrix (FIM) provides a fundamental estimation limit referred to as the Cramer-Rao Lower Bound (CRLB) [12]. This bound has been analyzed thoroughly in the literature, primarily for AOA, TOA and TDOA =-=[13, 14, 15, 16, 17, 18, 19, 20]-=-, but also for RSS [21, 22] and with specific attention to the impact from non-line-of-sight [23, 24]. The Fisher Information Matrix J(p) is defined as � �T d log pE(y − h(p)) d log pE(y − h(p)) J(p) ... |

17 |
On relation among time delay and signal strength based geolocation methods”, in
- Qi, Kobayashi
- 2003
(Show Context)
Citation Context ...it referred to as the Cramer-Rao Lower Bound (CRLB) [12]. This bound has been analyzed thoroughly in the literature, primarily for AOA, TOA and TDOA [13, 14, 15, 16, 17, 18, 19, 20], but also for RSS =-=[21, 22]-=- and with specific attention to the impact from non-line-of-sight [23, 24]. The Fisher Information Matrix J(p) is defined as � �T d log pE(y − h(p)) d log pE(y − h(p)) J(p) = E d p d p ⎛ = E ⎝ � d log... |

15 |
Posterior Cramer–Rao bounds for sequential estimation, in:
- Bergman
- 2001
(Show Context)
Citation Context ...e dynamic case in terms of the Cramer-Rao Lower Bound (CRLB) has been studied in [30, 31, 32] for the TDOA case. We here present a general derivation. The CRLB for non-linear filtering was derived in =-=[33]-=-. In short, they presented a recursion for non-linear models similar to the information filter version of the Riccati equation that computes a lower estimation bound Pt. The random walk model and velo... |

14 | Particle filter and Cramér-Rao lower bound for underwater navigation,” in
- Karlsson, Gustafsson
- 2003
(Show Context)
Citation Context ...on the environment and spatial map resolution. The map can also be a commercial street map for automotive terminals [9] as examplified in Figure 1. A support to GPS for sea navigation was proposed in =-=[10]-=- using sonar depth measurements and a depth map. • Inertial sensors: yt = hINS( ˙pt, ¨pt)+et describe the terminal movements relative a fixed coordinate system based on accelerometers, gyros, etc. . E... |

14 | Novel closed-form ML position estimator for hyperbolic location.
- Urruela, Riba
- 2004
(Show Context)
Citation Context ...arch methods for non-linear optimization are summarized in Table 3. Example 4 below illustrates typical performances. For further illustrations on computing the ML estimate for TDOA measurements, see =-=[25, 26]-=-. Example 4 Consider again the TDOA estimation problem in Example 1. The least squares loss functions in Figure 2 reveals that the non-linear equations have no local minima, so both a gradient and a N... |

14 | A non-line-of-sight mitigation technique based on ML-detection
- Riba, Urruela
- 2004
(Show Context)
Citation Context ...untered in two ways: • Use a robust error distribution. For instance, a TOA can have a long tail for positive errors. 12 44s• Include some logic in the search algorithm to rule out such outliers. See =-=[27, 28]-=- for some algorithms in the second category. In the FIM context, it is evident that a non-line-of-sight measurement does not add significant information. The associated measurement error standard devi... |

13 |
Cramér-Rao lower bound for geolocation in non-line-of-sight environment
- Qi, Kobayashi
- 2002
(Show Context)
Citation Context ...een analyzed thoroughly in the literature, primarily for AOA, TOA and TDOA [13, 14, 15, 16, 17, 18, 19, 20], but also for RSS [21, 22] and with specific attention to the impact from non-line-of-sight =-=[23, 24]-=-. The Fisher Information Matrix J(p) is defined as � �T d log pE(y − h(p)) d log pE(y − h(p)) J(p) = E d p d p ⎛ = E ⎝ � d log pE(y−h(X,Y )) d X � 2 d pE(y−h(X,Y )) d log pE(y−h(X,Y )) d X d Y In case... |

