CS F303: Computer Networks - Take Home Assignment

Write client-server programs to upload one file from each client, c1 and c2, to a server, s1. Client c1 sends a file name.txt, and client c2 sends a file id.txt simultaneously to the server using stop-and-wait ARQ protocol over unreliable channels, which can drop the packets (Data and Acknowledgments) but does not corrupt the packets. The clients c1 and c2 make two separate TCP connections to the server. The server should receive the packets alternatively from the clients' c1 and c2 and write them into a file list.txt. (Protocol is the same as Q.2 of the mid-semester exam.)

Implementation details:

1. Packetization:

The name.txt comprises student names separated by a comma. Client c1 sends one student name at a time in a packet using stop-and-wait protocol by encapsulating it in the form of a packet structure. Similarly, id.txt comprises student ids corresponding to the student's names in the name.txt, separated by a comma. Client c2 sends one id at a time in a packet using stop-and-wait protocol by encapsulating it as a packet structure. A full-stop (.) character terminates the files name.txt and id.txt. Remember, c1 and c2 are independent and send packets to s1 without any coordination between them. `

Note: Create name.txt and id.txt files and add a few names and ids in the specified format to test your programs. However, we will use our files for evaluation purposes with a similar format.

2. Packet Structure (Same for c1 and c2):

Besides the payload, the packet structure should contain the following information in the form of header fields.

• The size (number of bytes) of the payload
• The seq_no (byte sequence number) specifies the offset of the first byte of a packet to the file. The character comma (,) is used as a delimiter between student names and ids, hence do not count it for offset calculation.
• Use one Byte field TYPE to differentiate between DATA or ACK packets. (Note: The seq_no field in the ACK packet would correspond to the DATA packet with the same seq_no value received from the client.

Note: You can use any other header field(s) required to implement the desired functionality. Provide a brief justification for that within the code file as a comment.

3. Random Packet Drop

To simulate the erroneous channel (described above), the server randomly drops a received packet with probability PDR and does not send ACK back to the client. Keep the value of PDR at 10%. While evaluating, we may test your program's working for different PDR values.

4. Retransmission Timeout

Keep the retransmission timeout duration at the client as two seconds. Keep a copy of the transmitted packet to facilitate retransmission instead of reconstructing a new packet from the input file again.

5. Buffering

Server s1 is not allowed to buffer any out-of-order packet. For example, s1 is waiting for a packet from c1, and before it receives a packet from c2, s1 should discard the packet received from c2. Similarly, s1 should discard an out-of-order packet received from the same client.

6. Write to File

The server first writes the student name received from client c1 and then the corresponding student id received from client c2 to the file list.txt. The list.txt format is as follows:

Archit Nigam,2021a7ps0089,Dharma Teja, 2020phxf0042

7. Output Format

The client and server programs should produce the output in the following format: (The output format is the same for c1 to s1 and c2 to s1 communication.)

Note: If you have used any additional packet header field, you can add that to the client and server output for better clarity.

Key functionalities to be implemented by you are as follows:

• Input file reading, creating packets, and sending them over TCP sockets by c1 and c2.
• Receiving packets by s1 alternatively from c1 and c2.
• Writing the data received from c1 and c2 to the file list.txt by s1.
• Output format at c1, c2, and s1, as discussed above.
• Packet drop simulation and retransmission of dropped packets.