Software Receiver Design

This blog is mainly based on 《Software Receiver Design》.

Step1: The Big Picture

1 A Digital Radio

1.1 What Is a Digital Radio?

First, at the heart of any working receiver is a structure that is able to function in the ideal case. The classic approach to receiver design (and also the approach of Software Receiver Design) is to build for the ideal case and later to refine so that the receiver will still work when bad things happen.
Second,many of the basic ideas are clearer in the ideal case.

The job of the receiver is to take the received signal (such as that in Figure
1.2©) and to recover the original text message.

One fundamental limitation to data transmission is the trade-off between the
data rate and the bandwidth.

A common theme throughout Software Receiver Design is that a solution to
one problem often causes another!

Step2: The Basic Components

2 A Telecommunication System

2.1 Electromagnetic Transmission of Analog Waveforms

There some experimental(physical) facts that cause transmission systems to be constructed as they are. First, for efficient wireless broadcasting of electromagnetic energy, an antenna needs to be longer than about 1/10 of a wavelength of the frequency being transmitted.

1. the relationship between wavelength and frequency;

A second experimental fact is that electromagnetic waves in the atmosphere exhibit different behaviors, depending on the frequency of the wave.
2.by FT、DFT、FFT to analyze the frequency of analog wavesforms.

2.2 Bandwidth

The various definitions of different bandwidth.

1. absolute bandwidth
2. 3-db bandwidth
3. null-to-null bandwidth
4. power bandwidth

2.3 Upconversion at the Transmitter

s(t)=w(t)cos(2·pi·f₀t)
Analyze the spectrum of the process of modulation(upconversion).
Whatever the modulation scheme used, the idea is the same: a sinusoid is used to translate the message into a form suitable for transmission.

2.4 Frequency Division Multiplexing

FDM
TDM
CDM

2.5 Filters that Remove Frequencies

Perhaps the most important property of the Fourier transform is the duality between convolution and multiplication,which says that

• convolution in time<——>multiplication in frequency,and
• multiplication in time <——>convolution in frequency.
  Various linear filter:

1. LPF
2. BPF
3. HPF

2.6 Analog Downconversion

By downconversion and LPF to recover the original waveforms.
Introduce the issue of carrier synchronization.

2.7 Analog Core of a Digital Communication System

No transmission system operates perfectly.
For instance,

• the bandwidth of a filter may be different from its specification(e.g.,the shoulders might not drop off fast enough to avoid passing some of the adjacent signal),
• the frequency of an oscillator might not be exact,and hence the modulation and/or demodulation might not be exact,
• the phase of the carrier is unknown at the receiver,since it depends on the time of trval between the transmitter and the receiver,
• perfect filters are impossible,even in principle,
• no oscillator is perfectly regular,there is always some jitter in frequency.

Multipath interference can be viewed as a sum of weighted and delayed versions of the transmitted signal.

2.8 Sampling at the Receiver

The analog-to-digital conversion(A/D) is known as sampling.
Two of the chief design issues in a digital receiver are the following.

• Where should the signal be sampled?
  The sampling

1. could be done at the input to the receiver at a rate proportional to the carrier frequency,(if carrier frequency is high then it needs very high sampling rate to sample,it is very expensive.)
2. could be done after the downconversion, at a rate proportional to the rate of the symbols,or(need to be accurate,it is expensive.)
3. could be done at some intermediate rate.

• How often should the sampling be done?
  Nyquist reconstruction principle.

2.9 Digital Communications Around an Analog Core

Digital (relative to fully analog) communications have the following advantages:

• digital circuits are relatively inexpensive,
• data encryption can be used to enhance privacy,
• digital realization tends to support greater dynamic range,
• signals from voice, video, and data can be merged for transmission over a common system,
• digital signals can be easily compressed
• noise does not accumulate from repeater to repeater over long distance,
• low error rates are possible, even with substantial noise,
• error can be correct via coding.
• reconfigurable,upgradable

Disadvantages of digital communications:

• more bandwidth is (generally) required than with analog,
• synchronization is required.

2.10 Pulse Shaping

The spectrum of the pulse shape must be bandlimited.
Unfortunately, the sinc is not entirely practical because it is doubly infinite in time. In any real
implementation, it will need to be truncated.

2.11 Synchronization：Good Times and Bad Times

Synchronization may occur in several places in the digital receriver.

