The WindoWSIL

Given that my first post was a disaster of sorts, I’ll see if I can get things right this time.

Today was the Motorola SABA visit day and some of us presented posters. My poster consisted of one large 40″x30″ sheet of paper; I used a different WNCG template than the standard 9-slide approach.

Several people (including Robert) wanted to know how I got that to work. You can ask Lori for this template (or I can e-mail it to the group if there is sufficient interest). After making your poster (and saving it in an accessible location), go to the HP Studio Classroom2 in ECJ and log into any computer using your ENGR account. There, you want to use the HP DesignJet 755 CM Color Plotter (you may have to search for this printer and add it). After selecting your preferred paper size and the print quality, you can let the plotter go to work.

Note that the paper size that the plotter uses is much larger than the WNCG posterboard. Also, each of us has a print quota of 1000 pages per semester and each page printed on the plotter is worth 10 pages for the quota. In addition, make sure to roll up your poster after the plotter finishes up to avoid creasing it (you can look at my SABA poster to see how not to transport your poster from ECJ to ENS).

Robert brought up a good point about this template; since everything comes out on one big piece of paper, you don’t have to force 9 slides onto this template. Instead, you have the freedom to experiment and be creative…

I recommend trying this approach when you have the chance (the WNCG Board Meeting and/or the Texas Wireless Summit would be good opportunities).

Thanks to Prof. Vishwanath and Wei Wu for this “big poster” idea.

For those that are interested in an academic job, I’ve got a great reference (from an insider’s perspective) on how to secure your entrance into the academic community. As an aside, the site where I found this link (PhDs.org) has lots of information for Ph.D. students.

Yesterday’s post got me thinking…just how many relevant journal resources are available? Today, after googling and looking through UT’s journal references, I’m making an attempt to summarize all relevant journal sources. I encourage all WSIL members to contribute to this list. I have restricted the journals to areas traditionally associated with research in signal processing for communications (i.e. falling under Prof. Heath’s areas of expertise). Thus, I have not included applied mathematics, control theory, computer science, or image processing topics. If Prof. Heath desires, we may also include them.

