- This topic has 101 replies, 1 voice, and was last updated 13 years, 10 months ago by Avnish Kumar.
6th December 2002 at 14:46 #32572LuigimexGuest
The answer is Yes there are related, blocking swithing has to do with it.
A switch (lets say telephonic exchanges) has a switching matrix which is designed to have a GOS of 1/100000, it is designed on several issues:
Performance of the processor
Traffic capacity for trunks groups (E1 and suscriber lines that produce BHCA).
Call service processor performance
Size of memory to handle simultaneous calls and services, etc.
The mathematical model is so complex that is hard to describe, The model I saw 10 years ago was complicated and oriented to vendor technology, other vendors follow similar model with some deviations in accordance their own switch arquitecture.
Comments or corrections are welcome12th December 2002 at 04:25 #32573Nitin GuptaGuest
One more question. How do we find out, what is the population a 10,000 BHCA switch can cater to ? Whats the co-relation / formula ? This is in case of the IP Network.
Nitin12th December 2002 at 04:59 #32574Nitin GuptaGuest
I also need to know the co-relation between BHCA, simultaneous call processings on the server, corresponding population supported in case of the IP network.
Nitin21st January 2003 at 01:02 #32575OgieGuest
How can I compute the number of simultaneous calls made during the day? Given the number of calls made and average ACD time.22nd January 2003 at 01:54 #32576ErlangGuest
I think there is no unique solution to Ogie’s question. Certainly, at one critical end, the no. of simultaneous calls can just be the no. of calls if all the calls are made at the same time. At the other end, these calls may spread evenly among the day. In other words, you can only get the upper and lower bounds to the answer:
upper = no. of calls during the day
lower = no. of calls during the day * average ACD (say in sec)/24*60*60 sec
Please correct me if I’m not right.23rd January 2003 at 19:32 #32577ChrisGuest
I would like to add one thing to Erlang’s answer. I agree with his upper and lower bounds. I would add that when we use Erlang-B to provision a trunk (or whatever) we are estimating the number of simultaneous calls. For example, if we have 15 Erlangs in the busy hour and 1% blocking we would provision 24 servers (i.e. DS0s). In essence this is saying that 1% of the time we have 24 simultaneous users (maybe more that are blocked.)
That is one way to estimate the number of simultaneous calls (assuming you can use Erlang-B). There are other tools you could use to calculate the number if you had a system that you could take measurements on as well (but I think that is another discussion.)
Let me know what you think and feel free to send me an email (firstname.lastname@example.org) if you want to discuss this in more detail.
Chris26th January 2003 at 06:10 #32578OgieGuest
Thanks Erlang and Chris! I really appreciate your answers. I just ask this question because I want to determine the time wherein the number of simultaneous calls reached the maximum.10th February 2003 at 05:08 #32579BAhaGuest
What is/are the difference(s) between Erlang B and Erlang C tables?10th February 2003 at 13:31 #32580ChrisGuest
Erlang B is used in a blocking system and Erlang C is used in a queueing system. With Erlang B the assumption is that an arriving call is either accepted into the system (is assigned a resource) or it is lost (blocked, sent to treatment, etc.) With Erlang C a call can queue for a period of time to see if a channel becomes free. If the time expires, the call is blocked. Erlang B is used in most public telecom networks (trunk provisioning, cell site provisioning, etc.) Erlang C is used in some networks where people call in to speak to a customer service rep (or something like that.)
I hope this helps.
Chris10th February 2003 at 17:14 #32581BahaGuest
that was really helpfull thank you Chris, now where can I find these tables? for both formulas? I’ve looked through the forum and I couldn’t find a link to them!11th February 2003 at 04:25 #32582ChrisGuest
If you look at http://www.erlang.com/calculator/ will find the calculators. I have not tried them. Let me know how they work for you.
Chris20th February 2003 at 10:12 #32583Saurabh RastogiGuest
a. What are the basis of planning of no of E1’s from a Mobile network to PSTN .
b. Details on traffic dimensioning for a Mobile Network.21st February 2003 at 04:39 #32584Balakrishnan S NaickerGuest
ASR => “Answer Seizure Ratio” not “Attempting Size Ratio” as sombody said above.
ASR is the ratio of the number of successful calls
over the total number of outgoing calls from a carrier’s network23rd February 2003 at 06:12 #32585madhuGuest
i would like to know the meaning of DCTA and NVM24th February 2003 at 06:07 #32586Balakrishnan S NaickerGuest
NVM: Norstar Voice Mail.
one of the functionlities of NVM is Automatic answers for customers’ calls.
Plz refer at
http://www.daviscom.com/cust_secure/nvm40_set_Op.pdf for further details.