How is Data Received? An Overview of Digital Communications
Вставка
- Опубліковано 12 вер 2024
- Explains how Digital Communication Receivers work to turn the received waveform back into data (ones and zeros). Discusses the basic fundamental situation for wireless and mobile communications, and for Ethernet, digital subscriber lines (DSL), and optical fibres.
This is Part 2 of an overall system overview. The Transmitter was covered in Part 1 "How is Data Sent? An Overview of Digital Communications" • How is Data Sent? An O...
* Note that I made a minor error in showing the Detector before the Equaliser. What I had intended to do was to show them as a single block, since they are most often implemented as a single function (but I was trying to think of too many things at once). If they are implemented separately, then the equaliser comes first. I have fixed it on the Summary Sheet.
Other related videos: (see iaincollings.com)
• What is a Matched Filter? • What is a Matched Filter?
• What is a Constellation Diagram? • What is a Constellatio...
• What is Intersymbol Interference ISI? • What is Intersymbol In...
• How are Different Equalization Methods Related? (DFE, ZF, MMSE, Viterbi, OFDM) • How are Different Equa...
• How are Data Rate and Bandwidth Related? • How are Data Rate and ...
• How are Throughput, Bandwidth, and Data Rate Related? • How are Throughput, Ba...
• How are Bit Error Rate (BER) and Symbol Error Rate (SER) Related? • How are Bit Error Rate...
• What are Channel Capacity and Code Rate? • What are Channel Capac...
• What is Entropy? and its relation to Compression • What is Entropy? and i...
• What is Pulse Shaping and the Square Root Raised Cosine? • Pulse Shaping and Squa...
Full Categorised list of videos and PDF Summary Sheets: iaincollings.com
.
Thank you professor, I have finished my Master degree in Electrical engineering but I still occasionally watch your video to refresh my mind. I really like your explanation style since there is a lot of sublets, pitfall, and misunderstanding when I study wireless communication. Again thank you very much
That's great to hear. I'm glad you're still finding the videos interesting.
Thank you professor, Im, by degree, a theoretical computer scientist but after taking a stochastic process class and watching your videos, I took many more signal and systems/processing classes and performed very well.
Im often finding myself applying that knowledge to computer science problems, it is especially useful in having a stronger understanding of communications, data science and machine learning which are both ultimately based on signal theory.
It gave me a way to look at these topics in a way other students fail to understand, and it even helped me get into quantum computing groups and projects and perhaps PhD in the future.
That's so great to hear. It's a great motivation for me to keep making these videos. These fundamental topics have many applications, and it's good to hear about what you've been doing.
This video is very nice👍 I have been self studying wireless communication and is quite lost in the detail description in the course material. This video helps put all the things together!
Glad it was helpful! Have you seem my website? It has a fill categorised listing of all my videos: iaincollings.com
@@iain_explains yea sure. In fact I have watched some of your other vids before too. Thank you for making these videos for sharing your knowledge. Keep it up!
never believed that receivers can be explained like this in 9 minutes thanks sir for video
I'm so glad you liked it
Hi Iain, long time no see, you wouldn't remember me, but I was either in your class of 1999 or 2000 in University of Sydney. Good to see you still doing your thing (now on UA-cam). I am just doing some revisions to try to do a tech talk on communication systems. I never got into telecommunications, but software engineering is a lot of fun, which is what I do now. Hope all is well :)
It's so great to hear from you. It's always fun to hear from past students and find out a bit about what they're up to these days. Sounds like you're doing software related to communication systems, which would you interesting, I'm sure. I'm glad you found my UA-cam channel. Hope it's helpful for what you're doing.
Thank you for your outstanding video series. If possible, could you make tutorials about UWB signal processing? This technology is becoming more interests today.
Thanks for the suggestion. I'll put it on my "to do" list.
@@iain_explains thanks a lot. Looking forward!
Insightful Video. I am currently an ongoing masters student in Telecommunications Engineering, and have actually been viewing your videos to help me understand more this Digital communication Class.
Now, I was thinking of actually putting these ideas to actual product.
What Chips would I need? For example using your receiver signal flow model, I would probably need RF amplifier chips, band pass filter chips, Local oscillators, modulator(Maybe FSK), DSP, DACs in that order... would this be all or is there something else?
