摘要:
A CDMA demodulator which is improved in interference canceling effect even when signals are sent from many communicators. The demodulator receives signals sent from a plurality of communicators and having a frame structure in which a pilot signal having a known pattern is inserted between information signals at regular time intervals. The demodulator includes interference cancellers which demodulate the data transmitted from the communicators, and estimate interference signals. The demodulator estimates the variation of each path of the channels, performs the demodulation of received data, and estimates the interference signal replica. The errors of the interference signal replica decrease as the data pass the interference cancellers, producing a marked interference canceling effect. When the interference cancellation is executed after detecting the reception level of each channel from the total sum of the received electric power of the paths of the channels and determining the ranks of the channels in order of reception level, a further marked interference canceling effect is exhibited.
摘要:
A weight coefficient is feedback-controlled so as to maximize the ratio of desired signal power to interference power (SIR), whereby it is possible to improve reception quality and to increase the number of subscribers per cell and it is also possible to increase the speed of establishment of diffused code synchronization and the speed of convergence of the weight coefficient. A coherence adaptive diversity arrangement to which absolute synchronization detection is applicable is adopted. The weighht coefficient of each diversity branch is adaptively feedback-controlled, whereby it is possible to reduce the influence of interference power from other communicators, so that the number of subscribers per cell can be increased. Furthermore, a direction in which a highest reception SIR is obtainable is determined by rotating an antenna with directivity of predetermined angle being given to it. After that, by setting the initial value of the weight coefficient to the direction, the speed of convergence of the weight coefficient can be increased.
摘要:
Each of receiving spreaded signals from K users via L channels for each user is de-spreaded with a spreading code allocated to each user. Interference components among signal components in the de-spreaded signal vector Z obtained are eliminated by a de-correlation filter (33) and pilot signals are detected by a pilot detection part (34) from the interference-eliminated signal vectors for multiple frames. Each of the detected pilot signals is divided by the pilot signal's known symbol to obtain a transfer function in a transfer function interpolation part (35) and a transfer function between consecutive pilot signals is estimated from the transfer functions of the consecutive pilot signals. A de-spreaded signal vector Z is supplied to a de-correlation filter (36) and interferences among signal components are eliminated and transmission distortion is also eliminated using the estimated transfer functions. Regarding the KL signal components obtained in the above process, L signal components corresponding to each user are composited by a phase compensation and composition part (37) so that K composite signals are generated. Those composite signals are level decided by a decision part (38) to regenerate K symbols from K users.
摘要:
A determinant, which represents the relationship between a sequence of code division multiplexed transmission symbol vectors b (k) from L communicators and a sequence of despread output vectors y (k) of received signals taking into account the cross correlation between spreading code sequences of transmitted signals, is limited within the range of ±g symbol duration about symbol timing k, and despread outputs of received signals within the limited range are inputted into 2g+1-stage shift register 24. On the other hand, partial correlation matrixes R k+h (1), R k+h (0) and R k+h (-1) for each symbol duration within the limited range are calculated by a partial correlation computing part 26 with respect to h = -g, ..., g , then a correlation matrix R k between the spreading code sequences corresponding to the L communicators is derived from the partial correlation matrixes in an inverse matrix computing part 27 and an inverse correlation matrix R k ⁻¹ of the correlation matrix is calculated. The inverse correlation matrix R k ⁻¹ and a sequence Y k of the despread outputs y (k) from the shift register are multiplied by a multiplier 25 to obtain an estimated symbol vector sequence.