Adaptive multichannel sequential lattice prediction filtering method for range estimation in cognitive radios
Mehmet Tahir Ozden · 2014
A range estimation method for cognitive MIMO-OFDM radio systems is considered in this presentation. The range estimation problem is formulated as time delay estimation problem, and then adapted as multiple channel estimation problem. The multiple channels are estimated by using sequential processing multichannel Givens lattice stages (SPMGLSs). A complete modified Gram-Schmidt orthogonalization of multichannel input data is accomplished, matrix operations are avoided, and only scalar operations are used. A highly modular, regular, and Vertical Bell Laboratories Layered Space Time (V-BLAST) type radio receiver architecture, that has a suitable structure for software defined radio implementations, is achieved, such that the same receiver structure can easily be reconfigured as an equalizer or a spectrum sensing filter depending on mode of operation of cognitive radio. The dispersed cognitive radio spectrum is utilized by the multiple channels of the proposed method, the multiple time delay estimates are joined together by equal combining method to attain a single time delay estimate, and the range estimate is subsequently computed from the combined time delay estimate. The performance of the proposed method under timeinvariant and time-variant channel conditions, and also taking into consideration the inter-symbol interference (ISI) and interchannel interference effects (ICI) is presented by means of standard deviation and mean bias of range estimate vs. signal-to-noise ratio (SNR) curves, and compared with the Cramer Rao Lower Bound (CRLB) and with the performance of maximum likelihood (ML) method, implemented in each channel.