Providers delivering a carrier-grade IP trunking solution will see rewards.
After ten years of proving out the technology, and after an economic downturn in 2010, businesses are purchasing IP telephony like never before. The business benefits from decreased costs, more efficient networks and increased productivity have driven demand to what has now become a robust market with mainstream technology.
Migrating businesses from traditional PBXs and key systems to an IP trunking solution offers the potential for huge decreases in ongoing operating costs, such as T1 leasing and maintenance. But beyond simply reducing costs, SIP trunking can provide enhanced services and increased productivity to end users while generating increased revenue and subscriber loyalty for service providers.
Infonetics Research reports that the worldwide SIP trunking market is expected to grow from the current 2.6 million SIP trunks deployed in the first half of 2011 to 22.3 million by 2015, an increase of more than 770 percent in 5 years.
Few markets currently offer this level of growth potential for either telco or cable service providers.
SIP trunking network
As attractive as the business case may be, any service provider planning to move into the SIP trunking market must plan and implement a solid infrastructure and process in order to deliver a carrier-quality SIP trunking product.
There are many moving parts in a carrier-grade VoIP network, and service providers must carefully plan and implement their SIP trunking offering with an end-to-end view of how they will provision, deliver and support their service.
While there are differences in the required network equipment for MSOs versus carrier networks, the majority of the core equipment and the major concepts of delivering a SIP trunking service are identical. Figures 1 and 2 indicate the different equipment in a cable network and a telco network.
The key components of a SIP trunking deployment are:
- SIP trunking application/feature server and media server – Responsible for providing trunk group features, as well as add-on applications.
- PSTN-connected media gateway or softswitch – Responsible for bridging the service provider’s SIP IP network to the public switch telephone network (PSTN).
- Session border controllers – Responsible for providing voice-aware firewall functions, as well as network-based network address translation service.
- Backbone IP network – Responsible for carrying the SIP and RTP IP traffic between IP-enabled media and application servers.
- Backend provisioning systems – Responsible for the provisioning of all required network components to support SIP trunk groups.
- Network management server – Responsible for gathering and managing call data required to effectively manage and troubleshoot the SIP trunking network.
Service delivery challenges
Unlike TDM-based networks, where each call is dedicated the required resources to facilitate the call on a channel basis, SIP trunking places calls into IP packets and places the packets onto an IP network for delivery. Each packet must compete for bandwidth with other packets on the network. When the number of packets exceeds the capacity of the network, the voice packets can become delayed or dropped, causing the voice conversation to be choppy.
In order to ensure carrier-grade voice service, providers must understand how they will address QoS in their SIP trunking network.
The first, and perhaps still the most common approach, is to deliver VoIP and data traffic on two separate Internet connections, with the VoIP connection sized to preclude any potential congestion.
When access links are shared, the typical solution employs a QoS router at the customer site. This provides absolute priority to outbound VoIP packets and protects inbound VoIP packets through active traffic shaping – i.e., by signaling remote TCP hosts to throttle inbound TCP data flows so there is enough capacity for inbound VoIP packets.
One of the largest challenges faced by the first service providers to offer SIP trunking is that SIP, while a standard, can be (and most often was) implemented differently from one IP PBX to the next.
This meant that early service providers were required to perform interoperability tests with each SIP-based IP PBX or IAD, in which they planned to connect, identify any issues and work directly with their vendors to resolve issues before they could support a new IP PBX.
In order to resolve these interoperability issues, the SIP Forum formed a new user group called SIPconnect to define specific recommendations of how the SIP interfaces were to be implemented (SIPconnect 1.0), provide standard testing to ensure that devices would interoperate and certify vendor equipment as SIPconnect-certified.
Since the initial SIPconnect standard was ratified, the SIPconnect user group has expanded to include cable providers to also address the requirements of how SIP trunking is to be delivered over packet cable networks, as well as to further tighten the recommendations to ensure interoperability between devices in SIPconnect 1.1.
Without a doubt, SIPconnect 1.0 and SIPconnect 1.1 have done much to alleviate the pain of managing the interoperability of many IP PBXs and IADs, although there is still no guarantee that all features will work, even between SIPconnect-certified devices.
In order to address this issue and to allow a greater level of flexibility in the number of devices that can be supported in their SIP trunking network, many service providers have started to deploy SIP normalization gateways.
SIP normalization devices are most often deployed on the customer premises between the IP PBX or IAD and the service provider’s SIP network. These SIP normalization devices generally come with a wide range of certified IP PBXs and IADs, as well as the ability to perform SIP header customizations to allow service providers to quickly and easily certify new IP PBXs or IADs. Although fairly new to the market, these devices are increasing in popularity.