13 |
Survey of Maneuvering Target Tracking. Part I. Dynamic Models
- Rong, Vesselin
- 2003
(Show Context)
Citation Context ...andom walk, with or without velocity measurements, are summarized in Table 4. Any model suggested in the target tracking literature is plausible also for this application, see for instance the survey =-=[29]-=-, but we focus on the simplest cases. Higher order models can decrease estimation error slightly, but will not drastically change our conclusions. Random walk model: pt+1 = pt + Tswt Velocity sensor m... |

13 | Particle Filters for Positioning in Wireless Networks
- Norlund, Gustafsson, et al.
- 2002
(Show Context)
Citation Context ...ce positioning [9]. The advantage is that all currently available information described in Section 2 is easily incorporated, including power attenuation and street maps. Some examples are provided in =-=[36]-=-. 15sLMS ˆpt = ˆpt−1 + µH T (ˆpt−1)(yt − H(ˆpt−1)ˆpt−1) RLS ˆpt = ˆpt−1 + PtH T (ˆpt−1)(yt − H(ˆpt−1)ˆpt−1) Pt = 1 λ � Pt−1 − Pt−1H T (ˆpt−1) � λRt + H(ˆpt−1)Pt−1H T (ˆpt−1) � −1 H(ˆpt−1)Pt−1 EKF ˆpt ... |

11 |
Fundamentals of Signal Processing: Estimation Theory
- Kay
- 1993
(Show Context)
Citation Context ... 1 0 −1 −2 10 3 10 0.3 3 1 0.1 0.1 0.3s3.2. Fundamental Performance Bounds The Fisher Information Matrix (FIM) provides a fundamental estimation limit referred to as the Cramer-Rao Lower Bound (CRLB) =-=[12]-=-. This bound has been analyzed thoroughly in the literature, primarily for AOA, TOA and TDOA [13, 14, 15, 16, 17, 18, 19, 20], but also for RSS [21, 22] and with specific attention to the impact from ... |

9 |
A Report on Technical and Operational Issues Impacting The Provision of Wireless Enhanced 911 Services
- Hatfield
- 2002
(Show Context)
Citation Context ...le gaming could benefit from accurate and timely position estimates. Furthermore, the emergency call accuracy requirements by the Federal Communications Commission (FCC), have been tightened recently =-=[5]-=- to 300 m for 95% of the calls and 100 m for 67% of the calls using network-centric methods, and 150 m and 50 m for 95% and 67% of the calls, respectively, using mobile-centric methods. Therefore, it ... |

9 |
Cramer-rao bounds for the estimation of multipath parameters and mobiles’ positions in asynchronous ds-cdma systems
- Botteron, Host-Madsen, et al.
- 2004
(Show Context)
Citation Context ... Matrix (FIM) provides a fundamental estimation limit referred to as the Cramer-Rao Lower Bound (CRLB) [12]. This bound has been analyzed thoroughly in the literature, primarily for AOA, TOA and TDOA =-=[13, 14, 15, 16, 17, 18, 19, 20]-=-, but also for RSS [21, 22] and with specific attention to the impact from non-line-of-sight [23, 24]. The Fisher Information Matrix J(p) is defined as � �T d log pE(y − h(p)) d log pE(y − h(p)) J(p) ... |

5 | On geolocation accuracy with prior information in non-line-of-sight environment
- Qi, Kobayashi
- 2002
(Show Context)
Citation Context ...een analyzed thoroughly in the literature, primarily for AOA, TOA and TDOA [13, 14, 15, 16, 17, 18, 19, 20], but also for RSS [21, 22] and with specific attention to the impact from non-line-of-sight =-=[23, 24]-=-. The Fisher Information Matrix J(p) is defined as � �T d log pE(y − h(p)) d log pE(y − h(p)) J(p) = E d p d p ⎛ = E ⎝ � d log pE(y−h(X,Y )) d X � 2 d pE(y−h(X,Y )) d log pE(y−h(X,Y )) d X d Y In case... |

4 |
Effects of system and environment parameters on the performance of network-based mobile station position estimators
- Botteron, Host-Madsen, et al.
- 2004
(Show Context)
Citation Context |

4 |
Statistical theory of the effects of radio location system design parameters on the positioning performance
- Botteron, Fattouche, et al.
- 2002
(Show Context)
Citation Context |

4 |
Effect of biased measurement errors on accuracy of position location methods
- Koorapaty, Grubeck, et al.
- 1998
(Show Context)
Citation Context |

