BRAD     |     EMILLIE

Friday, May 28, 2010

What is Brad up to nowadays...

Most anyone who is reading this knows that the job I took in Ireland is a post-doc. But exactly what is a post-doc, and what do they do? Generally speaking, a post-doc position is a research position taken on by a recently graduated doctoral student. The goal is to solidify the research skills that the student just acquired during graduate work, and to branch out into new areas and expand their knowledge. After a few years, the idea is to continue on to a full-time research position at a university ('post-doc for hire') or company, or try for a professor position. To be successful, the post-doc must carry out research and publish the findings in peer-reviewed journals and conferences; the length of an individuals publication list is now generally seen as the mark of a successful post-doc, although some employers may be a little more discerning. The day-to-day work can involve meetings with fellow researchers and organizational work with granting institutes (the folk who pay our wage), but generally, most of a post-docs day is spent mulling over problems. Once you've established that a problem is worthy of consideration and will add to your areas knowledge base, you solve and publish it (easier said than done... :) ).

My own post-doc work is with a group that is primarily interested in improving wireless LAN networks, specifically the 802.11 standard, or WiFi. Why should more research be done on a technology that I can get access to for $50 at the local shop? Well, that's part of the reason... it is now a ubiquitous part of many an urbanites life. The other part is that the basic mechanisms behind 802.11 can get a lot of improvement by tweaking parameters that are widely accessible. Since they are so common-place, the impact of improving these networks would be appreciated by millions. So far, the group has made some very cool findings (like that the 6 Mbps setting is actually worse performing than the 11Mbps in almost all cases, which is the opposite than what is logically assumed... generally a slower signal has a lower chance of getting an error). The work that I am personally focusing on involves methods of allocating channels amongst competing access points, with the presumption that they cannot communicate with one another. You can imagine this a very useful thing on Commercial Drive, where every cafe has it's own WLAN, and most are probably all using the same frequency, causing a lot of interference to one another.

No comments:

Post a Comment