• Symbol phase synchronization–choosing when (within each interval T) to sample.
• Symbol frequency synchronization-- accounting for different clock (oscillator) rates at the transmitter and receiver.
• Carrier phase synchronization-- aligning the phase of the carrier at the receiver with the phase of the carrier at the transmitter.
• Carrier frequency synchroniazation–aligning the frequency of the carrier at the receiver with the frequency of the carrier at the transmitter.
• Frame synchronization–finding the “start” of each message data block.

The point of view adopted in Software Receiver Design is that many of these synchronization tasks can be stated quite simply as optimization problems. Accordingly, many of the standard solutions to synchronization tasks can be viewed as solutions to these optimization problems.

• The problem of clock(or timing) recovery can be stated as that of finding a timing offset t to maximize the energy of the received signal. Solving this optimization problem via a gradient technique leads to a standard algorithm for timing recovery.
• The problem of carrier phase synchronization can be stated as that of finding a phase offset a to minimize a particular function of the modulated received signal. Solving this optimization problem via a gradient technique leads to the phase-locked loop, a standard method of carrier recovery.
• Carrier phase synchronization can also be stated using an alternative performance function that leads directly to the Costas loop, another standard method of carrier recovery.

2.12 Equalization

The problem of equalizer design can be stated as simple optimization problem, that of finding a set of filter parameters to minimize an appropriate function of the error, given only the received data (and perhaps a training sequence).

2.13 Decisions and Error Measures

P36 Figure 2.13 PAM system diagram is important.

Average squared error can be used as in Chapter 13 to adjust the parameters of an equalizer when the source message is known.

The error between the soft decisions and the associated hard decisions is used in Section 13.4 as a way to adjust the parameters in an equalizer when the source message is unknown, as a way of adjusting the phase of the carrier in Section 10.5, and as a way of adjusting the symbol timing in Section 12.3.

No matter what the exact form of the error measure, the ultimate goal is the accurate and efficient transmission of the message.

2.14 Coding and Decoding

the theory of information
channel capacity
source coding–Huffman code
linear block codes

2.15 A Telecommunication System

P36 Figure 2.13

2.16 Stairway to Radio

Introduce the structure of this book.

This chapter illustratives the modulars and structure of the communication system and its functions,and some indicators that are used to depict the communication system.

3 The Six Elements

3.1 Finding the Spectrum of a Signal

关于Matlab仿真部分，自己首先要搞懂函数的功能、语句的作用、自带函数的作用，然后要保证阅读速度，对于自己提出的假设要积极验证。
有一个体会是，自己要多读多看，不浪费自己的精力。现在我有一个问题，该怎样平衡好学习过程中的阅读与实践？其实自己的每一个问题对于自己都是有意义的，自己有责任与义务为自己解答这些问题，可以去尝试，可以去请教。

3.2 The First Element:Oscillators

Generally,there is an amplifier and a feedback circuit that returns a portion of the amplified there ia an amplifier and a feedback circuit that returns a portion of the amplified wave back to the input.When the feedback is aligned properly in phase,sustained oscillations occur.

3.3 The Second Element: Linear Filters

This section provides an overview of how to implement simple filters in MATLAB.
MATLAB 中firpm（）的用法
While the idealized spectra are completely flat in the passband, the actual ones are rippled.While the idealized spectra completely reject the out-of-band frequencies,the actual ones have small energy at all frequencies.

Exercise3.12 噪声与时延

Exercise 3.13 带外信号未滤除干净

Exercise 3.14 不会

Exercise 3.15 通过滤波器之后，信号幅值有衰减，功率降低，其half-power bandwidth 自然会增加。

3.4 The Third Element: Samplers

The Fourier’s result says that any signal that is periodic in time has a spectrum that consists of a collection of spikes uniformly spaced in frequency. Analogously,any signal whoes spectrum is periodic in frequency can be represented in time as a colletion of spikes uniformly spaced in time,and vice versa.

Periodic in Time <=> Uniform Sampling in Frequency.
Uniform Sampling in Time <=> Periodic in Frequency.

If the spectrum of x[k] is bandlimited to 1/m of the Nyquist rate,then downsampling by m loses no information.Otherwise,aliasing occurs.

Exercise 3.16 aliasing 当满足Nyquist抽样定理时，信号频谱正常；当不满足Nyquist定理时信号出现混叠现象。f=50,51,60Hz出现aliasing

Exercise 3.17 Nyquist抽样定理 Ts=1/50,1/90,1/100出现aliasing

Exercise 3.18 f=33,43Hz出现aliasing

3.5 The Fourth Element: Static Nonlinearities

Linear system such as filters cannot add new frequencies to a signal, though they can remove unwanted frequencies.Nonlinearities such as squaring and quantizing can and will add new frequencies.These can be useful in the communication system in a variety of ways.

