I am always on the lookout for top-quality graduate students, and the ones I have the privilege to supervise are truly awesome. While my research page describes my research program in general terms, and my publications page gives a deeper technical insight, this page is intended to give you an understanding of my philosophy of graduate student supervision, as well as some of the technical mechanics of doing graduate studies in E&CE at Waterloo. In addition, it lists my current graduate students, former graduate students, which may give you some idea of the type of students in my research group, and their life subsequently. If you would like to do research with me, please first read this page, my research page, and some of my relevant publications before contacting me.
|Name||Thesis||Completion Date||Last-Known Location|
|Miao (Michael) Jiang||Modeling Management Metrics for Monitoring Software Systems||September 2011||Google, Waterloo|
|Kamran Jamshaid||Centralized Rate Allocation and Control in 802.11-based Wireless Mesh Networks||January 2010||KAUST, Saudi Arabia|
|Mohammad A. Munawar||Adaptive Monitoring of Complex Software Systems using Management Metrics||September 2009||RIM, Waterloo|
|Name||Thesis||Completion Date||Last-Known Location|
|Nicholas Armstrong||Just-In-Time Push Prefetching: Accelerating the Mobile Web||September 2011||?|
|Yi Luo||Cost-based Automatic Recovery Policy in Data Centers||May 2011||?, Toronto|
|Alice Yeung||Delivering IBM Tivoli Provisioning Manager as a Virtual Appliance||April 2011||IBM, Toronto|
|Robert Robinson||An Architecture for Reliable Encapsulation Endpoints using Commodity Hardware||April 2011||?|
|Giyeong Son||Experimental Performance Evaluation of Bit-Rate Selection Algorithms in Multi-Vehicular Networks||January 2011||VP, Samsung, South Korea|
|Yu (Henry) Xu||Problem Determination in Message-Flow Internet Services Based on Statistical Analysis of Event Logs||May 2009||RIM, Waterloo|
|Adeolu Adeoye||A Framework for the Self-Configuration of Wireless Mesh Networks||May 2009||Alliant Technologies, New York|
|Allen Ajit George||Providing Context in WS-BPEL Processes||September 2008||AeroFS, Palo Alto|
|Edwin Chan||QoS-aware Mobile Web-Service Discovery using Utility Functions||May 2008||IBM Toronto Lab|
|Cecil Reid||Achieving Soft Real-Time Guarantees for Interactive Applications in Wireless Mesh Networks||January 2008||Mona Geo-Informatics Institute|
|Belal Tassi||Fast and Simple Deployment of a Linux Cluster Data Warehouse||July 2007||IBM Toronto Lab|
|Miao (Michael) Jiang||A Cooperative Game-Theory Model for Bandwidth Allocation in Community Mesh Networks||May 2007||Google, Waterloo|
|Lei (Lily) Li||The Efficacy of Source Rate Control in Achieving Fairness in Wireless Mesh Networks||May 2007||Engenuity, Toronto|
|Sultan Al-Ghamdi||n/a||May 2007||?|
|Kevin Quan||Problem-Resolution Dissemination||January 2007||IBM Toronto Lab|
|Ann Lee||Achieving Fairness in 80.11-based Multi-channel Wireless Mesh Networks||August 2006||Speilo, Moncton|
|Evan Jones||Practical Routing in Delay-Tolerant Networks||April 2006|
|Dushyant Bansal||Third-Party TCP Rate Control||September 2005||Microsoft|
|Tao Huang||A Dynamic Cluster-Timestamp Creation Algorithm for Distributed-System Management||September 2005||TD Securities|
|Dwight S. Bedasse||An Efficient Computation of Convex Closure on Abstract Events||January 2005||IBM Jamaica|
|Jia Jun Wu||Collecting Transaction Data in Event-Monitoring Tools||October 2004
|Mohammad A. Munawar||Multi-interface Multi-channel Wireless Mesh Networks||June 2004||RIM, Waterloo
|Vijay Dheap||Event-Driven Response Architecture||June 2004||IBM North Carolina
This section is intended to describe my philosophy of graduate-student supervision and, in some measure, spell out what I believe is the role of the supervisor and the student. I presume at this point that you have looked at my research program and believe that it fits with your research interests. I start with this comment because it is a critical point, and one that might not always be readily apparent to would-be graduate students. Graduate studies are quite unlike undergraduate studies. The primary requirement of a graduate degree, even a Masters degree, is to learn how to do research, in part by doing research. In that regard, it is critical that a would-be graduate student have some idea of what research s/he would like to be involved.
