The capacity-based interconnection model must entail a transparent and non-discriminatory offer from PTC to OSPs of a given capacity of interconnection services, optionally to the temporized interconnection model, at geographic Points of Interconnection (PoIs) specified in the RIO, at a fixed price (i.e. flat interconnection tariff).
The flat interconnection tariff is based on the capacity purchased, irrespective of the volume and duration of the traffic actually routed. Capacity purchased is measured in multiples of the basic capacity unit to be defined later.
The capacity-based interconnection model involves PTC?s provision of network resources aimed at filling interconnection orders from operators purchasing a given capacity to route eligible traffic, in accordance with agreed quality and availability goals, with associated traffic transfer costs, to encourage efficient and sound use of capacity-based interconnection.
3.1 Capacity-based interconnection offer beneficiaries
3.2 Traffic and services eligible for capacity-based interconnection
3.3 Definition of basic capacity unit
3.4 Resale of capacity-based interconnection units
3.5 Traffic overflow conditions
3.7 Deadlines for poi creation/extension and circuit migration
3.8 Minimum service agreement time and quality of service parameters
3.9 Price calculation methodology for capacity-based interconnection
3.1 Capacity-based interconnection offer beneficiaries
ICP-ANACOM foresees no reasons to restrict the types of organizations which may benefit from the capacity-based interconnection offer based on the existing specifications of the RIO. Beneficiaries will thus be those currently specified in the RIO (public telecommunications network operators and providers of telephone services at a fixed location, mobile telephone services and data transmission services).
| Question 1: Do you agree that the capacity-based interconnection offer?s beneficiaries should be those currently specified in the RIO? If not, indicate who the beneficiaries should be, specifying reasons why. |
3.2 Traffic and services eligible for capacity-based interconnection
The capacity-based interconnection model is compatible with voice traffic and dial-up Internet access traffic.
In addition to standard switched traffic interconnection services, call origination and termination services, essentially indirect access, make up the fundamental interconnection support. For this reason, the following traffic should be eligible for capacity-based interconnection:
a.) Call origination: local, single transit and double transit
b.) Call termination: local, single transit and double transit
Access to the following services is excluded from eligible traffic for capacity-based interconnection due to their highly variable rate schemes, complexity in terms of additional dispersal over various existing support routes and the unique features of the final services offered, all of which have impacts on the size and management of interconnection resources:
a.) intelligent network services, namely: toll-free (800), local-call-rate (808), virtual calling card (882), etc.
b.) emergency services (112) and short numbers (117, 118, etc.)
c.) value-added services such as audiotext (601), televoting (607), etc.
d.) international call termination and transit traffic
| Question 2: Do you agree that the types of traffic to be used in capacity-based interconnection should be alike (i.e. voice and data)? Do you agree that the services eligible for capacity-based interconnection should be access (call origination) and call termination services at the local, single transit and double transit interconnection levels? If not, indicate which services should be eligible for capacity-based interconnection, specifying reasons why. |
3.3 Definition of basic capacity unit
Currently, temporized interconnection between operators is structured around a basic transmission network unit, the 2-Mbps circuit6https://www.anacom.pt/render.jsp?contentId=55134. If the basic interconnection capacity unit were a 2-Mbps circuit, the capacity to be purchased by an OSP would then be a multiple of 2 Mbps.
A second definition for basic capacity unit could also be explored: the 64-Kbps circuit7https://www.anacom.pt/render.jsp?contentId=55135. Such an approach would present two distinct advantages:
1. Permit easier access to the interconnection model by OSPs currently having no need to purchase a 2-Mbps circuit merely for this purpose
2. Permit more flexible planning in terms of the capacity to be purchased, adjusted to the needs of OSPs, in particular at lower-traffic PoIs, minimizing risks of inaccurate traffic forecasting
On the other hand, adopting a 64-Kbps circuit as the basic capacity unit could lead to certain complications, namely:
1. Substantial change to the network structure (since physical support would always be a multiple of 2 Mbps, unable to be unbundled)
2. Increased complexity in planning, implementing and managing interconnection, which may be viewed as disproportional since impacts would only be seen at the local level, given that the primary beneficiaries of the RIO currently use multiple 2-Mbps circuits per PoI for single transit and double transit
3. The need for greater processing capacity at switching exchanges
| Question 3: Which basic capacity unit should be considered: 2 Mbps or multiples of 64 Kbps? Specify reasons why. |
3.4 Resale of capacity-based interconnection units
The possibility of reselling capacity-based interconnection units to third parties exists in Spain, both in the capacity-based interconnection model (Point 9.4 of Telefónica?s RIO) and the temporized interconnection model.
