Coherence bandwidth vs.coherence time

Coherence bandwidth vs.coherence time

concept

I understand that coherence bandwidth or delay spread is the difference between the ray arriving first and the one arriving last and also that coherence time or doppler spread is the time or bandwidth in which the impulse response of the channel remains constant.

• Coherence bandwidth and delay spread are inversely proportional. Any easy way to think about it is that the coherence bandwidth is the range of frequencies over which the frequency is flat, or has constant gain (no frequency selective fading). That is why in the figure on the right, the coherence bandwidth exists between two fades. In time, the delay spread (usually the RMS delay spread) gives you a sense of the multipath in the channel (copies of the transmitted signal that arrive at a receiver at different times and with different gains). RMS delay spread gives you a sense of the power distribution of paths over time that arrive at the receiver. A very small delay spread (one path) will result in a very large coherence bandwidth since there will be little or no frequency selectivity. If there is a lot of multipath, then the coherence bandwidth will be small and there will be a lot of frequency selective fading which will cause intersymbol interference (ISI).
• Think of coherence time as the time over which the channel (whichever that may be) remains relatively constant. This is generally very large in static conditions. Now consider a situation where the TX or RX are moving, the observed channel at the receiver will change much quicker with time. Or if the channel has many blockers, cars, buses, etc. the channel will change frequently. It’s a time measure of how long the channel taps/gain/phase remain constant. Doppler is spread is the inverse of this. Doppler spread describes the spectral broadening of a signal in the frequency domain. Imagine if you are in a moving car and the TX is stationary. If the TX sends a single tone, the RX will observe that frequency at some offset of the center TX frequency since the observed velocity of the signal is a difference between the speed of light and the speed of the mobile. Now in this situation, the general definition says that multipath will come from every and all directions, but the highest power of the spectral broadening will occur at -fD and +fD (direction of RX movement / opposite direction of RX movement). Take a look at Jake’s doppler. I think the graph on the left makes sense but isn’t very intuitive since it is trying to show time coherence between two large fades.

---

another understanding

So do I understand this correctly if I say that delay spread is the effect of obstructions that are fixed in terms of position whereas doppler spread is the effect of multipath from objects that are in motion?

On a different note you seem to mention the term “channel tap” which I have seen being mentioned in literature here and there as well but haven’t really managed to get a good understanding of. Could you perhaps expand a bit on it?

• Think of the wireless channel as a filter. Filters have discrete taps, like a tapped-delay line, where at different delays the channel has various gains and phase shifts. Page 7 of this link has a quick description of it as well. http://www.cse.wustl.edu/~jain/cse574-08/ftp/channel_model_tutorial.pdf
• In the most basic form, your received signal y(t) is a convolution of the channel (filter) and the signal that you send: $y(t)=h(\tau )\ast x(t)$ where $\ast$ is the convolution operator. $\tau$ is just the delays of the channel. A single LOS path channel will have a single delta, channel tap, whereas a NLOS channel will have many taps with various gains and phases at different delays. In frequency, this is Y(F)=H(F) times X(F) by the convolution theorem that states that convolution in time is multiplication in frequency. This is why coherence bandwidth and delay spread have reciprocal properties and same with Doppler Spread and coherence time.

–

思考

问题：所以MIMO的多径信道模型刻画的是一个即时的模型，还是刻画了一个时间段内的模型呢？

• https://zhuanlan.zhihu.com/p/67326348：通信最基本问题是：在信道的约束下，尽可能的提高传输速率。
  由于 传输速率 = 带宽 * 频谱利用率
  传输速率的问题可看作提高带宽，提高频谱利用率两个子问题。