Dapeng Wang

My new homepage

Wednesday, 17. January 2007, 07:21:56

http://home.ustc.edu.cn/~dpwang

This is a program for my friend, it workds as follows:
Coding Part
1. Read "a.dat"
2. Build the Haffman Tree
3. Store the Haffman Coding, and store the Haffman Code in "code.dat"
4. Translate the source file to "01" string via the "Haffman Code", and store them in "b.dat"
5. Translate the "01" string to ASCII Code, and store them in "ASC.dat"
-----------------------
DeCoding Part
6. Load the HaffmanCode
7. Translate the ASCII Code into "01" string, and store them in "d.dat"
8. Translate the "01" string into the file, and store them in "e.dat"

Thus "e.dat" will be the same as "a.dat"
Final.rar

No comments

How is a MAC protocol implement?

Monday, 4. December 2006, 04:03:44

MAC protocol is implemented partly in hardware, partly in firmware, and partly in the device driver of Network Interface Card. Thus, changing the MAC layer can require hardware and firmware changes, a highly expensive proposition.

No comments

Develop a ping programme for MM

Tuesday, 28. November 2006, 13:22:06

1. The basic idea of this programme:
Model:
Client-- ping -->Server
|
calculate (rtt)
|
Client<--pingback--Server

In this programm, we do the following things:
1) Client sends a packet to the server, with the time vector <timeval payload> in its payload;
2) Server gets the packet from the client;
3) Server picks out the sendtime of the packet from the payload, and calculate the one-way rtt by <current_time - payload_time>;
4) Server sends the rtt back to the client;
5) Client gets the rtt, and print it out.

2. How to compile these files?
Input the following lines:
1) gcc pingd.c -o server
2) gcc ping.c -o client

Then you get two excutive files, server and client.
You could run both of them on the same machine, or on two different machines.

If you want to run them on the same machine:
in one terminal:
3) ./server
open another terminal:
4) ./client localhost (which means you ping yourself). Or you could input
./client 127.0.0.1 Or
./client <the IP of your machine> etc. "./client 192.1.42.23"

If you want to run them on two different machines:
in one machine:
3) ./server
in the other machine:
4) ./client <the IP of the sever Machine>

Howto
ping.c
ping.h
pingd.c

No comments

When Intel decide to lose weight...

Tuesday, 5. September 2006, 00:39:26

From google news, I knew that CEO Paul is going to give a show on "How will intel lose weight?" It is believed that Intel is going to fire more than 10,000 workers in the coming months. This quake was caused by the pressure from AMD, the biggest competitor of Intel. The whole process of this reorgnization started from April 2006 and guessed to be finished at the end of this year.
It is interesting that the price of intel's stock is runing up these days with the news of reorgnization. Experts said that this is because that this "reorg" news encourages investors of Wall Street.

No comments

Investigating Jeoder's UWB Simulation Projects

Friday, 1. September 2006, 02:59:27

AbstractI looked into Jeoder's project on simulation of UWB. I am sure this will be the basic work for our first step on simulating UWB Mesh Network. In the following main section, I list out how it works and the files different from standard ns-allinone-2.29.3.

Keywords UWB NS2 Analysis

Main Section Through beyondcompare, I found the following files are different in Jeoder's codes.
[1] mac/interference-phy.{cc,h} : These two files are modified from phy.{cc,h}
[2] mac-ifcontrol.{cc,h} : These two files are modified from mac.{cc.h}
[3] mobile/modulation-codedpm.{cc,h} : These two files are modified from modulation.{cc,h}
[4] mobile/ppm.cc : This file is called in modulation-codedppm.cc
[5] mobile/propagation-tarokh.{cc,h} : These two files are modified from propagation.{cc,h}
[6] noah/noah.{cc,h} : This module finish a routing method in one hop and multihop scenarios
[7] tcl/mobility/noah.tcl : This file is used for implementing the procs adding a noah agent in the simulation.

No comments

What about this Idea? "Video Messeger over Campus Wireless Mesh Network"

Wednesday, 16. August 2006, 08:35:18

Abstract:
I have a dream, that in my USTC. My group set up a wireless network using cheap WiFi Wireless Mesh Routers, which could cover almost anywhere of this university. Users here could make a short video message by their laptop or phones equiped with camera, then send it to the receiver. Controlled in a Peer-to-Peer way, the Wireless Mesh Work may forward this short video message to the receiver in a non-real-time way. Thus a transmission of video message is accomplished. As well as the video message, news of University could be pushed to all users by broadcast method. Also some little advertisement could be added in each message, for example a little logo or what ever. Due to the low cost and easy layout brought by wireless, we could keep it work without too much money, maybe balanced by the income of advertisement. Further step, maybe the size of the network could be enlarged to a much larger size, even to a city mesh network.

