公开(公告)号：US06356926B1

公开(公告)日：2002-03-12

申请号：US09284752

申请日：1999-06-17

申请人： Tore André

发明人： Tore André

IPC分类号： G06F1714

CPC分类号： G06F17/142

摘要： The present invention relates to a device and a method for calculating FFT, IFFT or like transforms with the aid of a calculating unit and at least two memories. The calculating unit reads input values from the memory positions in the memories, performs a group of calculations in at least one calculation step, and writes the output values obtained with these calculations into memory positions in the memories. The invention comprises the steps of reading-in and immediately storing the input values for calculation in different groups of calculations. When all input values necessary for a current group of calculations have been read in and the calculations in a preceding group of calculations have terminated at least one first calculation step, calculations in the current group of calculations are commenced. The output values from the current group of calculations are stored intermediately until the output values from the preceding group of calculations have been written, and are then written back into the same memory positions as those from which the input values were read, whereafter the output values can function as new input values. This enables at least two groups of calculations to be commenced at one and the same time.

摘要翻译： 本发明涉及借助于计算单元和至少两个存储器来计算FFT，IFFT或类似变换的装置和方法。 计算单元从存储器中的存储器位置读取输入值，在至少一个计算步骤中执行一组计算，并将通过这些计算获得的输出值写入存储器中的存储器位置。 本发明包括在不同的计算组中读入并立即存储用于计算的输入值的步骤。 当已经读入当前计算组所需的所有输入值，并且前一组计算中的计算已经终止至少一个第一计算步骤时，开始计算当前的计算组。 来自当前计算组的输出值被中间存储，直到前面的计算组的输出值被写入，然后被写回与读取输入值的存储位置相同的存储位置，之后输出值 可以作为新的输入值。 这使得至少两组计算能够在同一时间开始。

公开(公告)号：US07010277B1

公开(公告)日：2006-03-07

申请号：US09807924

申请日：1999-10-13

申请人： Tore André ,  Uno Torstensson

发明人： Tore André ,  Uno Torstensson

IPC分类号： H04B1/04

CPC分类号： H04B15/04 ,  H04B1/123

摘要： The present invention relates to an apparatus for cancellation of electromagnetic interferences (n), including means (3) for dividing a reference signal (sc) into a first and second signal (scl, scQ) having different phases, tow multipliers (4, 5) for each multiplying a weight coefficient to one of the phase displaced signals (scl, scQ) each and for outputting a multiplied signal each, and adder (6) for adding the two multiplied signals and for outputting a corrected reference signal (sn), and a subtracter (7) for subtracting the corrected reference signal (sn), from an input signal (sd) and for outputting an error signal (se). According to the invention different weight coefficients (l(f), Q(f)) are used for different frequencies or frequency bands (f) of the interference (n).

摘要翻译： 本发明涉及一种用于消除电磁干扰（n）的装置，包括用于将参考信号（s）分为第一和第二信号（s）的装置（3） 具有不同相位的两个乘法器（4,5），每个乘法器将权重系数与一个相移位置信号相乘（s，c， 并且用于每个输出相乘的信号，以及用于相加两个相乘信号并用于输出校正的参考信号（s）的加法器（6），以及 减法器（7），用于从输入信号（s）中减去校正的参考信号（s），并且用于输出误差信号（s＆lt;＆gt; / SUB>）。 根据本发明，不同的权重系数（1（f），Q（f））用于干扰（n）的不同频率或频带（f）。

公开(公告)号：US06757834B2

公开(公告)日：2004-06-29

申请号：US09826962

申请日：2001-04-06

申请人： Torbjörn Randahl ,  Albin Johansson ,  Anders Söderbärg ,  Tore André ,  Allan Olson

发明人： Torbjörn Randahl ,  Albin Johansson ,  Anders Söderbärg ,  Tore André ,  Allan Olson

IPC分类号： G06F132

CPC分类号： H04L25/028 ,  H03G3/004 ,  H04L25/0272

摘要： To minimize power dissipation in a line driver (3) in a central office (CO) for driving a DSL connection to a network terminal (NT) with a predetermined maximum constellation size, the central office (CO) supplies the line driver (3) with a predetermined supply voltage during a training sequence, and transfers data with a predetermined power spectral density in all available channels. The network terminal (NT) measures the signal-to-noise ratio of each channel, and transfers information to the central office (CO) about channels having signal-to-noise ratios that enable transfer of data. The central office (CO) reduces the power spectral density to a lowest possible value that still enables transfer of data in all channels having signal-to-noise ratios above a predetermined threshold as set by the maximum constellation size, calculates a sum of power to be transmitted in said channels, converts said sum into a corresponding line driver output voltage, and reduces the line driver supply voltage in correspondence thereto.

摘要翻译： 为了最小化用于以预定的最大星座大小驱动到网络终端（NT）的DSL连接的中心局（CO）中的线路驱动器（3）中的功率消耗，中心局（CO）提供线路驱动器（3） 在训练序列期间具有预定的电源电压，并且在所有可用信道中以预定功率谱密度传送数据。 网络终端（NT）测量每个信道的信噪比，并将关于具有使能传输数据的信噪比的信道的信息传送到中心局（CO）。 中心局（CO）将功率谱密度降低到最低可能的值，该最小可能值仍然能够传输具有由最大星座大小设定的高于预定阈值的信噪比的所有信道中的数据， 在所述通道中发送，将所述和转换成相应的线路驱动器输出电压，并且相应地降低线路驱动器电源电压。

公开(公告)号：US06690736B1

公开(公告)日：2004-02-10

申请号：US09555668

申请日：2000-06-01

申请人： Tore André

发明人： Tore André

IPC分类号： H04K110

CPC分类号： H04L5/0046 ,  H04L5/0007 ,  H04L5/006

摘要： A quality factor is associated to each carrier within a multicarrier system wherein a loading constant is computed as the quotient of a difference and the number of carriers that have no bits allocated, the difference being the difference of the sum of the bit error rates for carriers that have no bits allocated and the total remaining number of data bits that are not allocated, and allocating a number of bits to a carrier that has no bits allocated, the number of bits being the difference of the quality factor and the loading constant rounded to nearest integer, repeating the allocation until a predetermined fraction of the carriers have bits allocated, and reiterating the above steps until all carriers have bits allocated.

摘要翻译： 质量因子与多载波系统中的每个载波相关联，其中加载常数被计算为差分的商和没有分配比特的载波的数量，差值是载波的误码率之和的差异 没有分配比特和没有分配的总剩余数据比特数，并且分配多个比特到没有分配比特的载波，比特数是质量因子和加载常数的差值四舍五入到 重复分配，直到载波的预定分数具有分配的比特，并且重复上述步骤，直到所有载波具有分配比特。