• IEEE Journals: No doubt we’ve all heard of the Institute for Electrical and Electronics Engineers. In the US, we revere IEEE publications with the highest regard. With Transactions on Signal Processing, Communications, Wireless Communications, Information Theory, Vehicular Technology, Antennas and Propagation, Selected Areas in Communication, Solid State Circuits, and more, IEEE is a great go-to-guy for information on anything communications as well as Electrical Engineering.
• PROs: Very respected and relatively rigorously peer reviewed. If it’s happening at a US university, you’ll find it here. Complete UT electronic access offered here.
• CONs: Turnaround time from submission to publication can take several years. Formatting can be somewhat restrictive (no color).
• EURASIP Journals: EURASIP includes a number of relevant, well respected European journals including the Journal on Applied Signal Processing, Journal on Information Security, and Journal on Wireless Communications and Networking. EURASIP Journals are what is known as “Open Access Journals” meaning that it’s free for all. It is claimed that EURASIP journals have a fast turnaround time, but I have no personal experience to verify.
• PROs: Open Access, fast turnaround time, peer reviewed, and color publishing allowed.
• CONs: Less visibility compared to IEEE especially in the US.
• IET Proceedings: The recently formed (Sp ‘06) Institution of Engineering and Technology is a consolidation of the old IEE and IIE of Europe. IET has a series of new proceedings now that the old IEE proceedings have been absorbed by IEEE Xplore. The new IET proceedings include peer-reviewed Communications and Microwave Antennas and Propagation.
• PROs: If the IET proceedings receive as much attention in Europe as the old IEE proceedings then this publication will be a good source for research in Europe.
• CONs: Newly formed so will likely receive little visibility initially. Currently, we don’t have access through UT (I’m contacting the librarian on this one).
• IEICE Journals: The Institute of Electronics, Information, and Communication Engineers publishes an array of journals in our area including Transactions on Fundamentals of Electronics, Communications and Computer Sciences, Communications, and Information and Systems. These journals are accesible through Oxford’s journal resources online (click here for access through UT).
• PROs: These publications come almost exclusively from Japan and Asia, giving it a refreshing perspective.
• CONs: These peer-reviewed publications don’t receive much visibility and recently plagiarsim has been rampant.
• Wiley-Interscience Journals: As mentioned yesterday Wiley offers research journals with varying levels of prestige. For instance the International Journal of Circuit Theory, the International Journal of Satellite Communication, Microwave and Optical Technology Letters, and Networks all receive decent worldwide visibility. Other journals such as European Transactions on Telecommunications, **International Journal of Communication Systems, the **International Journal of RF and Microwave Computer-aided Engineering, and Wireless Communications and Mobile Computing are either new or less well-known.
• PROs: Nice paper formatting with color, peer-reviewed, and enjoy decent visibility outside of US.
• CONs: Generally poor visibility in the US with less activity than IEEE. **Some journals not available through UT.
• Springer Journals (inlcudes Kluwer): Springerlink offers a large quantity of research journals with contributions primarily out of Europe. Relevant journals include the International Journal of Wireless Information Networks, Telecommunication Systems, Wireless Personal Communications, Lecture Notes in Computer Science, Journal of VLSI Signal Processing, Mobile Networks and Applications, Communications Technology and Electronics and Wireless Networks.
• PROs: Good visibility in Europe. UT access available here.
• CONs: Poor visibility in US.
• El Sevier Journals: El Sevier offers some well-known journals including Computer Networks, Signal Processing, Computer Communications, the AEU International Journal of Electronics and Communications, and Digital Signal Processing.
• PROs/Cons: See Springer above. UT access available here.
• Miscellaneous Journals: I have termed the following journals “miscellaneous” because they don’t belong to any Publisher or established International Institution: African Journal of Information and Communication Technology, Communications in Information and Systems, International Journal of Computers, Communications, and Control, Journal of Computer Science, Journal of Signal Processing.
• PROs: Generally these publications are free for all. Topics not publishable in other journals may find a place here.
• CONs: Low visibility and content is questionably reviewed for uniqueness and worth.

Again, I’d like help filling out this list more with worthwhile journals as well as an analysis of each.

When creating figures (using Illustrator or Powerpoint) or creating slides for your presentation, one often needs to utilize mathematical symbols and equations. Unfortunately, presenters often either use Microsoft Equation Editor, or cut and paste from a *.pdf file. If you are a careful observer, it’s easy to pick these guys out…and generally, looks sloppy. The good news is, there is a better way! Several utilities exist to create LaTeX generated images. Most utilities also give you the ability to make the background transparent or a single color background so that the inserted image blends in nicely. Below I’ve summarized your options for various operating systems:

• Windows: The well-known utility is TexPoint. This is an add-in for Powerpoint that allows you to create LaTeX images. Back when I used TexPoint I was always having problems with it making Powerpoint crash…I’m not sure if this is still an issue. TeX4PPT is supposedly an improvement on TexPoint.
• Mac OS: There are three utilities available. Equation Service is the utility I use. It’s simplicity makes it a great choice. The two other utilities available are Equation Editor and LaTeXiT.
• Linux: Only choice I know of is Equation Editor, but let’s face it, if you’re using Linux exclusively, you’ll probably be doing your slides with LaTeX.

During the last two weeks, at least two students have lost laptop hard drives. Other students have lost data stored locally on the hard drives on their desktop PCs. Data loss can have a real impact. Imagine losing the last month of revisions on your thesis, or that conference paper the night before the submission deadline, or having to retype your camera ready paper from an old printout.

I want to emphasize that you should make sure that you have a reasonable data backup solution. Local data stored on PCs or laptops is not backed up. This seems obvious but is usually ignored until too late. You need a backup solution that you will use every day.

The most natural solution is to back up your data to your ENGR directory. The ENGR server is backed up frequently, at least daily. Note that you should only backup research related files. Backing up mp3s, jpegs, etc may cause a problem with the college.

Another option is to synchronize multiple PCs, for example your PC at UT and your laptop.

A final option is to use an external hard drive. Be careful though. An external drive can get lost. If the drive is on and a virus attacks you may in any case be out of luck!

