A B C D E F G H I J K L M N O P Q R S T U V W X Y Z All
Alshehri, Abdullah Ali
- Optimal Rake Receiver for CDMA Communications Using Time-Frequency Analysis
Authors
1 Electrical Engineering Department, King Abdulaziz University KAU, Rabigh, SA
Source
Digital Signal Processing, Vol 3, No 5 (2011), Pagination: 210-214Abstract
RAKE receiver that is used in both uplink and downlink in CDMA and W-CDMA mobile communications systems is tapping the received signal several times in order to reduce the multipath fading. The conventional rake receiver considers only the time delays due to the arrival of the replicas of the original signal and neglects the effects of Doppler shifts due to the mobility of the receiver which proves to have a significant effect specially at high speed movement of the mobile receiver. At rake fingers the received signal is correlated with the local spread signature and summed to increase the energy of the transmitted signal to provide a bitter correct decision rule for the transmitted binary bits. In this paper instead of the tapping scheme that is used in the classical rake receiver we consider obtaining first an estimate of the channel parameters such as delays, Doppler shifts and attenuation factors who then were adoptively employed into the rake receiver. Our propose channel estimation scheme is based on using the joint time-frequency analysis of the received signal via the discrete evolutionary transform DET. The channel impulse response obtained from the time-frequency kernel leads to the spreading function of the channel which provides a good estimate of all channel parameters required for rake receiver.Keywords
Discrete Evolutionary Transformation DET, Multipath Channel, Time-Frequency Analysis, Spread Spectrum and RAKE Receiver.- Segmentation of Non-Stationary Signals based on the Characteristic Function of their Evolutionary Spectrum
Authors
1 Electrical Engineering Department, King Abdulaziz University KAU, Rabigh, SA
Source
Digital Signal Processing, Vol 4, No 9 (2012), Pagination: 389-392Abstract
Segmentation and separation of non-stationary signals is of great interest for many engineering fields and applications. In this paper we present a segmentation approach using the characteristic width function of the joint time-frequency evolutionary spectrum of non-stationary or multi-component signal. The proposed segmentation algorithm is based on the characteristic function of the time-frequency evolutionary spectrum which identifies the different segments of the signal due to its frequency change or any other dynamic changes that will change its statistical properties. The distribution of the energy density at the evolutionary spectrum provides the measures of these changes as a function of both time and frequency. Time-frequency representation of the signal is computed using the discrete evolutionary transform DET. At the experimental part we applied our algorithm for multi-component sinusoidal signal which has a wide range of use in many engineering applications. The obtained results at the segmentation level provide a clear and précised measures of the desired segments and their boundaries corresponding to their time domain.