Keywords:
non-real-time, Peer-to-Peer, Wireless Mesh Network, Video Messenger

From the view of Intel's plan "Integrated Computing and Communication in one wireless end device", the idea comes up. The research issues contained in it mainly contains the following four parts:

1, We need a wireless Mesh Network for a campus size communication area. This may be implemented by IEEE 802.11a/b/g, as well as IEEE 802.11n with higher bandwidth. But Actually, we want to implement a non-real-time video Messenger, so the bandwidth will not be a very important issue. The protocol stack for this mesh network will be a good research topic, and this is just what we have done in last year. And it would also be my research topic of My Ph.D.'s Thesis.

2, A Peer-to-Peer solution for the message transmission. My video message does need to be a real-time one, so the bandwidth will not a problem. However, a good P2P solution will help a lot. Though I am not quite familiar with P2P, I know that P2P is becoming hotter and hotter, which will take the market for the near future. What's more, one hand the government and the ISPs could get a penny from it, P2P is said to be killed one day. On the other hand, P2P is really a money-saving method, that everyone loves it. So I believe some days later, it will be charged but the price should be very cheap. In the campus, the ISP will be the University, so the whole project could be totally free service.

3, How to make the messenger a safe one? The security is also a important part. After I have read the security section of IEEE 802.11s proposals, I know that security will a forever topic in wireless communication. How to keep all the user obey the basic rules, protect the Campus Wireless Mesh Network from being attacked, and manage the communication policies are the main points we should consider.

4, A good application both for Windows and Linux. Our group member Shi Lingzhi has implemented a real-time voice over IP before the start of this summer vocation. I do believe we could finish this programming work in a short time.

Conclude
This is just a interesting idea, I do not know which part is more difficult, the research part or the engineering part? What ever. Here from a research point of view, my group "Wireless Network Group" members will be interested in the 1st part, the members of "Grid & P2P" will be interested in the 2nd part, while the members of "Network Security" will be interested in the 3rd part. For those guys who are purchasing their Master's Degrees, they could try the 4th part. Wahu~~! This is so great a project, I really need much more guys help me to finish it. If we could get some founding from Intel or Zhongxing, it will be easier.

Or, maybe it is just a dream...

No comments

Compare Simulation Codes from Venkat's and Standard ns-allinone-2.29.3

Wednesday, 16. August 2006, 01:58:58

I always stands on friends' shoulders to look forward. Venkat is a friendly Indian guy, who has finished a project on 802.11s extentions for NS2. After several letters, he promised to give me the code he finished half a year ago. And now I am able to reduce the time for my recent work. It is a really simple implementation, and he told me that only the RM-AODV routing method was loaded in the mac module. Due to my understanding how "manual.cc" works in Hyncinth, I think it will not difficult for me to understand what Venkat has done.
Using "Beyond Compare Version 2.4.1", I am able to find the modified files. I would like to make a list here to provide a reminder for my future work. The modified files are as follows:

Directory ~/ns2.29/mac:
mac-802_11.h
mac-802_11.cc
phy.h
phy.cc
wired-phy.h
wireless-phy.h
wireless-phy.cc
lanRouter.h
channel.h
channel.cc
mac.h
mac.cc
mac-timers.h
mac-timers.cc
ll.cc

Directory ~/ns2.29/common:
connector.h
bi-connector.h
object.h
packet.h

1 comment

Intel Communication Technology Lab: "Directions of Wireless Research"

Tuesday, 1. August 2006, 06:06:01

From www.soi.wide.ad.jp/class/20060000/slides/03/1.html

Intel Senior Fellow, Director, Communications Technology Lab Corporate Technology Group, Intel Corporation
Date: June 9, 2006
Place: Tokyo University, Japan
------------------------------
------------------------------

No comments

DDC UWB has been installed on ns-allinone-2.29.3 successfully

Friday, 28. July 2006, 08:36:03

Our object is to implement the functions in the proposal "802.11 TGs Simple Efficient Extensible Mesh (SEE-Mesh) Proposal"(IEEE 802.11-05/0562r0).
Want to know more about this proposal? Just find more here:
http://www.csie.nctu.edu.tw/~yctseng/WirelessNet06-02/ieee802-11s-see-1.ppt
The offical draft is not open now. Though I tried my best to google it, nothing was found. I was surprised that they keep this "secret" so "good". What I found is just this Power Point File. It will tell you more about this Intel proposed draft for 802.11s. Fortunately, I have got a copy of it from my manager, but it is only for Intel Internal use, what a pity. We are preparing for the implementation of the two key routing methods in it for UWB Mesh Network.
I found the UWB Patch for NS2 here:
http://www.joergwidmer.org/
And I installed it on ns-allinone-2.29.3 with Fedral 5.0, just a little problem. The little problem is related to "gnuplot", which will be investigated later.

For the next week, I am going to talk on this overmesh project with the other three interns of IT Research and the people from IT FLEX. It seems that they are all interested in this project.
And I get a good news today, Kerry.Jiang, the manager of IT FLEX told me that Kai.Miao has got the budget for this project. However, the scope of this project will be a little lower, because Intel does not feel good since AMD's growing up recently.

No comments

1 2 3 4 Next »

Showing posts 1 - 10 of 33.