AshleyMills.com

Actually, this is an old paper, written before my Intrasite Scheduling paper. But it is a Journal paper so took much longer to publish:

Ashley Mills, David Lister, Marina De Vos: Understanding Static Inter-Cell Interference Coordination Mechanisms in LTE Journal of Communications, Volume 6, Number 4, July 2011 (c) Academy Publisher

If you didn't know, I work for Vodafone R&D. I did my EngD there too. I wrote this paper a couple of months ago toward the end of my program:

Ashley Mills, David Lister: Intrasite Scheduling for Interference Avoidance in LTE In Proceedings of the IEEE 73rd Vehicular Techonology Conference, IEEE VTC2011-Spring, 2011, Budapest, Hungary. (c) 2011 IEEE

It is about leveraging intrasite knowledge to improve scheduling performance in LTE... as the title says.

Finally got round to uploading my Vegas paper:

Ashley Mills, David Lister, Marina De Vos, and Yusheng Ji: The impact of MS velocity on the performance of frequency selective scheduling in IEEE 802.16e Mobile WiMAX In Proceedings of IEEE Consumer Communications and Networking Conference, 2010, Las Vegas, Nevada. (c) 2010 IEEE

I just finished an internal report on something a lot more interesting. I'll be submitting a paper version of this report to a journal in the next few months I hope.

I'm flying to Las Vegas tomorrow. Well, that is if Gatwick Airport is open. It was closed yesterday morning because of snow.

Yes, we have had some snow:

This was yesterday. And of course, it being England, everything ground to a halt. I know, I know, what an awfully clichéd statement. But there is some truth in it, mainly because we don't have the experience to deal with heavy snow fall here, and it doesn't happen often enough to warrant a heavy investment (god that also sounds awfully sheep like of me). I have some suspicions that the local roads have not been gritted enough so I have sent a freedom of information request to West Berkshire council to ask for the gritting budgets between 2007 and 2009 and for the miles of road gritted, just to see if they cut the budget this year.

The snow, the bane that it is, is nethertheless cool (pun unintended). I do need to get to Las Vegas however. I am set to be presenting at CCNC 2010, on "The impact of MS velocity on the performance of frequency selective scheduling in IEEE 802.16e Mobile WiMAX". Remember that paper submission I was bitching about a few months back? Well I re-wrote the paper and submitted it elsewhere, it got accepted at CCNC 2010. It may sound stupid, but I never thought I'd actually have to attend. I just thought they'd publish it anyway, but no, it turns out if I don't go and present it, then they can refuse to publish it. How naïve of me to think that, I mean, what did I think the conference was for?

Well anyway, so I'm off to Las Vegas. Incidentally it also happens to be when CES 2010, the biggest electronics show on the planet, is being held, and registration was free. I don't have to present until Sunday and so I was planning to goto CES on Saturday. Unfortunately I somehow lost the pass they sent me. I have searched high and low, mainly low, and cannot find it. I've sent them an email explaining the situation so hopefully I might still be able to get in. I doubt I'll have another practical opportunity to go. That was silly Ashley wasn't it.

Continuation viva results

As you may be aware, having perhaps read my previous blog post regarding the continuation viva. I think that previously that I had called it a transfer viva, but as I also mentioned, EngD doesn't have an actual transfer viva since the experient of an EngD is registered on that program from the start, whereas a PhD registers first on MPhil and then has a transfer viva at one year to transfer to the DPhil.

The viva was surprisingly less painful than I imagined.

First, I had to present to the department, and I delivered a presentation. Some people started asking questions and I got super over defensive and basically accused someone of not knowing what they were talking about. It was pretty unprofessional in retrospect. I did apologise to this person afterwards however.

Anyway, after the presentation I had the oral exam. I was asked quite frankly whether I thought I could finish the EngD and produce something useful out of it. Good question.

I said yes, and I think this is true. I produced a small time plan indicating what I wanted to do in the next two years. Basically for the next year and a bit I will be working on getting the results, and thereafter I will be writing up the thesis.

It was quite sobering to reflect on what I had done and to really think about what I have to do over the next 2 years to get this EngD. I hope I can do it.

Well what happened to my blog post a day vow. I guess it broke. At least however I am now certainly updating the blog more frequently, so that's a good thing.

Today I have my transfer viva. Its only a transfer viva, perhaps people should call it something else. The only award for passing this viva, is that I get to continue the EngD. Its not like the final viva.

In anycase I finished the document and sent that off on Tuesday to give the examiners plenty of time to read it. I just sent them the timeplan for the next 20 months (that's how long I have left...).

I have finished the presentation on my first paper, and will be delivering this at 14:15. Me and Dave (my supervisor here), will set off at about 12:30 to goto Bath University, and arrive around 13:30. From there we will go to pickup Marina (my academic supervisor) and then goto the PG Seminar. After the PG Seminar, Dave, Marina, Russel (the other academic supervisor), will goto some room and they will examine me on the content of the work I have done so far for the EngD. <gulp>

In preparing for this, I really started to think "shit, what have I been doing". In hindsight its always easier to think what one could have done differently (is it always easier? that's what people say, what if you had a brain damage in the interim? surely then in hindsight it would be much more difficult than it was at the time).