Then for the earlier Transmitter signal flow model, which components/chips would I need for implementation?
Thank you
I'm not a hardware expert, but there are many approaches you could take, depending on the level of integration you're trying to achieve, and the level of access you need to the various signals along the transmission chain.
You are Legend
Thanks. I'm glad you like the videos.
hey lain , at time of receiving the data at the rerceiver the syncronization bits which is used for timming recovery is used only once or whether its used after each frame [does start stop bit is same as the synchronization bits ]
Thank you. So a wifi adaptor for example has all of these elements?
Yes.
Outstanding
Thanks. I'm glad you liked the video.
The first section, the AMP, need some form of regulation i.e. AGC or AVC...
Hi, thanks for the video. From the video I understand that the frequency lock and clock lock are implemented in hardware, what are the benefits of that over implementing in software (after we sample)?
Yes, it's possible to do it in software, but you need to sample at a much higher frequency so that you retain information about the carrier phase rotations and frequency offsets.
@@iain_explains 😊 thanks
very nicely explained video professor!
but I want to ask a small question why there is a need for frequency locking at the receiver? We must be knowing already the frequency which we have to tune in so why there is a need for frequency estimation locking of the frequency?
Just because you know what the frequency should be, it doesn't mean your crystal oscillator will generate that _exact_ frequency. Every crystal is different, and needs the oscillations to be tuned, so that there is an exact match between the oscillator frequency in the transmitter, and the oscillator frequency in the receiver.
Lain, don't we do down conversation in the receiver?
Yes, I probably should have mentioned a bit more about that. Digital downconversion can be done as part of the Matched Filter (conceptually, at least - this would be the optimal thing to do), or it can be done as a seperate step just before the MF (which is most often what is done in practice, using a mixer).
Thanks, professor,
(a) Do we use the ASK process in wired channels?
(b) I watched your matched filter video too. Does matched filter function = S(T-t). That means the Machedfilter function S(T-t) is the function of input S(t). mean matched filter only works for known signals?
(c) Does matched filter should always comes next to the amplifier right? In the case of using ADC, ADC comes after matched filter.
Please let me know. Thanks again
Answers: (a) ASK is essentially "PAM with a carrier". And some wired channels do use PAM. (b) By definition, the _matched_ filter is _matched_ to the transmitted waveform (as explained in my Matched Filter video). But this doesn't mean you can't use the same waveform in multiple ways to send different digital data. For example you could sent +s(t) to represent a digital 1 and -s(t) to represent a digital 0, and then you would only need a single matched filter and look to see if its output was positive or negative to work out if a 1 or a 0 had been sent.
Hi Lain, thanks for the vidéo...A basic question. For ASK, you did not show sinusoids. But I guess we don't have rectangular shape nécessarily, no?
In fact, I am a bit confused to
make the links among line coding, DAC, modulation, symbol mapper and up converter.
I know that if we have an analog signal, we convert the analog signal into digital signal. While doing that, we have quantized samples. Then we do pulse shaping, then up convert? Where is the position of different terms between which I am missing the links here?
If you haven't watched this video, it should help: "How is Data Sent? An Overview of Digital Communications" ua-cam.com/video/MAddbFfCsIo/v-deo.html
sir amazing video but it would be better if you make the next video on the process of symbol time recovery and why its important
over all thanks again sir
Yes, thanks for the suggestion. It's an important topic, but it's not one that I spend much time thinking about. I'll give it some thought.
hello sir i have just steped my foot into the communication field thanks for the video
at 3:25 what does the collection of energy means sir if the energy is collected by match filter then what is the role of intigrator and my next question is why we need to collect energy cant we sample the upcoming pulse at the instant at its maximum amplitude , why there is collection of energy comming into the picture . i am badly confused sir please help me
Hopefully this video will help: "What is a Matched Filter?" ua-cam.com/video/Ci-EjiMJo3I/v-deo.html
hey lain,
I guess before we start matching our incoming pulse and start collecting its energy we need to find out the starting point of the symbol received at the receiver so my point is which algorithm is used at the receiver for the detection of the starting point of the received symbol at the receiver end and what are the impacts of STO(symbol time offset) on to our received symbols and if there is some
Yes, this is what I called 'clock acquisition'. This video might give more insights: "What is an Eye Diagram?" ua-cam.com/video/ROhgIWBteQQ/v-deo.html
what is clock acquisition and why its necessary
cant be communication possible without that?