If you want to synchronize multiple PCs, or use an external drive you might want to invest in some backup or synchronization software. There are many free software packages available online. I will reimburse you if there is a paid package that you have to have.

For reference, here is my backup procedure with several layers of redundancy:

- Use Interarchy to mirror files between G5 and new laptop. Done at least once daily.

- Use SuperDuper! to backup G5 files to ENGR drive (mapped using Samba). Done every few days.

- Use SuperDuper! to create bootable G5 backup on FireWire external drive. This backup includes my whole hard disk. The disk is turned off most of the time to prevent an attack. Done every few days.

I thought I’d get the ball rolling with this first post on figure guidelines for paper submissions.

1. Never include too many figures in a paper. A good rule of thumb is to never have more figures than pages. IEEE Transactions on Communications restricts submissions to less than 10 figures (you are restricted to 20 pages single-column, double-spaced which translates to roughly 10 pages double-column, single-spaced). It is important to not have too few figures as well. A publication heavily dominated by text may be viewed as unappealing to someone flipping through IEEE Transactions publications or conference proceedings.
2. Figures are used to visually convey information that otherwise would be difficult to explain in writing. If a figure doesn’t convey information effectively, either remove it or edit it appropriately. Conversely, if you find yourself spending paragraphs to explain a simple idea, you might try to visually represent this information.
3. Be meticulous when you construct figures. Crooked lines, improper alignment, fonts that don’t match the text, text that isn’t centered inside boxes, etc. are all observations that I’ve made in published work. It looks sloppy and can negatively influence someone’s opinion of your work. It isn’t hard to construct good figures (and it’s kind of fun) so spend the extra time making your figures look slick.
4. Make sure that lines are 1-2 pt and fonts are 12 pt (I also use bold style) in Figures. In matlab this can be achieved by selecting “Figure Properties” under the “Edit” menu of the figure (see below). Just click on the lines or the text to edit the parameters. Make sure you save your matlab figures as *.eps images (for “latex”) or as *.pdf images (for “pdflatex”). One of the nice features of vector images is if you decide to make some figure format chages after the fact, this can be done with programs such as Adobe Illustrator.
   
5. Be creative. Remember, not only are we trying to write technically sound, innovative papers, but we’re also trying to pique the interest of outside observers. I’ve read whole papers before just because I liked their figures. I really like the figures in Kaibin’s new preprint (although the lines could have been a bit thicker). Be careful not to make your images look too “cartoonish”. We are publishing technical documents and we want this work to be taken seriously. I like to ensure that my figures have well defined edges and don’t look “soft” or “bubbly”. I think that’s part of what separates a technical paper from a magazine article.

Those of you who are new to research may wonder at first why I insist that we use LaTeX instead of Microsoft Word for writing papers, proposals, term papers, or anything else. The main reasons are the formatting of equations and the automatic referencing. Both make papers easier to read. Many of you purists believe that papers are evaluated solely on their technical merit. Well, I wish that were true. Technical merit is important but a poorly written or poorly formatted paper makes a bad first impression and bad first impressions are everything. A reviewer naturally wonders if you can’t format equations correctly or cite papers correctly if they can trust your derivations or simulation results.

We will always use the LaTeX package for writing papers. Most conferences and journals provide LaTeX templates for this purpose. See for example the IEEE LaTeX style files. There are some great online LaTeX references. See 1, 2, 3 for example. If you ever have to find how to do something specific in LaTeX, your best bet is to use google. You can buy reference texts but you won’t need them.

There are two main ways to get a pdf file out of LaTeX: use pdflatex or use latex, which creates a dvi file, then convert to pdf. I prefer that you use pdflatex because it ensures that type 1 fonts will be embedded. This is required for most electronic publications. It is possible to configure dvi2pdf to do this but I have always had problems with this approach. When you install LaTeX packages such as MikTeX for Windows, teTeX for Unix/Linux/Mac, etc, pdflatex is included. Incidentally, fink’s teTeX is worthwhile if you’ve got a Mac.

