It’s been a long time since the last blog post, but nothing ground-breaking has (yet) happened, so here’s just a minor update.
On my second try, I’ve actually survived the first year of PhD! Jokes aside, I’m actually having a lot of fun. After my lengthy prologue that included reading into the computer networks domain, converging onto a concrete PhD topic and surveying related work, I have started with the construction/assembly of my “experiment setup” by implementing my personal reinforcement learning environment for routing using a well-known network simulator. I am currently implementing the first learning algorithm sketches to conduct the first few experiments hoping to get a glimpse of how hard this problem actually is, and what algorithm/model approaches might work. The details on my work will hopefully follow in the first paper, which I expect to submit in late summer. Sometimes I’m still slightly nervous when realizing that by that time, half of the time period indicated on my PhD is already over, but nowadays I am rushing it much less than I used to, especially since my supervisors and fellow academics keep telling me that I’m actually doing great (again, I myself seem to be my own biggest critic). So I’m confident that things will fall into place eventually.
Doing the Industry PhD still, for me, is a huge privilege since I am allowed to freely research on my problem without any obligations whatsoever. Occasionally however, there is some corporate politics nonsense that affects me despite me not being directly involved. So far that has not severely threatened my PhD project (and I won’t go into corporate details here), but I’d like to advise you to check whether there’s some form of representation/spokesperson for PhDs at your industrial institution, because otherwise in a sufficiently large company you might just be crippled by some random decision/policy that saves the company a minor amount of effort/money, but robs you of a crucial opportunity for education, research, development, networking etc.
The past few months have mostly been about implementing my ideas, which actually went quite well. I was able to maintain a consistent coding pace throughout many weeks, although there have been phases with higher as well as lower productivity. The higher productivity phases were those when I had a clear vision of what I wanted to implement (mostly related to my experiment setup) and did not face big distractions. It’s fascinating how much work you can do in a week of back-to-back focus days. The slump phases where mostly triggered by a lack of clarity in certain aspects of my vision, or due to things that arose in my “personal” life. People often say that a PhD is a marathon rather than a sprint, and given how time already passed while working on just the first paper, it definitely feels like one to me. However, It’s alright: I feel like it’s currently kilometer 6 and I’m running it slow and steady to save my energy for the hard moments, instead enjoying the view the course has to offer.