CN_CHomeworks3: A C++ repository from sina-tb

#CN_CHomeworks3

Code Structure

The code consists of three classes:

Edge : This class has three attributes:
- node1 : The first node of the edge.
- node2 : The second node of the edge.
- weight : The wieght of the edge.

The picture below shows the Edge class:

Node: This class has two attributes. Also, this class has some methods that can be explained by their name, so the explanation of methods are unnecessary.
- Id : Shows the id number of node in graph
- Edges: A vector of Edge objects that their first node is equal to the Id of the node. in other word, a vector consist of all edges that are connected to neighbors of the node.

The picture below shows the Node class:

Graph: This class has one attribute. Also, this class has some methods that we explain them later.
- Nodes : A vector of all nodes that have been added to graph so far.

The picture below shows the Graph class:

The following picture shows Customer class:

Customer: this class is the base of inheritance for the Provider class (so a Provider can be a Customer as well).
_prvoiders: indicates the providers for this customer
_prvLengthPref: holds a provider ID along with its length of connection and its preference to that provider

The following picture shows Provider class:

_cstmLengthPref: list of customerIDs mapped to their length and preferences.
_peerLengthPref: list of peerIDs mapped to their path length and prefernce.
_routeTable: routing information about a prefix(ASID) mapped to the its nextHop and the path length.

Picture of BGP class:

_potentialCustomers: list of Customers and Providers(which can also be a customer) on the network.

Note: we can classify ASs as follows:

Stub AS
Multihomed AS
Transit AS

For simplicity we excluded Transit AS in out project, however one could argue that having a multihomed AS is more difficult than a Transit AS, because, a Multihomed AS can be a customer and a provider at the same time.

Code Verification

Topology

The input graph is the given graph in project description:

Show

The adjacency matrix is shown below:

Modify

The modify command has been done on the graph shown in Show section:

Remove

The remove command has been done on the graph shown in Modify section:

BGP

At first after entering 2 in the menu and bgp after that, user will encounter a BGP Menu as follows:

after choosing each option a guide for the chosen command will be printed for the user. Here we will go forward on a few examples:

Add Provider/Add Customer

after pressing 1 or 2 on the BGP menu, the command guide corresponding to each option will be appeared:

for example in this photo we chose the Add customer option the ASID corresponds to id of the Autonomous System

Advertise routing

after pressing 5 on the BGP menu, we are able to announce a new routing path. the first input would be the AS introducing this route (note that for simplicity we introduce ASs instead of prefixes).

The first input indicates the provider which is inroducing the path. The second input indicates the destination which would use this routing information (More likely another provider) The third input indicates the actual prefix(i,e AS) and the final input as it was described in the hint, the cost/length of the path.

Advertise Errors

Hijacked

Whenever an incoming route to a destination with a different source of what has already been in the routing table comes this error will be printed

Role of Relations

If a AS X wants to advertise some routing information to AS Y two of which must be connected. Either by being peer of one another or by being in provider-customer relatioship

Show Providers:

after pressing 6 on the menu each of the previously added providers will be printed one by one (their IDs)

LSRP

In this section we send information about directly connected links to all nodes (not just neighbors) then we used Dijkstra's shortest path algorithm to find the shortest path.It takes n iterations(n is number of nodes).Iteration start at source node and then we use information about directly connected links to this node and find shortest path,then after we confirm a new node(shortest cost),with information of it's directly connected links,we update the paths and find the shortest one again.after n iterastions we got shortest paths from source node.

The result is shown in picture below:

Distance Vector(DVRP)

The code of this secction is shown below. As you can see, we used Bellman Ford algorithm to find distance vector. we iterate over all edges in graph and update their weight and save the lowest-price from each node to other nodes. If updating doesn't occur for all edges, we stop the algorithm: