CST 311 Week 7 Journal Entry

Link Layer

This week, we learned about the link layer of the Internet Protocol Stack. The link layer transfers frames between nodes (routers and hosts) in a network. Communication is completed by connecting each node by a link. These links can be wired, wireless, or local area networks (LANs). To transfer datagrams received from the network layer above it, the link layer encapsulates a datagram in a frame before sending it to another node. This layer uses MAC addresses, which are stored in the frame headers, to identify the source and destination of the frame.

The link layer transfers frames between nodes in a network. Switches can be used to store and forward frames to their intended destinations. To do so, a frame's MAC address is read and selectively forwarded to the appropriate outgoing, using point-to-point links to access a segment. Each switch has a switch forwarding table that maps the MAC address to the appropriate interface. Switches are self-learning and use flooding to find their intended MAC address destination if not known. Switches can also be connected to each other to help support larger networks. They work very similar to routers in the sense that they both store and forward information as well as have forwarding tables. However, routers use IP addresses and operate in the network layer, while switches use MAC addresses and operate in the link layer.

Furthermore, the link layer also helps with bit error detection. Errors can be detected by means of parity checking or cyclic redundancy checks. Parity checks can use a single-bit parity or a two-dimensional bit parity to detect bit errors. In a single-bit parity check, an extra bit is added to the data bits in a frame and is used to determine if an error has occurred by looking at the number of 1s in the data bit.  In a two-dimensional bit parity, data is arranged as a matrix, with each row and column assigned a parity bit. A more powerful way to detect an error would be to use a cyclic redundancy check (CRC). In this method, a fixed-length binary sequence is computed based on the data being sent from the source to the destination. This binary sequence is known as the CRC code and is predetermined.   The receiver takes this binary sequence and divides it by the original data in binary form plus the CRC bits using binary division in modulo 2. If the remainder is zero, no errors were detected. 

Overall, this week was a bit overwhelming. In addition to the final, we also have the readings on the assignments above and the team project to start. While it was tough to manage, it was interesting to learn about the (almost) last layer of the Internet Protocol Stack. I enjoyed learning and reading about the CRC and how it is used to check errors. The binary division caught my attention as I am familiar with using modulo.