The clock waveform is needed so that the detector has a time reference, and knows when to sample. ie. when one data symbol ends and the next one starts.
hello sir, hope you fine
sir i have a question the scaling that we are doing in the constellation is scaling in both phase and gain of the received signal means is phase also has something to do with scaling because the constellation has also been rotated apart from the lost in energy due to attenuation
Yes, that's right.
thanks for video professor 👍
You are welcome
hey lain , i have watched your previous video about how to transmit data in which there is DAC is used so, at the receiver end is there any role which ADC has to perform ?? i have seen in some text that after BPF a ADC has been used.
The sampler shown in the middle of the page is an ADC (I probably should have mentioned that).
@@iain_explains ok got that now thanks prof
Thank you sir
My pleasure. I'm glad you found it useful.
hello , @5:25 i cant figure out which oscillator are you talking about at the receiver end in the block diagram , there is no such block of oscillator shown in the diagram could you please clarify?
The oscillator is used to generate the Matched Filter. This must be matched to the frequency that was used to up-convert the signal in the transmitter to the carrier frequency. In many cases, the MF is implemented using a correlator, and in that form the oscillator appears explicitly. See: "How are Correlation and Convolution Related in Digital Communications?" ua-cam.com/video/We5q5FJcbcU/v-deo.html
@@iain_explains thanks lain you are a great ambasdor of wireless communication your channel is my best discovery on youtube wish yo millions of subscribers cheers!
Sir can you explain the channel estimation and detection part . Not able to visualise it
channel estimation is having an idea about how the channel is affecting the signal that we have transmitted
suppose you have sent a Qpsk constellation (11)(cosine wave with a 45-degree phase shift) so at the receiver, you will not receive the same thing rather you will receive the constellation with the rotated phase shift and also with the degraded energy as it may have suffered attenuation and other channel effects, thus by estimating channel you are finding out by which proportion your signal has been rotated or attenuated so that you can decide the decision boundary accordingly.
This estimation is done by the pilot bits which are sent along the data bits and accordingly will decide the decision boundary
Excellent answer Lutz!
@@lutzvonwangenheim9682 thanks so much sir. You explained it so well
hello, professor lain @ 6:37 what do you mean by scaling the constellation? I didn't actually get it
The channel attenuates the signal by an amount that depends on the specific channel (eg. wire type and length, or wireless channel over a certain distance with a certain antenna), and the receiver amplifier amplifies the received signal (most often by a fixed amount), and then after that the receiver needs to scale the overall received detected level, so that the overall effect of the channel and the amplifier cancel each other out, so that the overall detected signal can be mapped onto the constellation to decide which actual constellation point (waveform) was sent. See this video for more details about constellation points: "What is a Constellation Diagram?" ua-cam.com/video/kfJeL4LQ43s/v-deo.html
@@iain_explains did you mean by cancellation is that the overall gain of the channel +amplifier = 1 , if this is the case then can't be possible that this scaling can be achieved by the EQUALIZATION ?
The term "equalisation" refers to undoing the effect of ISI from the channel. And yes, when you have ISI in the channel, then you will have an equaliser, and it will generally have an amplitude scaling so that the overall gain is 1. However, if your channel does not have ISI (or if your modulation scheme effectively avoids the detrimental effect of ISI - eg. OFDM), then you won't have an equaliser, but you'll still need to scale so that the overall gain is 1.
@@iain_explains as you have mentioned in the above comment that for the OFDM systems we are not having the ISI effect but still we are doing the FRequency domain equalization , than what is the purpose of equalization if the ISI effect is not there in case of the OFDM systems i mean how we are doing scaling in case of OFDM systems ??
what stage is ADC ?
The "sampler" is the ADC.
@@iain_explains Thank you so much for replying, Iain. Absolutely love, love, love your approach and explanations. :)