It is important to note that when you use graphics packages with LaTeX (e.g. graphics, graphicx) pdflatex and latex handle these differently. I suggest you use the graphicx package if you are using pdflatex and *.pdf figures since it is a very efficient vector format. Using pdf will allow you to use pdflatex to directly create pdf files for your paper. I recommend the graphicx package for including figures. The following matrix shows which image formats can be used with the different packages within the different latex conversion tools.

It is best to use either encapsulated postscript (*.eps) or *.pdf images since they offer the capability of vectorized drawings. If you use rasterized image formats (such as *.png, *.jpg, *.gif, etc) the image perception is degraded when images are scaled to different sizes. Additionally, with vectorized images, it is possible to go back and change line widths, fonts, and other relevant parameters. In my opinion Adobe Illustrator is the best image creation tool. However, a simpler (and more widely available option) is Microsoft Powerpoint. With Powerpoint, you have to ensure the image takes up the whole slide and you are restricted in the image formats available.

You’ll want to choose a good editor for writing your LaTeX files. I am a Mac user so I prefer TeXShop simply because it’s so easy to use. If you are a WinXP user the two best available editors are WinEdt and TeXnicCenter. Probably WinEdt is easier to use and I will reimburse you for a registration to avoid the anoying popup windows. For Unix/Linux I use emacs but there are alternatives such as VIM for which LaTeX add-ons are available.

A nice way of including references in LaTex is to use BibTeX. Essentially with BibTeX you mantain a library of references. After running LaTeX, you run BibTeX to insert references, then run LaTeX again. When using BibTeX an accompanying style file formats the bib entries according to the type of journal you are using. Very elegant. Reviewers of a paper always look carefully at the references in a paper (in particular because they are looking for their own citations but more on that at a later date). Formatting errors in the references also leave a bad impression. So use BibTeX files and remember to enter your data carefully. There are many software packages that will help you manage your bib files (sort, generate keys, etc). I use BibDesk on the Mac. There are programs available for almost any platform. For example there are several add-ons to WinEdt. Or you might try BibTeXMng for WinXP.

As a final note, there is a very cool web application built by Kjell Magne Fauske that helps you convert directly from IEEE Xplore to BibTeX! It’s beautiful. Incidentally, I suggest that you configure BibConverter to use the options IEEEtran.bst macros and first last author format. On the mac, using BibDesk you can even drag these references directly in to your library.

Thanks to Bob Daniels for additional information, especially on formatting.

Almost two years back I emailed this resource to the group, but since the group has changed (as well as the resources) I thought I’d send it out again. As a student at UT, you have access to many online textbooks, including many wireless and communication/networking technology books. For online texts through UT go to the etext web page on UT’s library web site. There are three notable resources on this site.

1. Digital Books Index - This site has many communications and mathematics texts. It used to be a good resource, but I haven’t been able to successfully download any of the books recently. Maybe there’s a proxy somewhere from UT, but I haven’t been able to find it. Maybe you could help me with this?
2. Netlibrary - Requires a proxy server login from the UT library link. Has many excellent texts (for instance Cover and Thomas used to be on there, don’t know if it is now) in both communications and mathematics.
3. Gale Virtual Reference Library - This one is new to me, but I searched for document titles under wireless and found the Wiley Encyclopedia of Telecommunications, so I’m sure there is more.

There may be more texts from the UT library site, but I haven’t gone through them recently. If anyone wants to go through and update it for me, please do! Outside of UT there are other great resources.

1. book.google.com - Google provides an excellent resource that allows you to search all of the books of many great libraries including the libraries at Stanford, Michigan, and the New York Public Library. There is a copyright protection feature which only allows you to view certain pages, but most of the pages are still there. I find this is an excellent resource before I check a book out in the library or purchase it from Amazon.
2. Project Gutenberg - I don’t really like this site, but there are some interesting old mathematics textbooks.
3. The Universal Library - CMU hosts this website of old, copyright-free (due to age of texts) books of great interest. Do a simple search of “wireless” and you’ll find some intriguing books from the turn of the century. Continous Wave Wireless Telegraphy, anyone? Should be some great, legendary math books in here too.

Good luck with these resources…everything you need is at your fingertips!