BenTheMan said:
And you're not concerned about ceiling, at all? i.e., a candidate with 10-20 years of experience probably isn't going to get much better at their job than they already are (no offense, of course), but has all the skills you want; whereas a new Ph.D. probably doesn't have the precise skill set you want, but may have a much higher ability to acquire that skill set.
Sure, but there's a lot to be said in favor of someone who can "hit the ground running" and be productive immediately, versus someone who will need to go through a learning curve. But there's room for both, and I will certainly prefer a promising Ph.D. with good skills but no industrial experience who does a good job explaining how his skills and background would be a good match for the job, over a nominally experienced candidate who doesn't present a persuasive case for himself during the interview.
I can understand turning candidates away for not being able to program (if the job requires a lot of programming) or knowing the rudiments of DSP (if it's a job where that's important), but do you really expect people to know everything about every wireless system that's out there? If you only want a Wikipedia-type response about wireless systems, isn't that something that you can read and learn in a few hours?
Yes indeed, it is, and I'm not demanding more than that. But I certainly expect at least that much. It's a bit obnoxious to show up for an interview without having even done a modicum of homework about the technology under discussion. Also, if the candidate does have at least that level of understanding, then it's also reasonable to expect that he will have spent some time thinking about how his skill set and knowledge base would be applicable, and ideally he will not only have tailored his resume appropriately but also will show up to the interview with those skills freshly reviewed.
So, for example, if he did a bit of homework and knows that the technology in question is (for example) CDMA, he shouldn't show up at the interview not remembering how the correlation of two signals in the time domain manifests itself in the frequency domain. He should understand at a basic level how CDMA works and why it works, and what are some of the factors that can limit its performance and what are some of the things a designer might do to try to overcome those limitations.
He should also recognize that since CDMA is a type of digital communication system, maybe it's a good idea to brush up on that subject. Maybe be prepared to draw a block diagram of a simple CDMA receiver and be prepared to drill down a bit in detail on how some of the key components work. E.g., OK, you'd better be producing a replica code and you'd better find a way to align it, and keep it aligned, with the incoming signal. How do you do that? Oh, and there's some kind of data modulated on this signal? What kind of modulation scheme is used, and how might you go about demodulating that data? And so on.
It's surprising how many candidates can't even be bothered to do that basic level of homework before showing up. It really does not reflect well upon them.
It's been my experience that the brightest people don't really care to know the details of every system around, but if you give them a manual and the internet, they can learn.
No one is expecting them to know all the details of an unfamiliar system. What we are hoping to see is that they have shown enough interest in the subject to learn some of the basic facts before occupying a collective 8+ hours of busy people's time. They are, after all, going to be expected to learn way more than the basics very quickly if hired. Why show up at an interview completely cold? To many interviewers, that suggests a lazy or arrogant attitude, and attitude is really important in an interview, because the most important question in every interviewer's mind, especially in a small group, is: "do I want to work with this guy?"