Sergio Servetto

Grading | Machine Problems | Reading Materials | Syllabus | Back to Teaching

ECE 445: Digital Communications over Packet-Switched Networks

Announcements:

12/22, 10:50am: Final grades have been assigned, see this file. Also, if you want to see how I computed these grades, here is a C program I wrote for that purpose (view), and here is the input file I gave it (view). You were a great bunch to work with! Time to enjoy a well deserved break now.

Older announcements:

12/11, 4:15pm: This is the schedule for the oral interviews (all in my office, 326 Rhodes). On Friday 12/12:
- 11:00-11:15am: fukami-ahlawat
- 11:15-11:30am: wilson-tsung
And on Saturday 12/13:
- 8:00-8:15am: lin-chuo-yang
- 8:15-8:30am: peh-palaiyanur
- 8:30-8:45am: chen-mor
- 8:45-9:00am: aizcorbe-li
- 9:00-9:15am: ---
- 9:15-9:30am: cheung-wong
- 9:30-9:45am: cao
- 9:45-10:00am: aprea-papaioannou
- 10:00-10:15am: kwon-tarn
- 10:15-10:30am: ---
- 10:30-10:45am: chung
- 10:45-11:00am: zhou
- 11:00-11:15am: cohen-lee
- 11:15-11:30am: gunawi
- 11:30-11:45am: ---
- 11:45-12:00pm: gordon-yum
- 12:00-12:15pm: ian-chawla
- 12:15-12:30pm: limdongjae
- 12:30-12:45pm: nguyen-siu
- 1:00-6:00pm: ---
- 6:00-6:15pm: rajagopalan-wong
- 6:15-6:30pm: schweizer-israel
- 6:30-6:45pm: varshney-oakslee
- 6:45-7:00pm: zhang
- 7:00-7:15pm: ---
- 7:15-7:30pm: wu-henson
- 7:30-7:45pm: kurtis-tsim
- 7:45-8:00pm: pang-rickard
- 8:00-8:15pm: morgan
So this means, I should be heading for Rulloff's at around 8:40 or so.
12/08, 1:45am: I have received directly (through people sending me email about this), or indirectly (through people that expressed their frustration to the TAs), various requests to make MP#5 shorter, on the basis that it is way too much work to be done during time in which you should be preparing for finals. The answer is NO. Reasons:
- If you read through the papers on video compression, you will find they make use of just about every tool we discussed in class. Hence, to study for your final, all you need to do is read these papers and read your notes from class.
- The papers on digital fountains you had to read before. So now all you should do is just review them. And if you didn't read them when I sent you to read them (some 3-4 weeks ago) and you are reading them now for the first time, then... well, let's say I am less than fully sympathetic to any complaints you might have now!
- The section on layering you needed for MP#4 anyway.
- The only part on which you might have a legitimate complaint is about MAC issues. But I made sure that what I assigned you to read would be of a complexity that even your grandmother could understand it, it should take you literally no time to read this. And the reason I put it is because I don't want you to get out of this class without at least knowing what the MAC problem is, and this year we did not have enough time to cover it properly.
Bottom line: I replaced a version of MP#5 which involved writing code and debugging it for one in which all you have to do is read stuff that will help you prepare for your final, and write a report. So I am sorry guys, but you have to do this.
12/08, 1:45am: Someone taking ECE 425 mentioned to me that in the project report for that class you also had to give a description of some wavelet routines similar to those to do Haar filtering in MP#2, and asked me if it would be ok to use that material as part of MP#5. This is not at all a problem. As long as you clearly identify what is the material that you are using for 425, go ahead and save yourselves some time this way. If you want to turn in a copy of your 425 report together with this one, that would be fine too, just for me to see in what ways you used this material in two different projects.
12/08, 1:45am: On Saturday evening, after I am done seeing the last person for the oral part of the exam, the course will be formally over. I plan to go to Rulloff's, and have a beer. So whoever wants to join me for that, you are welcome to! The first glass of beer will be on me, for whoever is 21 or older.
12/08, 1:45am: The review session will be on Monday and Tuesday, 3:30-5:30pm, in 310 Rhodes (it was just too darn complicated for Kim to reserve a room in Phillips). What we will do is, if too many people show up and we do not have enough space in that room, we will head for Phillips and use whatever room we find empty.
12/05, 11:50pm: MP#5 posted.
12/03, 2:25pm: >>>DEADLINE EXTENSION<<<. It appears that ECE 425 and ECE 467 (two classes I am only happy to see many people are taking together with this one!) have deadlines for submissions this coming Friday as well. Hence, I am moving the deadline for turning in MP4 from Friday midnight to Sunday midnight. Please note: you still need time to prepare for the final, and to work on MP#5, that is why I do not give you more time to complete MP#4. We will discuss MP#5 in the lab section on Friday.
12/03, 1:15pm: You should have received an email from Christina yesterday, with your feedback on MP3. Please let me know ASAP if you did not.
