The elongated inning of the existing National Numbering Plan is approaching an inevitable end. The plan, which was designed in 1965 and revised in 1990-91, served the country well for 35 long years. But the communication revolution has thrown up a host of compelling new services that are pressing too hard for the introduction of new codes and numbers. So the existing numbering scheme, which was designed for a monopoly operator, looks exhausted and calls for immediate replacement. Any delay will lead to violations of the ground rules of the plan, to accommodate new requirements. This will cause complexity in network operations and add to confusion among users.

The TRAI Proposal
Telecom Regulatory Authority of India (TRAI) had envisaged the situation well in advance and suggested a modified numbering plan on 24 July 1998. But while the recommended plan may take care of the present crisis, it will fall short of expectations within a short period of time. A discussion on the pros and cons of the TRAI-proposed plan is still on with major operators. In the mean time, in the last three years, the unprecedented growth of mobile and Internet users has challenged the long-term viability of the plan. A close look shows that TRAI’s plan is a compromise solution, which provides some openness to the existing numbering plan, while maintaining the existing network hierarchy.

As against the existing nine-digit PSTN format AB CDE XXXX, TRAI has proposed the ten-digit format N ABCDE XXXX (See Table 1). The proposal has reserved an N digit 2 to 5 for basic service operators and used a linked number plan. This has neither opened up enough numbers to last for future nor is easy to adopt. The reasons why TRAI’s numbering plan is not likely to take off are:

• Conservative approach to change the basic service number plan hampered the value proposition of the plan

• The plan does not simplify the network structure or call-handling process, as it has followed the existing hierarchy

• It has not opened up enough numbers for future; hence, it has not generated enough interest among operators to take up the challenge of implementation

• It has not considered some of the upcoming services like VoIP, which will be important for our country due their affordability

As the change in the numbering plan in a national network with millions of customers is a major exercise, it is worth reviewing closely the proposed numbering plan of TRAI.

National Perspective
Historically, in all our plans, we have under-estimated the potential and need of the population. Many of our plans did not survive their life-cycles due to under-planning. The present numbering plan never considered penetration of communication to masses. Hence, the available numbers are inadequate. The telephone-based numbering plan and trunking plan are of miniscule size in comparison to the future requirement and are not flexible enough for dramatic expansion.

Also, over-utilization of the present number plan has created certain anomalies in the network. New plan should aim at bringing uniformity and eliminating undue advantages of incumbent operators. The new plan should be capable of handling enough unique numbers to accommodate every individual of the country for the next 50 years.

Due to our large population, we need many more numbers than any of the developed countries. The government plans to provide 150 million connections by 2010. Our problem is more complicated than any of the developed countries and therefore borrowed solutions from other countries will not be sustainable.

India is a vast country with a variety of vintage networks. Any changes in this heterogeneous network will be difficult, time-consuming and mostly manual. However, in most cases, the network is voice-centric, hierarchical and follows a uniform tariff structure.

Changing the numbering plan for a small customer base is more practical and cost-effective today than in future. As the change in numbering plan is inevitable to support the expected explosion of customer base, it will be comparatively simpler to do it at this stage.

Modify for Growth
TRAI proposed a 10-digit linked numbering plan, applicable to all telecom services. For STD and ISD calls, as well as for carrier selection, one has to dial extra prefixes before these 10 digits.

TRAI proposed N ABCDE XXXX as the standard number length, to be used by all telecom users of the country.

We agree with TRAI to adopt the 10-digit common numbering scheme. However, we have suggested a different way of usage of this scheme, as NABC DEXXXX.

If no restriction is imposed, the scheme may accommodate 1x1010 – 1 numbers, ie 10 billion numbers.

At present, India has a population of more than one billion. In the next 50 years, it is expected that the country’s population will be two billion (with 2 percent simple growth). To provide a unique number as communication identity for each Indian, we need to design a scheme which can support at least three times of 2 billion, i.e. 6 billion numbers.

N being the most significant number in the scheme, loss of one digit in N means loss of 1 billion numbers.

It is suggested that N should be used in the most open manner and as far as possible it should not be used for identification purposes. This is important to keep the number length to 10 digits. Otherwise in the near future, we may need to proceed to an 11-digit scheme.

New plan allows utilization of all 10 digits. Billions of numbers are saved as the most significant digit N is neither used as access code nor as service pointer.

We suggest to allot N = 2 to 5 to all access providers, (as defined by NTP’99). This will provide for 4 billion numbers to access providers. Even with one-third utilization of 4 billion numbers, we may sustain for a long time.

As N=6 and N=9 are kept spare, 2 billion numbers will be available for future allotment. Hence, the scheme is meeting the requirement of 6 billion numbers.

By avoiding the use of the most significant digits NABC for identification of services or addressing level-1 and level-2 taxes, wastage of a big chunk of numbers can be avoided. A uniform 10-digit numbering scheme will be available for all end users and no prefix will be needed to dial any number within the country.

