Authors: HoLung Hung ChiaHsin Cheng
Publish Date: 2013/12/11
Volume: 77, Issue: 2, Pages: 923-949
Abstract
In this paper the applicability of the subspacebased blind adaptive algorithm in multiple access ultrawideband systems is investigated However in the multiuser transmission environment multiple access interference becomes a serious issue and results in the degradation of system performance In order to overcome this shortcoming we propose a novel and low complexity decision mechanism termed the decision timing instant DTI A major advantage of the DTI algorithm is that it admits an adaptive implementation with low computational complexity instead of singular value decomposition In the present paper we exploit the joint blind multiuser detection in UWB systems a combined scheme is proposed which couples the minimummeansquareerror and the DTI subspace tracking algorithm under UWB timevariant channels Simulation results show that DTI is able to fast and precisely trace the variation of the channel environment and to improve the performance of the blind adaptive multiuser detection with a subspace approach over timevarying channelsUltrawideband UWB technology is currently regarded as an attractive solution for many wireless applications where high resolution reduced interference and propagation around obstacles are challenging 1 2 3 Also the research of UWB systems has recently attracted a significant interest in both the academic and industrial community 1 2 3 It utilizes ultrashort pulse shapes transmitted at a very low power spectral density PSD in compliance to the Federal Communications Commission FCC rules that defined UWB signals with a fractional bandwidth greater than or equal to 02 of the center frequency or a bandwidth of at least 500 MHz 4 In recent years much research work has been devoted to accelerating the acquisition process of UWB signals Based on different algorithmic approaches several fast acquisition techniques were proposed 5 6 7 8 9 10 However the complexity aspect was generally less emphasized than the algorithmic one Indeed the correlations are computed in the time domain and acquisition systems are fed by stream processing sample by sample irrespective of the search strategy 4 The corresponding architectures are thus not optimal and may require relatively long processing times under challenging conditions eg inside an underground mine gallery a peculiar harsh and infrequently studied environment 1 11 where lineof sight LOS ranging applications are highly desirable for safety and efficiency purposes However UWB technology as a physical support with excellent temporal resolution is very promising for such applications 12 With ongoing widespread deployment of the UWB communication systems reliable signal detection is desirable Most existing approaches employ correlators to correlate the received signal with a template signal 13 This technique appears very powerful but not so satisfactory in a multipath and multiple access channels Therefore both of the unknown multiple access interference MAI and the multipath distortion need to be mitigatedRecent works have focused on obtaining lowcomplexity algorithms for rapid timing acquisition by making use of coarse bin search and exploiting coded beacon sequences in conjunction with a correlator bank or subspacebased spectral estimation 14 15 Reference 16 17 proposes a semiblind synchronization scheme based on maximumlikelihood ML techniques to recover symbol and frame timing Semiblind schemes combine the methods of channel and timing acquisition based on pilot symbols and blind channel and timing acquisition derived directly from the data conveying signal However each of these approaches requires one or more of the following assumptions multipath is absent timehopping TH codes are slow or even absent the multipath channel is known and the system can afford the prohibitive complexity of exhaustively searching over thousands of bins chips Evidently timing algorithms based on these assumptions are impractical for most realistic UWB settings In addition the estimation of the timeof arrival TOA is a particular case of a timing acquisition problem whose maximum likelihood ML solution is known 18 but has strong practical limitations due to the requirement of very high sampling rates and complexity Although there are different ML approaches that manage to reduce complexity considerably 19 there still exist practical limitations for their use in positioning applications Efforts have been steered towards less complex solutions yet aiming at near optimum performance Most of them are based on time domain approaches Under the assumption of known undistorted received pulses optimal correlationbased TOA estimators 20 and a simplified version of the generalized ML criterion 13 are known Also energy based TOA estimation schemes have received considerable attention as a viable alternative to correlationbased methods 21 22 due to their reduced complexity implementation at subNyquist sampling rates Indeed they do not require expensive pulseshape estimation algorithms and represent a good solution for low power and lowcomplexity systems at the expense of ranging accuracy Hence In order to achieve a lowcomplexity receiver an UWB computationallyefficient acquisition system showing explicit design characteristics that offer greatly improved computational cost and acquisition time was proposed in this paperIn this paper we evaluate the performance of this new UWB fast acquisition system suggested for ranging over this peculiar timevarying channel For the THUWB systems the main focus of our work is on the study of the miltiuser detection In the multiuser transmission environment MAI becomes a serious issue and results in the degradation of the system performance To efficiently suppress MAI a subspacebased blind adaptive linear detector which was first proposed by Poor 23 is chosen Besides since the UWB systems are mainly operated in an indoor environment it also causes serious intersymbol interference ISI Moreover we consider a mobile UWB transceiver Changing of the transmission position will sometimes suffer the transition of the channel models because of different transmission distance or characteristics for example from CM1 to CM4 However a subspacebased blind adaptive linear detector proposed by H V Poor is highly sensitive to the sudden change of the channel environment over a timevarying channel eventually results in large performance degradation and a slow convergence speed to steady state as illustrated in simulation results It is because the subspace tacking algorithm is not update the signal subspace components rapidly under the changing environment In order to overcome this shortcoming we propose a low complexity decision mechanism termed the decision timing instant DTI 24 The proposed DTI method is mainly designed to accurately detect the change of the channel environment over a timevarying channel In following sections we will introduce the design theorem of the DTI method and then the combination of DTI method and subspacebased blind adaptive linear detector for THUWB systems is investigated as well The remainder of this paper is organized as follows In Sect 2 the system model is given Section 3 details the proposed UWB fast acquisition system with its implementation issues followed by Sect 4 that provides simulation results and finally Sect 5 concludes the paper
Keywords: