Time domain self-adaptive equalizer and decision feedback filter included therein

Abstract

The invention relates to a time domain self-adaptive balancer, wherein input signal flows to a forward filter which output flows into an input of an adder, another two inputs of the adder is the output of a first feedback filter and the output of a second feedback filter, the adder adds the input signals, which output is connected with the input of a judger, an error generator according to the input and output of the judger generates an error signal, a tapping factor refresher according to the error signal generates a tapping factor to refresh the tapping factor of the forward filter and the judgment feedback filter. The first feedback filter uses a rotary structure, and the second feedback filter uses a non-rotary structure, which input has some time delay relative to the input of the first feedback filter. The stage numbers of the rotary structure and non-rotary structure are decided according to the length of last path.

Description

technical field [0001] The invention relates to a time-domain adaptive equalizer, in particular to a time-domain adaptive equalizer based on delay matching. Background technique [0002] The equalizer is a very important component in the practical application of many different digital information such as voice, data and video communication. Equalizers are used as echo cancellers (compensators) for loudspeakers in full-duplex communications, video deghosting in digital television or digital cable transmission, signal conditioners for wireless modems and wireless phones, etc. In the process of signal transmission, due to the existence of multipath signals in the channel, it will bring inter-symbol interference (ISI), and inter-symbol interference is an important cause of errors. In most single-carrier digital applications, the time Domain Adaptive Equalizer to correct ISI errors. [0003] FIG. 1 is a simplified block diagram of a conventional time-domain adaptive equalizer 1...

AI Technical Summary

Problems solved by technology

In this way, compared with the non-transposed structure, the consumption of hardware resources of the transposed structure will increase

If there are many stages of the feedback filter, in order to save hardware resources, it is necessary to limit the number of bits of the input signal, which not only limits the realization of blind equalization, but also is a major obstacle to the realization of multi-constellation point modulation.

[0007] To sum up, the hardware resource overhead of the non-transpose structure is less, based on the development of FPGA technology and ASIC technology, it is no longer a problem, but the timing relationship is tense

The output of the transposed structure is the same as that of the non-transposed structure, which eases the timing relationship, but the hardware resource overhead is very large

[0008] Therefore, the traditional time-domain adaptive equalizer cannot resist the long multipath in the channel transmission, especially the long and strong backward multipath

Images