TY - GEN
T1 - Practical issues in parameter estimation over fading channels with TBMA wireless sensor networks
AU - Gao, Ping
AU - Tepedelenlioglu, Cihan
PY - 2006
Y1 - 2006
N2 - Some practical issues in parameter estimation over fading channels with type-based multiple access (TBMA) sensor networks is studied in this paper, where the parameter of interest is estimated through the empirical distribution of the observations. In the literature, the asymptotically optimal estimator requires the transmitted signal waveforms to be orthogonal to represent different observations. Meanwhile, the channels between the sensors and the fusion center need to be non zero-mean. If the channel is zero-mean, channel state information (CSI) at sensor side is required. However, in practice, imperfect CSI, or interference between the orthogonal waveform dimensions, such as synchronization error, cannot be avoided, and may be unknown at the fusion center. How these practical issues affect the asymptotic optimality of the existing estimator and its performance are discussed in this paper. A unified framework for estimation with these practical constraints is proposed, which includes the imperfect CSI and synchronization error as special cases. The conditions under which the asymptotically optimal estimator requires the knowledge of the channel or interference statistics at the fusion center are derived. Furthermore, how the interference affects the performance of the asymptotically optimal estimator is characterized. Simulation results corroborate our analysis.
AB - Some practical issues in parameter estimation over fading channels with type-based multiple access (TBMA) sensor networks is studied in this paper, where the parameter of interest is estimated through the empirical distribution of the observations. In the literature, the asymptotically optimal estimator requires the transmitted signal waveforms to be orthogonal to represent different observations. Meanwhile, the channels between the sensors and the fusion center need to be non zero-mean. If the channel is zero-mean, channel state information (CSI) at sensor side is required. However, in practice, imperfect CSI, or interference between the orthogonal waveform dimensions, such as synchronization error, cannot be avoided, and may be unknown at the fusion center. How these practical issues affect the asymptotic optimality of the existing estimator and its performance are discussed in this paper. A unified framework for estimation with these practical constraints is proposed, which includes the imperfect CSI and synchronization error as special cases. The conditions under which the asymptotically optimal estimator requires the knowledge of the channel or interference statistics at the fusion center are derived. Furthermore, how the interference affects the performance of the asymptotically optimal estimator is characterized. Simulation results corroborate our analysis.
UR - http://www.scopus.com/inward/record.url?scp=50849154445&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=50849154445&partnerID=8YFLogxK
U2 - 10.1109/GLOCOM.2006.981
DO - 10.1109/GLOCOM.2006.981
M3 - Conference contribution
AN - SCOPUS:50849154445
SN - 142440357X
SN - 9781424403578
T3 - GLOBECOM - IEEE Global Telecommunications Conference
BT - IEEE GLOBECOM 2006 - 2006 Global Telecommunications Conference
T2 - IEEE GLOBECOM 2006 - 2006 Global Telecommunications Conference
Y2 - 27 November 2006 through 1 December 2006
ER -