OMEGALUL
This commit is contained in:
24
cst311/lec/lec1.md
Normal file
24
cst311/lec/lec1.md
Normal file
@@ -0,0 +1,24 @@
|
||||
# lec1 \
|
||||
First we'll define some terminology.
|
||||
|
||||
> Hosts \
|
||||
End systems - typically don't bother with routing data through a network
|
||||
|
||||
> Comm Links \
|
||||
Typically the actual systems that connect things together. \
|
||||
|
||||
## Network edges
|
||||
|
||||
Can be subdivided clients & servers and sometimes both at the same time. \
|
||||
|
||||
## Access network: cable network
|
||||
|
||||
Typically when have to share one line we can change the frequency of the signal as one method to provide a distinguishment between different data which may sometimes come from different sources.
|
||||
|
||||
### Home Network
|
||||
|
||||
Starting with the modem. All it does it take some signla and convert it to the proper IEEE data format(citation needed).
|
||||
Typically we then pipe that data to a router which then just goes around the house in whatever configuration you may need.
|
||||
Typically, especially for small businesses, star topology is common where there is some switch in the middle for traffic to be forwarded from the router/modem setup to the individual nodes on the outside of the star.
|
||||
|
||||
|
||||
52
cst311/lec/lec2.md
Normal file
52
cst311/lec/lec2.md
Normal file
@@ -0,0 +1,52 @@
|
||||
# lec2
|
||||
|
||||
> Packets
|
||||
When we send data(messages) we'll typically split the data into chunks or __packets__
|
||||
|
||||
## Network \
|
||||
> packet switching\
|
||||
Hosts break app-layer msgs into packets. These packets are moved along the same path like a train from router to router. Each of these moves from router to router we call __hops__.
|
||||
|
||||
## Store and Forward \
|
||||
We first store the entirety of the packet before sending it off. Usually we keep this in some kind of buffer where we can read what kind of packet we are dealing with through some simple parsing. This is done so that we know how large the packet is going to be since the header of that packet will tell us crucial information about it and others like it.
|
||||
|
||||
## Packet Switching: Queueing delay, loss \
|
||||
Since the routers store the packets in memory before sending them off, they store them in some queue which can sometimes fill up. For that reason if the buffer fills the router drops the packets and keeps going.
|
||||
|
||||
Packet loss at the network layer is expected which is why we usually have packet loss recovery at each end.
|
||||
|
||||
## Two key network-core functions \
|
||||
Packet headers usually tell us where the packet is trying to go and if where the other packets which stitch to them are supposed to go as well. \
|
||||
This information is usually kept in the header of the packet file.
|
||||
|
||||
```
|
||||
HEADER:
|
||||
DEST: 10101001
|
||||
DATA: 1001000 ... 00011110
|
||||
REL: <>
|
||||
```
|
||||
It is important that we keep some relativity in the stream of packets.
|
||||
|
||||
### Circuit Switching
|
||||
|
||||
FDM v TDM [Frequency Division Multiplexing] [Time Divisiono Multiple]
|
||||
> FDM \
|
||||
Multiple frequencies which remain static over time we can use for different connections.
|
||||
|
||||
> TDM \
|
||||
We use the same channel but this time we use slices of time allocated to different connections.
|
||||
|
||||
## Packet switching v Circuit Switching \
|
||||
_Packet switching let's more users on a network at once over circuit switching_
|
||||
The reason boils down to user behavior on a network.
|
||||
|
||||
Packet switching advantages: \
|
||||
* great for burst data \
|
||||
* easy to setup typically
|
||||
|
||||
Disadvantages: \
|
||||
* Packet delay and loss
|
||||
* congestion does happen
|
||||
* you need to implement some kind of protocol for data loss recovery
|
||||
|
||||
|
||||
Reference in New Issue
Block a user