Time Delay Estimation Methods Based On Signal Models

Today’s wireless positioning technology has penetrated into every aspect of ourdaily lives. With the rapid development of electronic information technology andInternet technology, wireless location technology is evolving gradually at full range bythe simple applications at first towards positioning tracking, information security,communication services, information services, in-car entertainment and navigation andso on. With the coming of the era of the Internet Of Things nowadays, wireless locationtechnology affects and changes our production and life deeply all the time.Simply speaking, wireless location technology ascertains geometric position of themobile station by measuring the parameters of the communication information betweenthe estimated base station and mobile station. Commonly used parameters are Angle OfArrive(AOA), Signal strength Of Arrival (SOA), Time Of Arrival(TOA). What’s more,Time Difference Of Arrival(TDOA). Corresponding to the different parameters, thereare different positioning systems and algorithms. In recent years TDOA locationtechnology is developing rapidly and has already become a generally concernedquestion as well as the focus of extensive research in the industry. The major work ofthis article is the research of time difference estimation algorithm, which is based on thedifferent kind of models. We select the autoregression signal model as well as the chirpsignal model. The main contents are presented below:1. With the study of autoregressive model and its spectral estimation, we presentInnovation TDOA estimation method of autoregressive signals, which associatesinnovation of autoregressive signals with TDOA estimation and significantly improvespositioning accuracy and resolution when compared to the traditional generalized crosscorrelation method.2. The high-order statistics method is applied to autoregressive Innovation TDOAestimation model with the help of high-order statistics toolbox of Matlab software.There shows better results in positioning accuracy and resolution in low signal-to-noiseratio environment.3. Statistical properties of the cyclostationary signal from the special class of non-stationary signal are discussed, especially for chirp signal in the field ofcyclostationary signal. We extend the theory to cyclostationary domain and do somecomparison, analysis and research. Then we analyse Cyclic Cross CorrelationAmplitude TDOA Estimation method. As what we expect, the final simulation resultsshow good resistance to the harsh environment of low signal-to-noise ratio and theroutes of transmission with correlated noise.4. Cyclic cross correlation phase TDOA estimation method is proposed. With thedetermination of cyclic autocorrelation function and characteristics of the phasedifference of the cross-correlation function, it’s convenient to achieve the timedifference. Related to noise interference, the method also has good resistanceperformance, which is also better than the generalized cross correlation method.