Ample Support for Future Services
To support convergence of services, various equipment capable of handling specific protocol will be introduced as new elements in the network along with existing voice switches. These elements and switches, which will handle various intra-operator, inter-operator and inter-service traffic, may be called nodes of the network.

A node may be a transit exchange; IP switch or a high-end router used in network to handle inter-operator traffic, which includes voice, data, voice over IP and multimedia traffic.

It is recommended that nodes of different services and of different operators be interconnected through a suitable interconnect exchange for network simplicity and for the ease of inter-operator protocol handling.

In NABC DEXXXX scheme, NABC is recommended as the address of any node. We further recommend NABC to be allotted as signaling point code of the node. Hence to direct a call to a node, no translation or digit-by-digit analysis is needed.

The most significant fourth digit of the national dialing code shall be the SPC of the called node. SPC is an unique address of a node in a complex network.

A node defined above may support 1 million numbers. These million numbers may be sub-divided to many network elements connected to the node. Network hierarchy below the node may be as per the specific service requirement.

An access provider may have one or more nodes. They are to be interconnected and to be in the same license area. Calls between nodes of the same operator in the same service area are expected to be handled by operator-owned/leased interconnect links or may be routed through interconnect exchange of SSA.

Addressing the Nodes
By using NABC, we can address 9,999 nodes. As N=0, 1, 7 and 8 are fully or partly allotted for other purposes (See Allocation of ‘N’ table), usable address for nodes will be 5,999.

These numbers are to be allotted to access providers (defined in NTP ’99), without wasting any series for special identification. Any access provider will get a minimum bunch of 1 million numbers for a node. The million numbers may be subdivided into smaller exchanges, according to the need of the network. At present, India has less than 400 Level-1 and Level-2 TAXs, a few tandem exchanges and about 100 MSCs of cellular operators working as inter-operator transit exchanges. Even after considering multiple operators in every service, 5,999 nodes with 6 billion numbers will be there for access service providers.

It is highly recommended that Level 1 and Level 2 TAXs be replaced by or upgraded to interconnect exchanges, with multi-protocol handling capability. NLDOs will collect/deliver the calls from or to interconnect exchanges and deploy only one or two inter-circle nodes in a circle. The interconnect exchanges and inter-circle NLDO nodes, though critical transit points of the network, will not be identified by subscriber-dialed numbers, as they are neither the originating nor destination nodes of the call. Hence, SPCs of these transit nodes should be different and are not addressed by the NABC part of any number starting with N = 2,3,4,5,6 or 9. Hence, 6 billion numbers specified for access providers are untouched.

For Maximum Utilization...
Some simple measures can be effective in increasing the addressing potential of the TRAI-proposed numbering scheme. Let’s look at some problem areas and the possible solutions.

Problem #1: Significant digits
Whenever we use the most significant or subsequently higher order digits for specific identification purpose, we lose a huge chunk of precious numbers. By allotting N=1 for emergency and special dialing numbers, for example, we loose 1 billion numbers out of 10 billion numbers of the 10-digit numbering scheme.

Wastage is also due to region-wise, circle-wise and LDCA/SDCA-based divisions. Our network is at present hierarchically divided into regions, circles, LDCAs, and SDCAs. A series of numbers are allotted to them. Numbers in different significant positions are used to identify each stage of network hierarchy. Number analysis is needed in each stage to physically route the call.

Solution: Service identification through significant numbers is to be avoided as far as possible.

Problem #2: Allotment
Allotment of 92, 93, 94 and 95 to mobile operators in the TRAI proposal theoretically delivers 400 million numbers for cellular operation. TRAI proposes to use ABC as MSC address. Then only it will allow 400 MSCs in the network each of 1 million capacity ie 400 million numbers. This is almost in line with our proposal.

Solution: All access providers should be grouped together and a common pool of number be allotted. Each operator should be provided with a minimum of 1 million numbers, without any discrimination. No prefix should be used for service identification, and service providers should be monitored for optimal utilization of all the million numbers allotted.

Problem #3: Number of digits
Let us take an example of two STD calls made to two adjacent SSAs, namely, Delhi and Gurgaon, from Ahmedabad.

In the call to Delhi, where 011 MNP XXXX has been dialed by Ahmedabad subscriber (011 is STD code, MNP is Delhi’s local exchange identity and XXXX is the end-user identity in that exchange). On the other hand, in the call to Gurgaon, 0124 MNP XXXX has been dialed (0124 is STD code of Gurgaon, MNP is Gurgaon’s local exchange identity and XXXX is the user identity in that exchange).

The number-length of both the calls are different—10 digits for Delhi and 11 digits for Gurgaon. This is because STD codes are of different lengths—011 for Delhi and 0124 for Gurgaon.

Solution: The suggested 10-digit format will remove this disparity.