- This topic has 14 replies, 1 voice, and was last updated 11 years, 7 months ago by
.
-
Posts
-
I’m evaluating the voIP solution of an IP-PBX company and it seems they have a distributed architecture and add lines & trunks depending on my needs.
They say that the Erlang model does apply only to the trunk to pstn part of the picture and they warranty that their PRI’s can support up to 23 simultaneous calls or their analog trunk switches can support also 23 simultaneous calls.
so my question is. How can I relate the Erlang traffic model to calls originated from the VoIP network?
(calls can be made from IPPhones, softphones or analog-to-voip converter’s in an unlimited way)
In other words.. I do not see how the Erlang model applies to the IP-PBX except in the trunk-topstn interface.
am I correct in my appreciations or there is something else I should take into account
thanks
any help would be trully appreciated
😉
The Erlang model is just as applicable to VoIP systems too, to argue it doesnt is to show a complete lack of knowledge of telecommunications principles (note I say telecommunications, not voice!)
In a voip network there will be a requirement for a certain amount of voice traffic, expressed as erlangs. CCS, call minutes or whatever. Erlang tables convert this to a quantity of ‘channels’ to carry that traffic. Each ‘channel’ is the equivalent of the voice compression standard (G.711, G.729A/B etc) plus IP overheads.
For instance, 10 Erlangs of traffic requires 18 lines/channel to carry that traffic. At 64kb (G.711) compression plus Ip overhead each ‘channel’ occupies 80kb, therefore the MINIMUN bandwidth for your network needs to be 18 x 80kb= 1.44Mb just to carry voice. This is easily achievable across a local 100Mb LAN, but must be considered if you are considering VoIP via say a 256kb DSL style connection.
Increase the number of phones, ERLANGS increase, bandwidth requirement increases. Once you hit b/w near 10Mb you may need to investigate the capability of hubs, switches and routers!
Throw in some IP video-conference and your LAN may just begin to sssllloooowwwwww down.
thanks for your reply. first of all.
I was not clear explaining my point originally.
I of course see a relation between the erlang traffic model and the VoIP transport across an IP network (bandwidth concerns)
For my particular case I’m starting from the point of a VoIP deployment in LAN conditions (100 Mb network, even Gb links), and for the amount of calls that I will ultimately intend to support, my bottleneck is the PSTN access (because I depend on the PRI’s/ Analog trunks)
I guess I’m curious to know how the Erlang model applies to this case (when you have enough bandwidth to support even VoIP Linear codec calls.)
I am using the Erlang model to calculate the distribution of PRI trunk channels versus VoIP endpoints depending on our foreseen traffic needs an obtain the peak BHCA/BHCC/CCS/Erlang figures that this equipment must support
I am using the number of PRI channels, the amount of peak trafic (typical BHCA/BHCC rate) (% of acceptable calls lost due to circuit unavailables) to get the figures. But assuming that the IP network is not an issue for bandwidth concerns, I don’t see how a VoIP approach is different than a traditional approach (this is what I am curious indeed)
Our WAN VoIP traffic is minimal and to some extent not prioritary these days (Inbound PSTN still produces 98% of our traffic).
thanks anyway for the information
VoIP benigger
Forget the fact that the transport media is IP, understand the behavior of the VOIP as..
TDM —VOIP —TDM connection, the same calculation applied to TDM circuits applied to IP circuits once the only thing that changes is the transport media and the compression provided by the codecs on the media gateways.
follow the result of the 0.35 milierlangs that you used to have on TDM, however this will work only for Prepaid calling cards, Corporative users or Home users, if you try to apply the calculation to wholesale carriers you will find that every carrier will use the available ports in a 1 to 1 relation begining PEAK-TIME.Regards,
Saul Bejarano
That’s true for conventional telecommunication. However, VOIP is runned over packet switched network. What about multiplexing gain ?
Could you please let me know what is the formuale to convert Minutes of Usage to erlang?
When calculating the bandwidth required on a link for a VoIP connection, I have not seen any consideration for Layer 2 (Data Link) protocols or silence suppression. The Cisco calculator page ignores both.
DSL could incur significant overhead in the PPPoE link layer, reducing the call capacity. Ethernet service such as VLAN will see the ENET header on each packet. And don’t forget the required inter-frame interval for Enet, where applicable–that’s like extra bits in each frame.
Yes, you are correct. Layer two protocols can be signigicant. You could check our VoIP Select product, which takes these into account.
how to derive erlang from minutes of usage? what is the relationship between erlang and minutes of usage?
pls.give me the answer
1 Voice call = 21 Kbps with compressed speech.
in TDM 1 call = 64 Kbps
Please clarify 1200 concurrent calls means how much bandwidth in terms of MBPS if I am using coded G723 and G729?
Suppose, My network Capacity is for 10K BHCA. For That 10K BHCA how many STM-1 we needed and How it is calaculate?
Through STM-1 you can connect 30×63 TDM concurrent calls. But in case of VoIP, this number will be total BW of STM-1 (approx. 150Mbps)divided by per call bandwidth requirement in Mbps.
10K BHCA is basically calculated for switch capacity and STM-1 is channel capacity for sending voice/data. This relationship will vary in case to case basis. In BD scenario, without bandwidth optimization technology and using G.729 codec, you can handle 10K ~ 12K concurrent calls.how can i translate “erlang in circuit switching” to “kbps in packet switching” ?
It depends on the Codec used on VOIP.. G711 or G729 or G723
- The forum ‘Voice over IP’ is closed to new topics and replies.