Exercise 3.20 False 频率分量多了，带宽自然就宽了。

Exercise 3.21 2f1,f1+f2,2f2 250Hz

Exercise 3.22 出现好多频率分量

Exercise 3.23 ( c ) -500<f1n-(f1-f2)m<500
( d ) 不知道，猜测和频率两两只差有关系

Exercise 3.25 （c)乘以三之后3的比例显著提升。

3.6 The Fifth Element: Mixers

One feature of most telecommunications system is the ability to change the frequency of the signal without changing its information content.RF（300KHz~300GHz）

3.7 The Sixth Element ：Adaptation

Adaptation is a primitive form of learning. Adaptive elements in telecommunication systems find approximate values for unknown parameters in an attempt to compensate for changing conditions or to improve performance.A common strategy in parameter-estimation problems is to guess a values,assess how good the guess is,and refine the guess over time.
adaptive elements occur in a number of places in the communication system,including the following.
In an automatic gain control.
In a phase-locked loop.
In a timing recovery setting.
In an equalizer.

3.8 Summary

The bewildering array of blocks and acronyms in a typical communication system diagram really consists of just a handful of simple elements: oscillators,linear filters,samplers,static nonlinearities,mixers,and adaptive elements.In addition,this chapter has emphasized the “how to” aspects by providing a series of MATLAB exercises,which will be useful when creating simulations of the various parts of a receiver.

Step 3: The Idealized System

4 Modeling Corruption

The path between there and here can be degraded in several ways, including multipath interference, changing(fading)channel gains, interference from other users, broadband noise, and narrowband interference.

4.1 When Bad Things Happen to Good Signals

This section discusses five kinds of corruption that are used throughout the chapter to motivate and explain the various purposes that linear filters may serve in the receiver.

4.1.1 Other Users

4.1.2 Broadband Noise

4.1.3 Narrowband Noise

notch

4.1.4 Multipath Interference

4.1.5 Fading

Fading may be caused because the transmission path changes.
Such time-varying problems cannot be fixed by a single fixed filter;rather,the filter must somehow compensate differently at different times.

4.2 Linear System: Linear Filters

This chapter describes the three representations of linear systems and shows how they interrelate.

4.3 The Delta “Function”

Exercise 4.4 解决了

4.4 Convolution in Time: It’s Linear System Do

Exercise 4.13 不会

4.5 Convolution？Multiplication

4.6 Improving SNR

Exercise 4.18 LPF

5 Analog (De)modulation

For efficient transmission, the analog version of the message must be shifted in frequency, and this process of changing frequencies is called modulation or upconversion.

5.1 Amplitude Modulation with Large Carrier

Here is a clear engineering trade-off; the value of the wasted signal strength must be balanced against the cost of the receiver.

5.2 Amplitude Modulation with Suppressed Carrier

The frequency and phase of the modulation cosine(located at the transmitter) can never be known exactly at the receiver.

Exercise 5.5-5.6 很直观地观察到了phase offsets 对recovered message的影响。

Exercise 5.8 frequency offsets =0.1014 0.1 时还好，=1，10时波形逐渐离谱。在观察频移的时候持续时间一旦超过某个阈值，恢复出来的波形都会变得不理想。无论频移有多小，迟早都会很离谱。解决方法是什么 ？

5.3 Quadrature Modulation

AM.m的波形前面有一小段的怪怪的东西与原波形不相符。
Exercise 5.14

6 Sampling with Automatic Gain Control

This chapter begins by considering the sampling process both in the time
domain and in the frequency domain. Then Section 6.3 discusses how Matlab
can be used to simulate the sampling process. This is not completely obvious
because analog signals cannot be represented exactly in the computer. Two simple tricks are suggested. The first expresses the analog signal in functional form
and takes samples at the desired times. The second oversamples the analog signal
so that it is represented at a high data rate; the “sampling” can then be done
on the oversampled signal.
When the goal is to determine the value of the signal at some particular point, it is called interpolation.
AGC accounts for changes in strength of the received signal.

6.1 Sampling and Aliasing

This section shows the “how” and “why” of sampling.
the spectrum of the sampled signal ws(t) differs from the spectrum of the
original w(t) in two ways:
Amplitude scaling
Replicas
Bandwidth can also be thought of as limiting the rate at which data can flow over a channel.

6.2 Down conversion via sampling

6.3 Exploring Sampling in MATLAB

evaluate a function at appropriate values (or times),
represent a data waveform by a large number of samples and then reduce the
number of samples.