Support and troubleshooting
In order to deliver a quality SIP trunking product, service providers must have the ability to support and troubleshoot the solution from end to end – from the core application server, through the core and access networks, to the demarc router or gateway, and into the customer’s local area network. There are many products that will allow service providers to monitor and troubleshoot VoIP traffic end-to-end, although most do not support the crucial ability to constantly monitor all calls 24/7. The specific product selected does not matter quite so much, so long as it can provide the crucial detailed historical data for all calls, all the time. Detailed historical call data allows the service provider’s support staff to effectively troubleshoot past issues, often while the subscriber is on the phone. Without this historical data, the only way to troubleshoot a problem is to set a trap for a specific network link, customer or individual subscriber and wait for the problem to recur.
Concurrent call model with over-provisioning
A common way to offer SIP trunking to customers is to provision and charge based on the number of concurrent calls required. For example, a customer with 100 users may only require having 20 concurrent calls active from the IP PBX to the service provider’s network at one time. In this case, the service provider provisions the user with 20 concurrent calls and will block any additional calls so that the 21st concurrent call is not possible.
If the customer has a requirement for 20 concurrent calls normally – however, due to seasonal possible bursts of call traffic requires up to 40 concurrent calls – this customer must be provisioned and billed for 40 concurrent calls to ensure that no calls to or from the business are blocked. This is referred to as over-provisioning.
Concurrent call with burstable SIP trunk groups
An alternative to this model requires that the service provider be capable of supporting a feature referred to as burstable SIP trunk groups. This feature allows the service provider to configure the SIP trunk group for a set number of concurrent calls, with the ability to define a second parameter that allows the customer to “burst” above his or her set limit. When a call “bursts” above the set limit, a call detail record (CDR) is generated to inform the service provider that the customer has exceeded the normal contracted capacity.
For example, a customer with a help desk that normally only needs 20 concurrent calls but could need as much as 40 concurrent calls during an outage, they could be provisioned and billed for 20 concurrent calls, with the ability to burst to 40 concurrent calls at a premium for each call over 20.
More revenue from add-on apps
With ever-more service providers launching SIP trunking products in order to take advantage of this quickly growing market, competition has driven down margins for origination and termination. In order to increase the average revenue per user (ARPU), many service providers are beginning to look to high-value, add-on applications and services such as hosted voicemail, conferencing, etc.
These add-on applications and services are especially attractive to customers using a legacy PBX with IP trunking to an IAD or gateway. Most legacy PBX’s have few of the newer high-value features, such as end user Web portals, find me/follow me, simultaneous ring, visual voicemail, etc.
SIP trunking provides an excellent delivery mechanism by placing the applications and services in front of the PBX to perform intelligent routing functions such as forwarding calls from VIPs to users’ mobile phones or ringing multiple phones simultaneously.
Even though many of these services may be available on the customer’s IP PBX, the additional cost of hardware and licensing may be cost-prohibitive. For example, the customers may be able to license hosted applications, such as auto attendant or audio conferencing, for a fraction of the cost required to purchase them from their IP PBX vendor. Figure 3 illustrates how service providers are targeting these types of value-add applications and services to multiple customer segments.
In addition, these hosted add-on services can be very “sticky,” making it much more difficult for the customer to shop for a new SIP trunking vendor based purely on price.
Business continuity using SIP trunking
Business continuity offers the customer the ability to continue sending and receiving calls in the event of a network outage, power outage or natural disaster. Because the SIP trunking solution resides in front of the customer’s IP PBX and provides advanced routing intelligence, the service provider has the ability to provide alternate routing if the customer’s PBX is unreachable. Most SIP trunking solutions provide the customer with a self-management Web interface that allows a customer administrator to configure how calls will be routed if the PBX is not reachable on the network. This could include configuring main numbers to be routed to another location and having the users’ individual DIDs routed to their mobile phones or voicemail. This ensures that all critical calls will continue to be handled.
If the customer has multiple locations, a common way of managing business continuity is to simply route all critical calls from one location to the other in the event that the IP PBX is unreachable. This solution requires that each location has the proper resources required to manage calls for both sites in the event of an outage. This requires that each site be either over-provisioned or provisioned with burstable SIP trunk groups. Figure 4 illustrates a customer with two locations configured for burstable SIP trunk groups.
SIP trunking is now the fastest-growing area of the VoIP marketplace, and all indications point to sustained growth for the foreseeable future, making it a very attractive business for both telco and cable providers.
While it is impossible to predict the future, two things are certain: 1) The majority of business customers are still using legacy TDM trunking to connect to the public network; and 2) The vast majority of business customers understand the value proposition and plan to move to SIP trunking in the next five years.
The rewards for service providers to effectively deliver a carrier-grade IP trunking solution are easily measured, both in terms of new revenue streams and satisfied customers.