NGN Signaling:
Revenue Protection for End-to-End IP Networks


Next: --->> BWA Licensing Issues

Sunday, 11 November 2007 10:01:43  - 0800 GMT


Country “A” has a plan on implementing three things (mostly regulatory)


a.         Mandatory interconnectivity between phone companies.

b.         Cheaper International calls to user.

c.         Easy MNP implementation.


1.         All the phone companies will be connected to Interconnection Exchanges. These Interconnection Exchanges should offer services like MNP, ENUM-assisted routing etc. These Interconnection Exchanges will have to connect to International (VoIP) gateways. You might consult the total layout at the end of this document. What can be the interfaces between these two layers? Currently all the Phone companies have bilateral E1 connections with each of them.


2.         Most of the Phone companies are using (some have plans for upgrading) Softswitches except for the incumbent operator. These phone companies are either owned by big telecom players and small operators have started operating afresh. Both have their kind of economic reasons for deploying Softswitches. The incumbent PSTN operator has TDM switches like some other countries.


3.         As there are no Interconnection Exchanges and proper International Gateways (VoIP) are in place in Country A, so, you can start designing with a clean slate by putting things like softswitches, session initiation protocol (SIP) application servers, media servers, and SIP-capable endpoints from the International (VoIP) Gateways to Interconnection Exchange layer! Only problem you might encounter with integrating older incumbent PSTN’s TDM exchanges.


4.         As we are planning a total IP infrastructure in both International Gateway and Interconnection Exchange layers (do not exist), we might use SIGTRAN that will interface between our incumbent’s PSTN – via a Signaling Gateway and MGCs to connect to all IP Interconnection Exchanges. This signaling gateway will eventually convert ISUP over MTP to ISUP over SIGTRAN. The MGC installed here then receives ISUP messaging; recreate the equivalent SIP for the Interconnection Exchanges. Do these PSTNs need standalone signal transfer points (STPs) when 3GPP IMS architecture introduces the IMS call session control function (CSCF)?


4a.       With my knowledge, the Interconnection Exchanges and International VoIP Gateway can be connected over the Ethernet with gigabit interface. If the incumbent operator does not upgrade their TDM based networks, they can always be connected with Interconnection Exchanges with TDM over IP as a temporary solution. Emulation of TDM circuits over the IP networks can be carried out using pseudowires (PWs). TDM over IP protocol operates over IP networks, including UDP over IPv4 or IPv6, MPLS, L2TPv3 over IP, or pure Ethernet. But how do I track number of calls (session handling) if those are encrypted? Should regulator intervene not to allow encryption on principle?


5.         We know that the Signaling networks are the best source of CDRs, not the switches. All networks want to protect their part of the revenue, what happened when incumbent sees these IP networks may incur loss to their revenue what have been done so far without hassle with TDM networks. The questions can be similar to these:


5a.       What happens when the Phone companies are allowed to use NGN signaling over IP to connect to Interconnection Exchanges and then to VoIP gateways? How the revenue is protected? What if the phone companies bypass the billing entities installed in these two layers and connect to another VoIP gateway?


5b.       Will phone companies be able to create VPN through these two layers to connect to their choice of VoIP gateways? Using SS7 over TCP/IP; like using M2UA and M3UA for MTP2 and MTP3 respectively, SUA for SCCP and SCTP with TCP might allow routing to their choice of destination.


5c.       Some solution providers (mostly, revenue assurance groups) are advocating for installing TDM interfaces between all the layers (i.e. Phone companies, Interconnection Exchanges and International VoIP gateways) to protect the revenue. It might incur more overhead cost when TDM interfaces gone obsolete. How does NGN signaling provide better visibility to what is coming into and leaving each of their networks? What kind of OSS will track all available UDP/TCP ports to prevent no VPN is being used?


6.         For MNP, ITU defines Signaling Relay Function (SRF) - a trigger less solution and intelligent network (IN) triggering - a triggered solution but the later impresses me most. What about using IN triggering when these Interconnection Exchanges will have IN functionality? For NP, as I read; ISUP-based trigger less solution needs to be used. Will it be possible to implement MNP without ISUP triggering from Interconnection Exchanges? How to get ready for MNP without installing extra solution in Interconnection Exchanges?


I know you might criticize the idea of using extra layers, but it has its advantage too. Newer services will be added to this Interconnection layer without hassle.


Should you need to consult the design part of it, here it goes: [Click on the image for a larger view]


 Figure 1.0 [Distribution of Layers]


Thanks in advance for the replies. You can post your solutions here.

[Link Removed: Solution found]



Raqueeb Hassan



Prospective Solution #>


Wednesday, 14 November 2007 06:43:21  - 0800 GMT


1. Well, there is an one stop solution for that. What if the whole IP network is exposed to external parties to run denial of service (DoS) and other malicious attacks? What if the revenue-producing elements in the service core, such as soft switches, application servers and media gateways are under attacks where service providers have built trusted and secure interconnect borders?


It's all about SBC, Session Border Controller.


2. At IP network borders, a session border controller (SBC) is used to mediate the differences in various IP networks as well as provide security, quality and cost control at the point when entering and going out of those networks. These properly configured SBCs can satisfy the critical security, service reach maximization, revenue and profit protection and SLA assurance requirements in IP interconnect and peering deployments.


3. Session border controllers are used at interconnect borders to mediate between networks and solve the challenges associated with IP hand-offs as opposed to costly TDM hand-offs for a variety of revenue generating service offerings and cost-saving applications,between service providers like Inter-connection Exchanges.

4. If you look todays complex areas of deployment - SBC's are getting common these days. It can be deployed to serve many applications but I was more concern about two service providers. SBCs can be used at the border between two service providers, where one or both have agreed to carry the other’s traffic. Thanks to Jon Hardwick's white paper as I borrowed this diagram from.


5. As in the solution note of Acme Packet it started like this "VoIP growth is undeniable. Many service providers are using IP networks to reduce the operational and capital costs of trunking PSTN traffic between central offices and the number of VoIP service subscribers is growing around the world. Yet, this growth resides in VoIP islands that, in most cases, require PSTN connections to go from one service provider network to the next. The next step in network evolution uses IP to interconnect these “islands” of VoIP to expand network reach and revenue potential while minimizing PSTN termination costs. IP interconnects enable service providers to realize several benefits."


6. Why bother reinventing the wheel? I hear you saying that. 



Raqueeb Hassan

14 November, 07



Communication has always been a human need.

We believe it is also a human right. ----- ITU


Next: BWA Licensing Issues