Following on from this, I view my role as a supervisor to be one in which I am training my graduate students in how to do research, which is nothing more than how to solve large, open problems, and to communicate the solution that is found. There are a number of different ways in which this training takes place, and it is largely tailored to the individual student. It is for this reason that graduate studies requires one-on-one interaction with a supervisor.
The role of the student, in turn, is to learn how research is done, in
part by doing research. What this means is that the student would be
expected to contribute to scientific and engineering knowledge, by
solving some open problem. The solution should be communicated, both
in a thesis and, typically, in scientific journals and/or conferences.
The student would be expected to present their research orally, within
our research group, and possibly also in a conference. Again, my job
as supervisor is to prepare you for that, so that you will pass these
I should emphasize that I have not had any student
fail. Every student I have accepted has graduated in a timely manner
with their desired degree.
(The last two sentences of the above paragraph were true for the first decade of my career as a professor. So what happened? What happened is that I discovered something else about graduate studies, which is implicit in what I write above, but not explicit: what I discovered is that graduate studies (indeed, all education) is a process, not a thing; specifically, it is a process that changes the person educated; after people have completed their education they think differently, they view things differently, they act differently. The process of education is, in part, the process of changing how the person being educated thinks; the speed of that process, in turn, will depend on the willingness of the person to allow himself or herself to be changed. If a student is unwilling to change, then that student will ultimately find it very difficult, or (in this case) impossible, to complete the course of studies.)
I emphasize this central role of research because it has been my experience that successful graduate students want to do research, and do not simply want to get a degree. The degree is what the student receives for doing the research. Students who focus on the research find that the degree will take care of itself. Students who focus on the degree, find the research is this very painful thing that is required for the degree. They hate it, but, per the earlier point, they still have to do it.
That said, it is my experience that most people who don't like research, dislike it because they do not understand it, and they do not see the value in it. They really just want a degree, from Waterloo, so they can get a high-paying job. What they do not understand is that the skill set necessary for success in industry is the same as the skill set necessary for success in graduate studies. What do I mean by that? Simply this: your boss in industry really doesn't care if you are good at courses, or can write exams well. Your GRE scores mean nothing to him/her. Rather, s/he wants you to innovate products. Innovate? Wow! Sounds like you must do something novel and significant. You must be able to describe this innovative idea to your boss, to persuade him/her to fund the idea. Wow! Sounds like you must communicate this idea, likely both orally, in a presentation, and in writing, like a publication. See? Same skill set.
Finally, some more general comments about my views of supervision. First, I believe that students need financial support, and as such all of my full-time graduate students are fully funded. Funding for my students comes from a variety of sources, including IBM fellowships, scholarships from NSERC, OGS, and Nortel, and funding from Microsoft, Nortel, Sandvine, Christie Digital, and others. That said, I do not have unlimited resources with which to support every qualified applicant, nor could I supervise them adequately even if I did have the resources. In general, I select students from the following sources, (listed in random order):
Students who have interacted with me while they were doing their undergraduate studies have a far greater likelihood of being accepted by me for graduate studies than any other category.
This section is intended to describe some of the mechanics of becoming a graduate student in E&CE at Waterloo. It does not have all of the mundane details, which are readily available on the department web site. Rather, it is intended to give an informal view of the practical process by which graduate students are accepted at Waterloo.
There are two categories of graduate students in E&CE at Waterloo, full-time and part-time. In addition, E&CE offers three graduate degrees, M.Eng., M.A.Sc., and Ph.D. (not counting the online M.Eng. in Electric Power Engineering). The M.Eng. program is a course-based masters, requiring no research supervisor, and so is not relevant for my purposes. That is not to say it is not a good program: it is eminently suitable for those who wish to upgrade their qualifications or for new immigrants wishing to acquire a Canadian credential, which helps in job searching. It does not, however, require a research component, and so has no research supervisor. The remainder of this section will therefore only pertain to the M.A.Sc. and Ph.D. degrees.
To become a graduate student in E&CE, either M.A.Sc. or Ph.D., you must satisfy the minimum admission requirements of the university, faculty, and department, as well as be accepted by a research supervisor. In general, the most difficult part of this process is finding a supervisor who will accept you. That is to say, the official minimum cut-offs are a poor reflection of what is actually needed to be accepted into the graduate research program.
