INTELLIGENT NETWORK: Brain for the networks

You decide to order a pizza having realized that it is well past

the meal timings and the only number you remember is a kind of 1600 free phone

number; one that pizza company has taken so much pains to advertise as its

delivery on call number. You punch in the digits on your cell phone and a 'number

error' flashes on your screen....

You pick up the calling card your company gave you to save on

the long distance calling from hotel and key in the access number and get to

hear a weird announcement telling you to check the number you have dialed....

What is wrong with these numbers? Where is the convenience of

easy to remember numbers which were toll free for information services? Why is

the calling card access denied from some phones?

These are the questions bombarded on hapless call center

executives of the telecom operators' across the country. And the reply? Just

that these are so called IN services which are not accessible across

operators!!! Meaning? A BSNL calling card will not be accessible using a AirTel

phone. Or an AirTel Toll free number will not be accessible from Reliance

phones.

The IN Concept



Intelligent network (IN) as a concept germinated out of competing operators'

requirements to offer differentiating services to their subscribers in the face

of these services being implemented by equipment vendors in a dissimilar manner.

In these systems, subscriber features and services were implemented as a large

monolithic piece of software making it difficult to make features interoperable

across systems developed by different vendors.

IN is an architecture which provides for deployment

telecommunication network architecture for provisioning of advanced services

quickly and with low risk of deployment over and above the existing

architecture. IN architecture relies on a rather simple concept. It takes out

feature handling from individual switching nodes and places the feature handling

logic and the associated data in a separate central node called the Service

Control Point (SCP). The exchanges continue to do what they do best-handling

of calls.

IN in India



The pre-privatization era of Indian telephony had already an IN architecture

in place. This was done some time in mid 1998 with the first IN service control

point in Delhi. The technology was our own indigenous CDOT who had come up with

a full-fledged IN SCP with services like pre-paid, free phone , universal access

number, virtual private network and tele-voting in place. Since there was only

one fixed line operator present at the time, the issues of inter-operator

handover were blissfully absent. All subscribers were owned by one operator and

routing them through IN was as easy as extending any other service.

The overlying advantage of IN is the ability of these so called service nodes to house huge and service specific databases independent of the actual service provider owning the subscriber

NTP '99 happened then and many more operators joined the

telecom bandwagon. The customers increased, tariffs crashed and expectations

rose. Making a voice call was not the only requirement a customer had. He wanted

a cheaper call and the mobile phones also exposed him to the possibility of a

pre-paid option. The BSNL toll free numbers also were getting to be noticed with

people actually starting to use it.

The mobile operators were using a limited addition of IN for

provisioning and handling the pre-paid subscriber base. These equipment and

solutions were usually bundled with rest of equipment including switches and

HLRs as a standard peripheral and were only restricted to the operator who was

providing these services. Any operator to roll out mobile services would do so

with a pre-paid plan, which was handled by the IN. This intelligent network, was

more of a customized pre-paid solution for the operator, rather than being a

multi-feature platform used to offer other services like free phone and virtual

private network.

The Issues



The ability to handover the IN calls across operators using the predestined

Points Of Interconnect (POIs) is technically not an issue at all. What makes it

complicated is the Interconnect Usage Charges regulations which defines the

revenue share between the originating, carrying (in case of long distance) and

terminating operator. This regulation at present is silent on revenue share to

be implemented for an IN call handover. In absence of this , because of slightly

different and complicated way of call routing, the revenue share becomes a

complex issue.

• Access providers' dilemma: The fact that no access
  
  provider wants to let go the control of their respective captive subscriber
  
  base, makes this interconnect more of a will and desire to connect issue
  
  rather than a capability one. Today, the access providers are lobbying hard
  
  for this service to be declared as an access service thereby restricting the
  
  operator base to access providers. This will deny the long distance
  
  operators to roll out any IN based service like a pre-paid card which some
  
  of them are keen to roll out.

Throwing a spanner in this already messed up service identity

issue, is the incumbent, who earlier came up with the issue of IN service being

his USP to his subscribers , thereby defying the basic concept of IN. Over a

period, and with regulator keen to go ahead with an IN interconnect, there was a

volte face demanding a completely seamless IN interconnect at the INAP level

across the IN nodes. This is another extreme of telecommunication compatibility

and is a near impossible scenario given the number of varied technologies and

systems in the country.

Indian telecom network contains almost all switching

technology and various IN platforms. Presence of CDMA, GSM and wireline makes

such interworking more complex as different interfaces are supported for IN.

Moreover, configuring all SSPs to handle call flows corresponding to all three

technologies has not been achieved in any product so far. Even for only wireline,

though every vendor comply to baseline of CS1 , they have implemented their own

variants wherever ITU has given an option (which mainly relates to charging and

the SII parameter). For announcements, as the IVR is integrated with SSF and

each operator will have their own announcement flow which would be in different

languages for different services ,the mapping of different language matrix would

again be a huge problem.

From an operations perspective, every operator would need to

share the service call flow, announcement scripts etc to all the other operators

when a service is launched or modified and this would be needed to be configured

in all the SSPs of all the operators. This is not only complicated and time

consuming but not feasible at all. One operator who wishes to launch a new

prepaid service for example could only do so when the call flow and respective

announcements have been configured in all the exchanges which are acting as SSPs

for all operators.

• Long distance service providers' dilemma: Private NLDOs
  
  are keen to get their share of pie revenue specially when features like CAC
  
  which allows subscriber to choose her long distance operators are not
  
  getting implemented. Absence of CAC enables the access providers to bind all
  
  their subscribers to the long distance operator of their choice. Integrated
  
  operators like BSNL do not handover a single minute of long distance traffic
  
  to the private NLDOs.

• Addressing the subscriber's needs: The regulator, TRAI,
  
  has tried to come to subscriber's rescue setting up an IN committee to
  
  study the issue and come up with the issues. The report, yet to be released,
  
  is supposed to dictate the IN interconnect scenario to follow and to address
  
  the concerns of all the stakeholders. Till such time it seems , that the
  
  concerns of the biggest stakeholder, the end subscriber will keep hanging.

Shyam Mardikar GM,

Technical, Bharti Infotel