联系方式

  • QQ:99515681
  • 邮箱:99515681@qq.com
  • 工作时间:8:00-23:00
  • 微信:codehelp

您当前位置:首页 >> Python程序Python程序

日期:2021-01-05 08:31

CAN201 Introduction to Networking Coursework 2
1
CAN201 Introduction to Networking
Coursework 2
Routing simulation
Contribution to Overall Marks 25%
Submission Deadline Monday 15th Jan. 2021, 23:59
Type Individual coursework
Learning Outcome [A][B][C][D]
How the work should be submitted?
? SOFT COPY ONLY!
? You must submit your work through Learning Mall.
Specification
Routing algorithms are very important to determine good paths from sending hosts to receiving host
through a network of routers. In this coursework, you will simulate the routing process using Python
Socket network programming. The goal of this project is to Bellman-Ford distance vector algorithm
Requirements
? You should design only one app for a node, which is used to simulate the routing behavior of a
router. For each app, it will start with the initial distance information of neighbors and IP addresses
+ port numbers of their neighbors. Each app should run without errors and load its own
information.
? Each app should implement UDP communication to exchange information with their neighbors.
You should create a protocol to exchange distance vector information, as required by BellmanFord.
? In order to simulate the group behaviors of routers, unfixed number of apps will be started. The
starting order of nodes is not fixed.
? Nodes should asynchronously exchange distance vector information as needed until the algorithm
converges. After convergence, each node should output its distance vector table to a json file (one
file per node with a given format).
CAN201 Introduction to Networking Coursework 2
2
Data Format (All files are located in CWD, ALL!!!)
Distance information of each node’s neighbors is stored in a json file named with the node
name + “_distance”. (They are given, your apps should load them)
For Graph 1 in the right side, there are four json files to represent these four nodes:
u_distance.json v_distance.json
{
"v": 2,
"w": 5,
"x": 1
}
{
"u": 2,
"w": 3
}
w_distance.json x_distance.json
{
"v": 3,
"u": 5,
"x": 3
}
{
"u": 1,
"w": 3
}
IP addresses + port numbers of each node’s neighbors and this node is stored in a json file named
with the node name + “_ip.json”. (They are given, your apps should load it)
For Graph 1:
u_ip.json v_ip.json
{
"u": ["127.0.0.1", 10000],
"v": ["127.0.0.1", 10001],
"w": ["127.0.0.1", 10002],
"x": ["127.0.0.1", 10003]
}
{
"v": ["127.0.0.1", 10001],
"u": ["127.0.0.1", 10000],
"w": ["127.0.0.1", 10002]
}
w_ip.json x_ip.json
{
"w": ["127.0.0.1", 10002],
"u": ["127.0.0.1", 10000],
"v": ["127.0.0.1", 10001],
"x": ["127.0.0.1", 10003]
}
{
"x": ["127.0.0.1", 10003],
"u": ["127.0.0.1", 10000],
"w": ["127.0.0.1", 10002]
}
Graph 1: An example
CAN201 Introduction to Networking Coursework 2
3
The output distance vector table of each node should be stored in a json file named with node
name + “_output.json”. (Your apps should output them)
For Graph 1:
u_output.json v_output.json
{
"v": {
"distance": 2,
"next_hop": "v"
},
"w": {
"distance": 4,
"next_hop": "x"
},
"x": {
"distance": 1,
"next_hop": "x"
}
}
{
"u": {
"distance": 2,
"next_hop": "u"
},
"w": {
"distance": 3,
"next_hop": "w"
},
"x": {
"distance": 3,
"next_hop": "u"
}
}
w_output.json x_output.json
{
"v": {
"distance": 3,
"next_hop": "v"
},
"x": {
"distance": 3,
"next_hop": "x"
},
"u": {
"distance": 4,
"next_hop": "x"
}
}
{
"u": {
"distance": 1,
"next_hop": "u"
},
"w": {
"distance": 3,
"next_hop": "w"
},
"v": {
"distance": 3,
"next_hop": "u"
}
}
What should be submitted:
Codes:
? Python 3;
? Your application can be implemented using multiple Python scripts;
? But there is only one application.
A development report:
? A cover page with your full name and student ID;
CAN201 Introduction to Networking Coursework 2
4
? 3~ 8 pages, single column, 1.5x line space, 2.54cm margins, Serif font1
, font size:12pt;
? PDF format, LaTeX is recommended;
? Including:
- Abstract
- Introduction: project requirement (do not copy from this document), background, literature
review (try to find research papers, development reports or testing report of similar apps), and
introduce what you did in this coursework;
- Methodology: proposed protocols (using FSM or mind map), proposed functions and ideas;
- Implementation: steps of implementation, program flow charts, programming skills (OOP,
Parallel…) you used, what difficulties you met and how to solve them;
- Testing and results: testing environment, testing plan and testing results (screenshot, tables or
curves for showing performance)
- Conclusion: what you did and why you did it. Future plan and so on.
- Reference [IEEE format]
Meanwhile, you have to follow the compulsory requirement (no tolerance2
):
? The file structure of your submission:

