Management of tasks

One of the problems faced by persons involved in research is that they are usually involved in mental tasks which require multiple sittings. It is usually not possible and sit down and keep on thinking till you find a solution. One need to find a way to manage all the task which one does. The objective should be to finish the task as soon as possible once it is taken up. It is not better to take up a task and leave it halfway, instead do'nt take up a task.

We can manage it using two task lists. First task list is where all the pending tasks are just listed. Let call it P list. We can also maintain another task list S. All the task which are part of list S should be done or attempted to be done one by one in cyclic fashion. This means all the tasks will progress steadily if they are in list S. Ideally, if all the tasks in list S are attempted in a day, we can move some important task from list P to S. If all tasks are not attempted in a day, we simply keep on attempting the tasks in cyclic fashion from list S. This usually will try to keep size of the list S, such that one attempts all the tasks in list S at least once a day. This regular attempt on problem, will help in gaining insight or identifying subproblems, issues which need to be resolved. Sometimes, when a problem has been attempted for multiple times, our subconsious part of the mind start working on the problem and solution is seen as flash of insight.


Some minor modifications can be made to this above task management algorithms. In the queue S, if you do'nt have resources for attempting the task, you can just move ahead to the next task and when in next cycle the turn of the jumped task comes, you make an attempt. If some task is has be given a higher priority, it can be listed multiple times in the S list.

               /-----------------\

               |   ---------\    |

--------\      \-->         |    |

  P     |--------->   S     | ---/---->
--------/          ---------/

                   <-------->
                  Length of S

Philosophy

(This is something which is evolving. You may come periodically to check for updates. I may not be clear a many places. If you are reading, you are requested only to use it as inputs to you own thought process and decision process. Please do'nt believe blindly without making your own judgement.)

Fundamentally, research is nothing but asking questions about whatever you decide to research on, and finding out the answers. It is possible that one can do research about whatever problems one faces on daily basis, problems which have been documented by researchers.

Most common practice followed by research students (pursuing Ph.D) is study thoroughly the research paper published by other researchers. Usually this should be done with no faith in anyone. Once you doubt the authors, you study the papers critically. You may end up in many questions. It is good idea to write all of them down, prioritize them. Pick few questions, start finding the answers for them. One need not study research papers for doing research. In general, the research can be done anyone after observing anything of his/her interest.

Whatever you observe, you can ask the following about it.

- What is it?
- Why is it?
- How is it?
- If you do (what you ever you have as an idea?), what is going to happen?
- If you want something at the end of your pursuit, what should be done as next step?

Fundamentally, the research is done as iterative refinement of understanding.

Problem -> Observe what ever you can related to problem -> You make progress by either asking more questions, or deciding on where you must observe next.

This procedure keeps on happening repetitively, untill you landup with solvale subproblem which need to solved first or you are in a position to form a hypothesis (model) based on your observations. In next step, you usually verify your hypothesis.

You also keep on solving the identified fundamental problems as separate research threads. Once sufficient subproblems are resolved, the solution to main problem will emerge.

What are the pre-requisite for doing research?

- One should not waste time. Observing or analysing once daily life is important to find how to modify the lifestyle, to devote maximum time to your research work. It is not only important to to do research work but also to do it as fast as possible.

- When you are not happy with the way your life is progressing, analysing (observing) and finding what methodical improvements are needed, should be done. Usually it is good idea, to form a list of suggestions (rules) and keep on modifying them after each such instance of analysis.

- When doing a work - either thinking, or mentally revising something for identifying clues for forming hypothesis, identifying subproblems, next observation space, a focus is needed. Diversion due to interuptions are not desirable. Typical diversions happen due to one of our five sensory organs (touch, smell, vision, sound, taste). Another major interrupt happens when we have urge to speak. Usually, when you are speaking you cannot do mental revision and observation session. So avoid speaking when doing your research work.

Some more problems

Optical Switching

1. All-optical AGC techniques for gain control of EDFA in optical fiber loop buffer? (17.06.2006
2. All-optical optical threshold detector leading to switching on-off of an optical source? (19.06.2006)
3. Performance of fiber loop buffer memory with different varities of TWCs to determine the best performing TWC? (19.06.2006)
4. How the high bit rate (20, 40, 80, 160 Gbps) payload packets can be generated for transport in all-optical packet switched network? (23.07.2006)
5. How recirculating loop buffer can be used to store variable length packets? (23.07.2006)
6. All-optical packet switch architecture designs and their performance analysis when WDM is also used in links. So far the link (connected to input or output of switch) are assumed to have single wavelength transport.(23.07.2006)
7. Bistable optical elements can be used as memory elements. How to build the optical random access memory (RAM) using them? What will be criteria for judging their performance? What kind of performance can be achieved with them? How they will compare with fiber loop buffer memory? (24.07.2006)
8. How to estimate the control unit complexity for all-optical packet switch? To make the comparative study of all-optical packet switches on the basis of control unit complexity. (09.05.2007)

Software system management

1. Design issues in development of distributed repository system of software, description file archive, distributed event notification system and software deployment and update agent service. (23.07.2006)

Wireless Sensor Networks

Wireless Adhoc networks

1. How the addresses to be assigned to each node? The mechanism should use distributed algorithm. Each node should able to decide its own address based on some local information. If communication is needed before address is freezed, how it will be done?
2. 
   

Reliability in overlaid multicast networks

One of the issues in overlaid multicast networks is reliability of multicast communications. sudden tunnel failures, node failure, degradation of tunnel may lead to failure or degradation of the communications. In such scenarios, what signalling protocols can be implemented for reconfiguring the overlaid multicast topology for reliability. How the proposed algorithms will perform? We need to decide on the performance criterion also.