|
This is a part of the series of blogs about me going to school:
Part 1: FML, Part 2: MLIA, Part 3: Back on track
Part 4: The best two weeks of my life?
And we're back on the air. Reality hit me like a brick to the face on a Sunday morning. School started and my god the pace is fast, I can't even compare it to anything. Last semester I got the opportunity to take 2 classes at the university and they must've been the easiest classes they had to offer at the time. Back then I could spend... 2-3 hours of day on schoolwork and still be ahead of schedule by like a week or two. Now my standard is 8 hours a day at school and then I read for an hour before I go to bed and my longest day so far was 12 hours at school and 2 more at home. But the thing about this is: I fucking love it. The classes I'm currently taking are challenging and I can see real life use for what I'm learning (something I've always needed to study).
But there has still been something that's been nagging me. I've never been a "worrier". I'm the guy who comes to school and just chilled. Like today. We had to interview a PhD for a class and the two classmates I was going to do the interview with were both really nervous. They knew the questions by heart and just had that "what if he realizes I'm over my head"-look. I just read trough the questions once and relaxed. But in three years (yes I know, it's a long time from now) we have to narrow down on a field we want to study in. We have "information and communication technology", "signals and systems" and "build in systems". It's basically if we want to go deep in the programming area, network area or hardware area. And in the back of my head I've had this little voice going "but what if you chose the wrong one, then you're stuck in a field you don't really like"
Turns out that interview might be the best thing that's happend to me in quite a while. The first question was a huge one "what are you currently working on", we knew it had something to do with WLAN and wireless sensors to do but not much else. So he started explaining and I just had this huge revelation. "This is what I want to do". Everything he said sounded so interesting, beeing interested in networking I also understood most of the things he said (on a more basic level ofc). So today I made up my mind for a descision I dont have to make for another 3 years. But it just feels so right!
TL;DR:
I interviewed a PhD today and he made me realize in what area I want to specialized!
  
|
Woot! Become the next rich :DD
|
Good for you 
|
On September 28 2010 23:57 Froadac wrote:
Woot! Become the next rich :DD
You don't "become" r1ch. r1ch was born with all the knowledge in the world. If you'd say r1ch was a normal human being I'd just be a grain of sand on his left foot as he walks trough humanity with his infinite knowledge. And that's just if I'm lucky!
|
lol, congratulations! I'm still looking for what I want to do... better find it before junior year starts 
|
Yeah - you'll probably still change your mind or end up doing something completely different later on.
|
Just so you know... CCNA cisco certificate wont get u anywhere beyond helpdesk/support and minor stuff. So you will (and smart by doing so) go to a college/university and get a expert/professional degree in cisco.
Here's a wiki link to the types cisco certificate if u want to know more
+ Show Spoiler +http://sv.wikipedia.org/wiki/Cisco_certifiering
|
On September 29 2010 01:07 Postaldude wrote:Just so you know... CCNA cisco certificate wont get u anywhere beyond helpdesk/support and minor stuff. So you will (and smart by doing so) go to a college/university and get a expert/professional degree in cisco. Here's a wiki link to the types cisco certificate if u want to know more + Show Spoiler +http://sv.wikipedia.org/wiki/Cisco_certifiering
We did that in my school (gymnasium) god so boring XD but I'm glad you like it, Julmust!
|
Good for you man. I have yet to figure it out myself ._.
|
On September 29 2010 01:07 Postaldude wrote:Just so you know... CCNA cisco certificate wont get u anywhere beyond helpdesk/support and minor stuff. So you will (and smart by doing so) go to a college/university and get a expert/professional degree in cisco. Here's a wiki link to the types cisco certificate if u want to know more + Show Spoiler +http://sv.wikipedia.org/wiki/Cisco_certifiering
yeah I'm in a university now, studying towards a master of science in engineering.
just noticed you were from sweden so its civilingenjör for you aswell 
|
Congrats! I haven't figured this out yet. It's reassuring to know that many people do.
|
Ah, wireless sensor networks! The joke around conferences in the area is that there are more Ph.D dissertations on sensor networks than actual sensor network deployments in the world. 
Sensor networks have a lot of deep implications in pervasive computing, and you can point to many useful applications that would benefit from sensor networks. Everything from industry to safety to transportation to medicine could use more data gathered around, and the easiest way to aggregate data from lots of sensing locations is through a WLAN.
The practical issue with such networks today is that the battery on the sensor nodes run out in a few months, lol. Sensor networks are awesome in that you can sit them down and let them run without needing to babysit them...except until you need to service the damn things. I get the impression that sensor networks is a field that has attracted the most interest in the networking community, yet most of the problems that need to be solved have more to do with lower-level engineering practicalities in communications, electronics design, and power consumption (i.e. outside the scope of traditional networking).