What then does it take to persuade a faculty member to accept you for graduate studies? While the answer to this is as individual as are the various faculty members, there are certain threads that weave through the process, and some general issues applicable to all faculty. In particular, there are two key hurdles that you must overcome: funding and fit. First, though the less significant of the two, is the issue of funding.
By official policy, every full-time M.A.Sc. and Ph.D. student in E&CE is funded. Current guaranteed funding rates are $16,500 (Canadian) for M.A.Sc. students and $19,600 (Canadian) for Ph.D. students. The relevance of this is that every graduate student must either have a scholarship or their supervisor must agree to fund the student. Funding from the supervisor is in the form of a Graduate Research Scholarship (GRS) or Graduate Research Assistantship (GRA). There are no guaranteed Teaching Assistantships (TAs) in E&CE. Furthermore, such TAs offered are considered payment for work, and are not counted in the funding component of the student. In other words, if you do not have a scholarship, and you cannot find a supervisor willing and able to fully fund your RA, then you will not be accepted as a full-time graduate student in E&CE. There is no exception to this rule. In particular, self-funding is not a currently accepted form by which students may enter the full-time graduate program. I emphasize this point about funding because it is the single largest point of confusion I have found among potential graduate students.
Scholarship funding comes from a variety of sources, including:
At this point I should observe that scholarship decisions are rarely a simple matter of grades. In particular, other things that weigh heavily include publication record, letters of reference, and research plan.
The publication record is largely made feasible by working on research at every opportunity you get. It is also critical to ensure this research is published. Unpublished research may just as well not have been done! Take the opportunity to work with faculty at your undergraduate establishment. I have described some of the ways it is possible in Waterloo. There are likely similar mechanisms at your establishment.
Such early research interaction will also lead to strong letters of reference, as well as an appreciation for research. Such letters are critical in the success of scholarship applications. The appreciation for research is needed if you plan to pursue graduate studies. Don't do graduate studies if you don't enjoy research (that said, if you don't enjoy research, likely you don't understand it, per my point above).
Third, you need a coherent research plan. This is made easier by having that appreciation for research that comes by doing it during your undergraduate career. It is also easier if it can be achieved in conjunction with your proposed supervisor. This is frequently not feasible. Where it is feasible, first take the time to read some of the professor's research papers. Then formulate some ideas, and suggest them to the proposed supervisor, and try to arrange a meeting. This is more crucial if you are planning on pursuing a Ph.D., where you will be working closely with the individual for ~4 years.
If you do these three things you will substantially improve your chance of receiving a graduate scholarship. There is, however, an additional reason for doing these things. If you truly wish to pursue graduate studies, then these things will increase your chance of being noticed by a suitable supervisor. All of the requirements of a graduate scholarship application are also present in a graduate-school application. They are present in both for a simple reason: they are an accurate reflection of performance within graduate studies and in a research career. If you attempt these things and find you do not enjoy them, it is possible that the problem is your situation or your supervisor. It is, however, also possible that you are not well-suited to a career in research. Trying it early and discovering you do not like it is also useful, since too many students waste many years in graduate studies not realizing that this is not a good vocation for them.
If a student has full, or even partial, scholarship funding, then they have removed, or at least alleviated, that second hurdle in becoming a graduate student. This makes acceptance easier, but by no means assured. I now turn to the third, and critical, issue, that of fit. Any supervisor will want to see how you might fit in his/her research program, and at a personal interaction level. The smartest student, with the best scholarship, may simply not be a good fit for a particular professor or a particular university. It may not even be an issue of differing views on research, but simply a matter of personality clash. This is one of the reasons why I try to make something of my personality show through on these web pages, especially when describing my family and my opinions.
Fit is hard to make general comments about, though some things still apply across the board. First, I must make a distinction between master's and doctoral students. A number of would-be supervisors are extremely reluctant to accept students directly into the Ph.D. program without having first had some interaction with the student in question. This tendency is, however, tempered by the fact that every faculty member in E&CE has an active research program, and doctoral students contribute far more, and to a greater degree, than do master's students. This is as much a matter of arithmetic as it is of ability. A master's student is required to take five courses and do a thesis. Since the master's is expected to be completed in about 5 terms, and since the courses will typically consume the first two-and-a-half terms, the student will usually do a term or two of actual research, before writing the whole thing up and graduating. A doctoral candidate, on the other hand, is required to take just three courses, likely completed by the end of the second term, and then spend the remaining three years doing research. Thus, the tension.