4 |
Barankin bounds for position estimation using received signal strength measurements
- Koorapaty
- 2004
(Show Context)
Citation Context ...it referred to as the Cramer-Rao Lower Bound (CRLB) [12]. This bound has been analyzed thoroughly in the literature, primarily for AOA, TOA and TDOA [13, 14, 15, 16, 17, 18, 19, 20], but also for RSS =-=[21, 22]-=- and with specific attention to the impact from non-line-of-sight [23, 24]. The Fisher Information Matrix J(p) is defined as � �T d log pE(y − h(p)) d log pE(y − h(p)) J(p) = E d p d p ⎛ = E ⎝ � d log... |

2 |
A novel estimator and theoretical limits for incar radio-location
- Urruela, Riba
- 2003
(Show Context)
Citation Context ...tion system in the measurement equations. 13 ⎞s4.2. Fundamental Performance Bounds Recently, location performance in the dynamic case in terms of the Cramer-Rao Lower Bound (CRLB) has been studied in =-=[30, 31, 32]-=- for the TDOA case. We here present a general derivation. The CRLB for non-linear filtering was derived in [33]. In short, they presented a recursion for non-linear models similar to the information f... |

2 | Efficient mobile location from time measurements with unknown variances in dynamic scenarios
- Urruela, Riba
- 2004
(Show Context)
Citation Context ...tion system in the measurement equations. 13 ⎞s4.2. Fundamental Performance Bounds Recently, location performance in the dynamic case in terms of the Cramer-Rao Lower Bound (CRLB) has been studied in =-=[30, 31, 32]-=- for the TDOA case. We here present a general derivation. The CRLB for non-linear filtering was derived in [33]. In short, they presented a recursion for non-linear models similar to the information f... |

1 |
Overview of 2G LCS technologies and standards
- Zhao
- 2001
(Show Context)
Citation Context ...raight-forward to introduce further measurements. This brief section exemplifies the measurements to illustrate 2sthe ideas. Details regarding practical implementations and standards are adopted from =-=[2, 3, 6, 7]-=-. Denote the two-dimensional mobile position at time t by pt = (Xt, Yt) T , and the known base station positions by p i = (Xi, Yi) T . The generic measurements relative reference point i, y i t are su... |

1 |
Field strength and its variability in VHF and UHF land-mobile radio service
- Okamura, Kawano, et al.
- 1968
(Show Context)
Citation Context ...ngth: y i t = hRSS(|pt − p i |) + e i t is averaged over fast fading, and depends on distance and slow fading. A model that solely depends on the relative distance is the so called Okumura-Hata model =-=[8]-=- y i t = K − 10α log 10(|pt − p i |) + e i t, where std(e i t) ≈ 4 − 12 dB depending on the environment (desert to dense urban). It is also possible to utilize a predicted or measured spatial digital ... |

1 | Cramer-Rao bound for location estimation of a mobile in asynchronous DS-CDMA systems
- Botteron, Host-Madsen, et al.
- 2001
(Show Context)
Citation Context |

1 |
Cramer-Rao bounds for time of arrical estimation in cellular systems
- Koorapaty
- 2004
(Show Context)
Citation Context |

1 |
Adaptive AR model based robust mobile location estimation approach in NLoS environment
- Zhen, Zhang
- 2004
(Show Context)
Citation Context ...untered in two ways: • Use a robust error distribution. For instance, a TOA can have a long tail for positive errors. 12 44s• Include some logic in the search algorithm to rule out such outliers. See =-=[27, 28]-=- for some algorithms in the second category. In the FIM context, it is evident that a non-line-of-sight measurement does not add significant information. The associated measurement error standard devi... |

1 |
A novel estimator and performance bound for time propagation and Doppler based radiolocation
- Urruela, Riba
- 2003
(Show Context)
Citation Context ...tion system in the measurement equations. 13 ⎞s4.2. Fundamental Performance Bounds Recently, location performance in the dynamic case in terms of the Cramer-Rao Lower Bound (CRLB) has been studied in =-=[30, 31, 32]-=- for the TDOA case. We here present a general derivation. The CRLB for non-linear filtering was derived in [33]. In short, they presented a recursion for non-linear models similar to the information f... |