? Only ZIP file is allowed to submit.
? The starting Python script file should be named as “main.py”;
? Run command: python3 main.py --node 3
Marking Criteria
Report (40%)
Marking Criteria Item Mark
Structure (5%) Structure 5%
Contents (30%)
Abstract 2%
Introduction 4%
Methodology 6%
Implementation 7%
Testing and results 8%
Conclusion and reference 3%
Format and language
(5%)
Report style and format 3%
Language 2%
Marking scheme:
1. Structure (5%)
1 Eg. Times New Roman, Modern No. 20 or Cambria. 2 It means that if you do not follow the compulsory requirement, your work will be marked as zero. 3 Eg. python3 main.py -- node x. Your app should open x_distance.json and x_ip.json and finally output x_output.json.
ZIP Compress
CAN201 Introduction to Networking Coursework 2
5
- Well organized structure: 5%
- Reasonable structure: 3% ~ 4%
- Disordered structure: 0% ~ 2%
2. Contents (30%)
2.1. Abstract (2%)
- Appropriate abstract (2%)
- No abstract (0%)
2.2. Introduction (4%)
- Excellent (4%)
- Lack of necessary parts (1%-3%)
- No introduction (0%)
2.3. Methodology (6%)
- Excellent methodology: sufficient and accurate figures and text description (6%)
- Reasonable methodology: clear figures and text description (3%-5%)
- Incomplete methodology (1%-2%)
- No methodology (0%)
2.4. Implementation (7%)
- Excellent implementation: sufficientsteps of the implementation with proper figures or charts (7%)
- Reasonable methodology: clear steps of the implementation with figures or charts (4%-6%)
- Incomplete implementation (1%-3%)
- No implementation (0%)
2.5. Testing and results (8%)
- Excellent: sufficient testing plan for different cases, sufficient results to show the performance with
proper analysis (7%-8%)
- Acceptable: clear testing plan, clear results to show the performance with analysis (3%-6%)
- Lack of testing plan, results or analysis (1%-2%)
- No testing and results (0%)
2.6. Conclusion and reference (3%)
- Excellent conclusion and reference with the correct format (2%-3%)
- Acceptable conclusion and reference (1%)
- No conclusion or incorrect reference (0%)
3. Format and language (5%)
3.1. Report style and format (3%)
- Beautiful and clear typography: 3%
- Acceptable typography: 2%
- Bad typography: 0% ~ 1%
3.2. Language (2%)
- Accurate and concise language: 2%
- Unclear and confusing language: 0% ~ 1%
? Code (60%)
Code running environment:
? The same virtual machine as CW1;
? You have to test your app using the provided virtual machine without any modification;
CAN201 Introduction to Networking Coursework 2
6
? The app that cannot be executed properly in this virtual machine will be marked as zero.
Code testing steps:
1. Your app (you have only one app with a single python file or multiple python files) will be
run on only one virtual machine VM;
2. Your app will be started in different terminals (run together, but maybe not start together):
python3 main.py --node
python3 main.py --node
python3 main.py --node
python3 main.py --node
python3 main.py --node
……
python3 main.py --node
3. Your apps can run without any error (RUN)
4. Each of your apps should be running until getting the final result and write to the
corresponding json file (eg. u_output.json …) (RES4 = [RES1, RES2…RESn5]).
Marking Criteria Item for testing Mark
Phase 1 (20%) RUN 20%
If you cannot pass Phase 1, your app will not go to the testing of the next phase.
Phase 2 (30%) 30% / n * len(RES) 30%
Coding style (10%) Comments / readability 10%
4 Results include only distances. The next hops will not be checked. 5 No more than 10 in this test.
An example:
Terminal1> python3 main.py --node u
Terminal2> python3 main.py --node v
Terminal3> python3 main.py --node w
Terminal4> python3 main.py --node x

版权所有:留学生程序辅导网 2019 All Rights Reserved 联系方式:QQ:99515681 电子信箱:99515681@qq.com
免责声明:本站部分内容从网络整理而来,只供参考!如有版权问题可联系本站删除。