12/03, 1:15pm: The exact date and time for the final exam is set: December 11th at 12pm, in 203 Phillips.
12/01, 10:15am: Some clarifications on MP#4 posted.
11/19, 11:50am: MP#4 posted.
11/12, 3am: MP#4 is not ready yet. Please check tomorrow (Thu 11/13) in the morning again. I will do my best to have it ready before class, so that you have time to look at it by Friday, to discuss it in the lab section. There is a quiz on Friday too.
11/12, 3am: The two papers that you need to read (both by John Byers et al) are available from the Reading Materials page.
11/02, 5pm: >>>DEADLINE EXTENSION<<< New deadline turning in MP#3: Fri 11/07, midnight. Two reasons for this extension: (1) there seem to be problems with ganesh; (2) I was out of town between Thursday and today, and unavailable to clarify some important questions about the MP that people had. We will talk about whatever is pending in class on Tuesday (11/04).
10/29, 10pm: PLEASE MAKE SURE YOU CLEAN UP YOUR QUEUES AFTER EACH EXECUTION OF YOUR PROGRAM. Right now there are too many dangling queues in the system, and people cannot run their code because of that. I just sent an email to the current "offenders". But I am not going to be here over the weekend, while all of you will be cramming to get your assignment done...
10/26, 8:30pm: MP#3 updated.
10/26, 3:15pm:
I have received some inquiries about a piece of code that I gave you, that was working before, and is not working any more. The problem is that this code I gave you before was not cleaning up queues properly upon exiting (even when exiting normally). Therefore, a few of you have run it, and now there are too many dangling queues in the system, and your program fails because it cannot create new ones. So I will ask you to do two things:
- Please take a look at this file, and if you user name is there, please run clqs with the queue numbers associated with your user name in this file. This will clean up those dangling queues.
- Please use the new version of the files that I put a few minutes ago in classes/ece445/mp3/. This should have taken care of the problem of not cleaning up queues upon exiting.
Also, as a general rule, please do check that when your program terminates, no messages queues are left dangling. There is a maximum that can be created at any point in time, and so this creates an annoyance for everyone.
10/19, 11:55pm: MP#3 posted.
10/17, 2:45pm: Means for first two quizzes. Quiz#1: 16.31/30. Quiz#2: 12.25/20.
10/15, 10:30pm: I have received a number of inquiries concerning class tomorrow. There is no class, due to the inauguration event of the new president. We still have a quiz on Friday, and it will be a longer one.
10/04, 4 pm: About a quiz for this coming Friday (10/10). I have had quite a few requests from people who would like to leave early for home (for Fall Break). We will do this, so there will be no quiz on Friday 10/10. However, we will need to make up for this. To not complicate my grading plans, we need at least 6 quizzes in the semester (preferrably 8), and this means we cannot have less than 1 every other week. So far, we have been having exactly one every other week, and therefore, if we are not going to have one this coming Friday (10/10), we will need to add extra quizzes later on.
10/04, 4:20 pm: At the request of one of you guys, I added a version of vplay.c, just like in MP#1. This means, I had to move the definitions of the functions Encode/Decode to libMP2.{ch}. I also corrected a bug in the definition of the isLeaf(tree) macro (pointed out to me by Pranay Ahlawat, so thanks for reading the code carefully!).
9/30, 12 midnight: MP#2 updated. New deadline: Tue 10/07 at midnight.
9/30, 12 midnight: There will be no quiz this coming Friday (10/03). The lab session will be entirely dedicated to going over questions you might have about MP#2. Note: with this last update, you have now all the information you need to complete MP#2.
9/19, 10am (Western european time), 4am (EDT): First version of MP#2 posted.
9/09, 11:50pm: In /classes/ece445/mp1/Debug/ you can find what the output should look like for a window of size 7x7 for MP#1.
9/05, 12:30pm: MP#1 updated.
9/04, 10am: To run gcc on ganesh, you need to add to your path /opt/sfw/bin, else you will not find it.
9/03, 10am: Instructions for how to use the labs, and an addendum to how to turn in homeworks.
8/29: First lab session, and first MP posted.
8/28: First lecture.

About the course:

Instructor:

Sergio Servetto.

TAs:

An-swol Hu.

Christina Peraki.

Ron Dabora.

Meeting times and place:

Lectures: TuTh 1:25-2:40, 206 Hollister.

Lab: Fri 1:25-2:15, 407 Phillips.

Office hours:

During labs, and after class (for 1 hour).

Prerequisites: ECE 301/310 (required), ECE 302 (recommended), and some knowledge of computer programming. You might find that concurrent registration with one or both of ECE 425 and ECE 467, and with CS 414, will be useful. You might also find that taking ECE 468 and CS 519 next spring will be useful as well. I would very much encourage you to do all of this, if your schedule permits.