Anyway, in this case I can look back and think, hmm, why didn't I do this or that, or go straight to work on X. But at the time it was a real struggle just to work out wtf I was doing, to learn about the cellular networks field, and to get some grasp of where I could actually do any research. I hope my output will be more prolific in the next 20 months, since now I reasonably know what I am doing and can focus on just trying to write some papers.

I'll let you know how the viva goes.

"You didn't post anything yesterday!" I hear you cry. Whilst it is true that I didn't post a blog entry, I did post a geometry entry so perhaps you will forgive me by virtue of this contribution. In fact I had intended to post a blog entry but, believe it or not, my 3G card stopped working.

I do not have Internet access in my new house/ rented room. The live-in landlady did not have Internet access before I arrived. Being a geek, I had previously considered it impossible that anyone could live without Internet access, but here is evidence to the contrary. Interesting. In anycase I am trying to get it setup but the cheapest way is if landlady extends her sky package to include it, and then I will increase my rent to cover the difference.

Thus the ball is out of my hands, no that's not right, the ball is in her court and so far a week has passed by with no Internet forthcoming. Up until yesterday I was able to survive by using the 3G card in my Vodafone laptop but yesterday evening it spontaneously disabled itself and now will not re-enable.

I will physically remove the SIM and put it back in later and see if that helps but I am preparing for the worst and consequently am submitting this blog entry now whilst still at the HQ. (Incidentally, it is past 17:15 and my official clock off time if your reading this Vodafone, so go take a running one if you were going to start accusing me of shirking).

I've been working all day on my 2nd year "transfer viva" document for the EngD. I call it a transfer viva because this is the term appropriated to the 1st year grilling rewarded to PhD students that survive the first 12 months. Should they fail at this point to demonstrate sufficient progress or otherwise convince the examiners of their efficacy, then they will be kicked off the programme.

The EngD viva is at 2 years because 1 year of this 2 years is taught courses, and hence only 1 year is availble for research. Thus the two streams are comparable in this respect. I have made sufficient material to pass but I am currently assembling it into some readable form that is appropriately convincing. I will have my viva on Friday July 3rd and it will consist of a departmental presentation followed by an oral grilling from 3 examiners.

The only problem I have is organizing my work into some kind of coherent form around a central theme. An EngD is different from a PhD in that, in recognition of the changeable nature of industry, the final report can consist of a portfolio of smaller projects that need not be as closely linked as the work in a PhD thesis. However, the document should contain a section which ties the work together under some umbrella abstract notion.

My thesis is about self-optimising cellular networks (SON). That is, cellular networks whose elements adjust themselves automatically to network change in order to optimize their operating efficiency and to compensate for problems. As a more familiar example, if you run a wifi network at home you might change the operating channel in the router settings so as to mitigate interference coming from a neighbor's wifi router. We want to do similar things automatically in cellular networks, but much more sophisticated.

I have the following items for my viva document:

1. A whole bunch of tutorial information regarding cellular networks to serve as background for readers.
2. A review of the industries current efforts to do SON.
3. A paper about resource optimisation, which I've submitted to a conference.
4. A whole load of other stuff I read related to SON, which I am struggling to fit into some kind of literature review.

The current status of the report is that it basically contains the first 3 of these items in various stages of completion, but that I am confident I can have done to a reasonable standard in time.

I am unhappy with the literature review section, the forth item in the list above. I want to integrate the papers I read about SON ages ago, but I am afraid that they are no longer really relevent or worth including. In SON there are a number of industry standardized "use cases", for example handover optimization so really I think the review should focus on those use cases which the papers are based on, not all possible SON activities because this is too broad.

After the viva I will be working on an actual SON use case and so then this will fit nicely into the thesis. My last paper however, whilst it is a SON type thing, doesn't neatly fit into a use case and was motivated by a problem I identified when reading about WiMAX and so its hard to see how to make it relevent to a particular use case.

WTF am I going on about? Good question el-nino.

I'm sure this is all riveting reading and so I'll be sure to keep you updated on every piddling development. But for now, its time to go home and probably continue writing this fucking thing (the document, that is, not the blog).

P. Tino, A. Mills: Learning Beyond Finite Memory in Recurrent Networks Of Spiking Neurons. Neural Computation, in print, 2005. (c) MIT Press

P. Tino, A. Mills: Learning Beyond Finite Memory in Recurrent Networks Of Spiking Neurons. In Advances in Natural Computation - ICNC 2005, (eds) L. Wang, K. Chen, Y.S. Ong. pp. 666-675, Lecture Notes in Computer Science, Springer-Verlag, 2005.