The issue of routing in ad-hoc networks (i.e. the networking back-end for sensor networks) has been investigated heavily already, and many people are still working on that outside of the context of sensor networks.
In any case, I think sensor networks are a decent research area, depending on the angle you take. But be warned that it was the hot stuff 10 or more years ago...and we haven't gotten much further since then.
More general WLAN research is also pretty interesting. Personally, I'm not so keen on QoS and especially security issues, but there are a lot of open problems out there that are interesting to many.
Btw I'm a grad student more in the wireless communications systems area. But I have some research experience with a professor who does some sensor networks and other WLAN problems, so hopefully I know a little bit of what I'm talking about. Of course, I'm probably biased in some way too, so keep that in mind.
|
nice man, moments like these are always exciting
good for you
|
Congratulations! There's nothing more satisfying (debatable) than that moment of clarity of realizing you career goal. You're this close to becoming a man.
|
Lucky you - i'm more in the "well, people from the banking industry are interviewing economics majors, so i guess i'll do something in the banking industry".
I'm sure its nice to feel like you WANT to do something 
|
On September 29 2010 09:19 Myrmidon wrote:Ah, wireless sensor networks! The joke around conferences in the area is that there are more Ph.D dissertations on sensor networks than actual sensor network deployments in the world. Sensor networks have a lot of deep implications in pervasive computing, and you can point to many useful applications that would benefit from sensor networks. Everything from industry to safety to transportation to medicine could use more data gathered around, and the easiest way to aggregate data from lots of sensing locations is through a WLAN. The practical issue with such networks today is that the battery on the sensor nodes run out in a few months, lol. Sensor networks are awesome in that you can sit them down and let them run without needing to babysit them...except until you need to service the damn things. I get the impression that sensor networks is a field that has attracted the most interest in the networking community, yet most of the problems that need to be solved have more to do with lower-level engineering practicalities in communications, electronics design, and power consumption (i.e. outside the scope of traditional networking). The issue of routing in ad-hoc networks (i.e. the networking back-end for sensor networks) has been investigated heavily already, and many people are still working on that outside of the context of sensor networks. In any case, I think sensor networks are a decent research area, depending on the angle you take. But be warned that it was the hot stuff 10 or more years ago...and we haven't gotten much further since then. More general WLAN research is also pretty interesting. Personally, I'm not so keen on QoS and especially security issues, but there are a lot of open problems out there that are interesting to many. Btw I'm a grad student more in the wireless communications systems area. But I have some research experience with a professor who does some sensor networks and other WLAN problems, so hopefully I know a little bit of what I'm talking about. Of course, I'm probably biased in some way too, so keep that in mind.
From what I understood during the interview, progress done by research often takes 10-15 years to be implemented in the private sector. So this would mean we'll see a great development in this field within the next 5 years.
And part of the thing I found interesting about sensor networks is that you need to keep it low energy. You can't write huge pieces of code but you still need to keep it efficient. The university I go to does alot of research in this field, but from what I understood of the professor they still work on problems that doesnt directly concern energy-consumption.
And I'm not sure I'd want to get a PhD in sensor networks, I'd love to, when I get out of school in 5 years, work with actually implementing the research being done into the private sector.
|
Sometimes research takes 10-15 years to be adopted in the private sector, but sometimes it doesn't. There have been commercial products for several years now, but few people use them because of the problems mentioned before.
I think you know this, but very few people do research that they apply in industry after they finish school. A lot of dissertation topics involve assumptions that don't hold in the real world or can't be implemented efficiently, even if they were relevant to begin with. Of course, that doesn't mean there's no value in the work at all--I'm not trying to put it down that much.
I'm not sure which problems are considered not to directly concern power, since so many of them do. Granted, I'm not exactly an authority on sensor networks. I guess node detection and localization problems, especially on startup after placement of nodes, would not really be considered power problems.
But you have to think about routing efficiency in terms of power rather than just some QoS measure or min hops. Min hops might actually require extra power to carry the same message across the network. Of course, MAC and PHY considerations are heavily tied to power as well. Management of sleep states and turning off the radio and amplification chains are crucial. That's not even mentioning the electronics design aspect.
Maybe I'm trolling the networking community, but my very brief (and likely inaccurate) impression is that a lot of people tend to be formulating and solving the wrong problems with regards to sensor networks and a few other issues. Some of this is because the more critical issues are out of their area of interest or area of expertise.
Maybe they would say the same thing to me, who knows.
But I digress. I think what I want to say is that it's good that the professors are excited about their research and see the potential. In fact, a lot of people see the potential, which is why it has been a big area and a major focus at your institution. However, the people you should most listen to in deciding a field are the grad students working on the projects. Also consider talking with other professors in the networking area, so you have a clearer idea of the particular issues that are most interesting to you about sensor networks as opposed to other problems. If you find something else more interesting, then that's just as well.
|
|
|
|
|
|
EPT