This tension between the master's and doctoral student reflects a broader issue of knowledge of the candidate. Any supervisor is more comfortable with a student s/he knows than with one s/he does not know. Similarly, the supervisor will prefer students from universities s/he is familiar with over those that are unknown. This can be a hard hurdle to cross, and emphasizes again why early interaction can be so critical. It also speaks of the responsibility of a graduate students to those who will follow after them. If you come from some university to pursue graduate studies in E&CE at Waterloo, you will, in some measure, reflect the quality of students from that university, and will be a benchmark by which succeeding students are measured. If you behave poorly, you will be hurting your prior establishment.
Turning to the question of preferred degree, there is no clear answer to the question of what is the preferred degree to apply for. The doctoral student is preferred, but comes at a higher cost of commitment on the part of the supervisor. My personal history is that I have taken students in almost all categories. In particular, I have accepted students directly into the Ph.D. program. I have also taken students who wish to do a Ph.D., but had them do a master's first, so that I could determine their abilities. Like most faculty, I will not knowingly take students who I am not convinced are capable of doing a Ph.D. That does not mean I only take students who want to do a Ph.D. A number of my students have completed a master's and that is all they wish to pursue. They are typically open about this before I accept them, and I make no distinction when I accept them. To be able to do a Ph.D. is not the same as wanting to do one, and many, many successful people will never do a Ph.D., because they know what they want in life, and a Ph.D. is not necessary for it.
Finally, a negative comment, but it must be said. I, and all of my colleagues, receive a large number of random e-mails from would-be graduate students. We rarely respond, partly from lack of time, but mostly because most such e-mails are so clearly being spammed to multiple recipients, with the hope that one would accept them. This is a poor way of getting attention. If you read and follow my advice above, you will be far more selective in your e-mails, but also far more likely to elicit a reply.
Now I should make some comments about part-time students. First note that, by official policy, part-time graduate students receive zero funding, and are not eligible for either TAs or space, either lab or office. Any lab space made available to such students is at the discretion of their supervisor.
Part-time students are viewed by some faculty as something between a non-entity and a nuisance. Specifically, such students are typically primarily driven by the desire to achieve a credential, rather than to learn much, if anything, about research. For my part, I am open to both classes of students, and evaluate them on a case-by-case basis. I currently have four part-time students. The remainder of what I write pertains to how I interact with such students, since there is too much variation in this area to say anything more general.
Part-time students are required to satisfy all of the requirements of the full-time program, though their time limits may be increased accordingly. I say "may" because there is an exception to the rule. If a student has, at any point, been a full-time student, they are subject to the full-time student time limits. This is a deliberate policy implemented to discourage students from completing the coursework requirement and then leaving for full-time employment. Part-time students, like full-time students, are required to do a thesis. Only M.Eng. students do not do a thesis.
Part-time students tend to fall into two categories, which are roughly defined by the level of interaction. On the one hand are those who operate with minimal interaction with the supervisor when taking courses. After completing their courses, they then interact closely to define the thesis topic, and perform the research. On the other hand are those who interact closely from the start, joining in the research program actively, typically coming to the university once a week, and doing research work on evenings and weekends. I have had both types of part-time students, and welcome both. My decision to accept, or not accept, part-time students is made more-or-less along the same lines as that of accepting full-time students, with an assessment of cost (not just monetary cost, but also, critically time cost) and fit.
Finally, I note that these comments pertain to genuine part-time students. These are are those students who have a job, and wish to pursue graduate studies. Some students who wish to pursue full-time graduate students, but who cannot acquire a scholarship or RA funding from any supervisor, attempt to enter graduate studies part-time. This is a usually bad idea. The opportunities for funding in such a case are extremely limited. I do not generally take on students in this category.
First, a negative comment: it is a really bad idea to do graduate studies just because you cannot get a job at present. Please don't do that. You won't like it. Do it because you want to do some neat research.
Now you know a little bit about my research, and something of my research philosophy. Still want to work with me? Drop me a line. In the subject, write "I want to do research." Tell me about what it is you want to do, why you think you'd like to work in my research group, something about your research ideas, and where your research ideas would fit within my research program. Oh, and click on that e-mail link above to do it. It's a different e-mail address than my usual one.
By the way, this is something of a simple intelligence test. People who are able to successfully complete a graduate degree in Waterloo E&CE should find it trivial to follow these instructions. People who are spamming around for any supervisor who will give them money find it somewhat harder to follow.
|Paul A.S. Ward Last modified: Wed Nov 16 00:25:24 EST 2011|