Thanks... Pls I wat to explain a scenario for you Pls see attached to this message a figure This is a VoIP network which deals with only voice packets. 4 WAN models are attached to a core network (3 routers) via an edge network ( ingress node and egress node).. The ingress node accepts incoming traffic to the core while the egress node transmits packets from the core back to the WAN Now assuming a user in WAN 1 wants to send voice packets to a user in WAN 4....the ingress node accepts the voice packet to the core... Within the core, I'm employing a dynamic protocol....DV...in your opinion, is this right?
$ns rtproto Static $n16 $n17 $n18 $n19 $n20 now after routing within the core, when the voice packet gets to the egress node for transmission to WAN 4, how do I configure the egress node to be able to know that the packet is destined for WAN 4 and automatically route the voice packet initially from WAN 1 to WAN 4? Thanks On 10/30/07, Pragasen Mudali <[EMAIL PROTECTED]> wrote: > > Sorry for the late response (I'm in South Africa and its currently 8am, so > there's probably a time difference) > > The gateway nodes are responsible for routing traffic that is destined to > an outside network (in your case the core network). They are also > responsible for accepting traffic from outside networks (again, your core > network). > I think that multicast should work fine at the core. > > On 29/10/2007, Brenda Lindsay Williams <[EMAIL PROTECTED] > wrote: > > > > Thanks for your prompt reply... > > Suggestion No 1) configure your edge routers as your gateway > > nodes....what are the function of gateway nodes? > > In my model, at the edge network, the ingress node serves as a router > > which accepts incoming voice traffic from the WANs to the core while the > > egress node serves as a router which routes voice traffic from the core to > > the WAN > > In the core, for Gigabit Ethernet, i'm using 1000Mb, while for ATM, I'm > > using 100MB > > I'm thinking of using multicast at the core area which employs a full > > mesh topology... > > > > Thanks > > > > Brenda > > > > On 10/29/07, Pragasen Mudali <[EMAIL PROTECTED] > wrote: > > > > > > I'm actually working on wireless mesh networks but I have a few > > > suggestions > > > > > > 1) configure your edge routers as your gateway nodes > > > 2) define your wired mesh core network as in your model (I'm not sure > > > what type of cables/links you're trying to simulate) > > > 3) define the routing protocol to be used (if possible... like I > > > mentioned earlier I'm more familiar with wireless networks :-) > > > > > > NB. the shortest path may not always be the best path. > > > > > > On 29/10/2007, Brenda Lindsay Williams <[EMAIL PROTECTED]> wrote: > > > > > > > > Hi > > > > > > > > Thanks for the reply... > > > > I'm working with VoIP core networks.... > > > > I'm basically connecting 4 WAN models to a core network via an edge > > > > network... > > > > Pls see attached to this mail a diagram....I hope you can see it > > > > this time > > > > I'm comparing two types of networking technology (ATM & Gigabit > > > > Ethernet) at the core.... > > > > I'm using QoS parameters such as throughput, packet loss and packet > > > > delay as my measuring metrics... > > > > lemme give you an example..a user in WAN 1 sends a voice packet to a > > > > user in WAN 4.... > > > > How does the core using a full mesh topology as shown in the diagram > > > > know the best traffic route to take within the core keeping QoS > > > > parameters > > > > in mind... > > > > Thanks > > > > > > > > Brenda > > > > > > > > On 10/29/07, Pragasen Mudali <[EMAIL PROTECTED]> wrote: > > > > > > > > > > Hi, > > > > > > > > > > I can't see the diagram that you referred to, but there are two > > > > > options. > > > > > 1) random traffic as generated by the ns2 traffic generation > > > > > scripts > > > > > 2) your mesh network is used to provide Internet connectivity and > > > > > therefore the overwhelming majority of your traffic will be between > > > > > your > > > > > nodes and your Internet gateway. > > > > > > > > > > Hope this helps > > > > > > > > > > On 29/10/2007, Brenda Lindsay Williams < [EMAIL PROTECTED]> > > > > > wrote: > > > > > > > > > > > > > > > > > > > > > > > > Can anyone give me an idea of what packet traffic will look in a > > > > > > mesh > > > > > > topology? > > > > > > I need a template to work with... > > > > > > Pls attached to this mail is a diagram of what I'm working with > > > > > > to give you > > > > > > an idea.. > > > > > > Thanks > > > > > > > > > > > > Brenda > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > Pragasen Mudali > > > > > Research Student > > > > > Department of Computer Science > > > > > University of Zululand > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > Pragasen Mudali > > > Research Student > > > Department of Computer Science > > > University of Zululand > > > > > > > > > > -- > Pragasen Mudali > Research Student > Department of Computer Science > University of Zululand >