At the time of its response to the public consultation regarding the imposition of obligations in wholesale markets for call origination and termination (markets 8 and 9)8https://www.anacom.pt/render.jsp?contentId=55137, PTC asserted that introducing a flat interconnection tariff would give rise to serious competitive imbalances, particularly for larger-sized operators, by offering surplus capacity to smaller operators at prices below cost.
With resale, operators buying interconnection capacity are not only responsible for correct forecasting of retail demand, but also wholesale demand from third-party operators through capacity purchased, thus avoiding surplus wholesale demand for capacity. In some ways, this can actually stimulate development of the wholesale market by providing the option of sharing previously-purchased capacity with other operators, thus enhancing business opportunities.
In this way, there appear to be no reasons at the outset for any restrictions to the possibility of reselling capacity-based interconnection units to third parties.
| Question 4: Can you see any disadvantages in reselling capacity-based interconnection units to third parties? If so, specify the disadvantages and indicating specific resale prevention methods and how these would be implemented. |
3.5 Traffic overflow conditions
By introducing the capacity-based interconnection model and installing (through PTC) the capacity forecast by OSPs, the required capacity may be more than the amount purchased, thus causing sporadic congestion of capacity-based interconnection resources at the outset. Note that according to the current RIO, traffic interconnection circuits should be sized so that the loss at each interconnection route does not exceed 1%.
Thus, regardless of whether operators forecast their capacity-based interconnection so as to avoid congestion, all traffic exceeding the capacity purchased under the flat-rate regime should be subject to transfer under one of two optional scenarios:
Option 1: Through circuits associated with temporized interconnection at the same PoI. In this case, a compensatory payment is required to offset additional costs from the incorrect sizing of the capacity-based interconnection by the OSP, which has impacts on the capacity of other resources. This amount must be high enough to encourage correct planning of capacity-based interconnection routes and effective sharing of the inherent risks of forecasting demand. Following overflow at capacity-based interconnection circuit(s), the OSP must pursue the necessary procedures to extend the number of circuits (in accordance with the current specifications of the RIO).
Option 2: When all capacity-based and temporized interconnection circuits at a given PoI are busy, transfer of eligible traffic should be done by means of the current agreed system between the operators, i.e. if overflow is done using interconnection circuits at another PoI, this PoI?s interconnection prices (temporized model) would apply for the interconnection level in question.
With regard to applicable pricing for traffic overflow Option 1, ANACOM believes an amount equalling 5 times the temporized interconnection price is appropriate. As previously discussed, this scenario has been used in Spain, and ICP-ANACOM has no knowledge of any inequities resulting from this solution.
It should also be noted that another possibility exists, also foreseen in Spain, involving the option of interconnection without overflow in which all exceeding traffic is simply lost. The CMT, in its resolution dated 10 July 20039https://www.anacom.pt/render.jsp?contentId=55138, chosed to introduce the no-overflow option, although given its reduced interest in Spain, it is not believed to be a relevant option to pursue in Portugal.
| Question 5: Do you agree with the proposed model in which all traffic exceeding the capacity purchased under the flat tariff is subject to overflow? Do you agree with setting a traffic overflow price to encourage efficient and sound use of capacity-based interconnection, and in particular, the reference price specified by ICP-ANACOM for ?Option 1? (equaling 5 times the temporized interconnection price)? If not, indicate the methodology you believe would be most fitting to set this price and its corresponding reference value. |
3.6 Procedures for capacity purchase agreements and for migrating from the current interconnection model to the capacity-based interconnection model
The capacity-based interconnection offer must be characterized by transparency, efficiency and speed, and thus warrants specific procedures to be implemented in the RIO, including:
1. Means of communicating capacity/migration order (responsibility: OSP)
2. Means of communicating acceptance/refusal of the order (responsibility: PTC)
3. Means of communicating service activation (responsibility: PTC)
A procedure must also exist for communicating anomalies in the deployment/migration process:
1. Means of communicating anomalies (responsibility: PTC)
2. Means of communicating resolution of anomalies (responsibility: OSP)
In this context, ICP-ANACOM sees no reason, in principle, to change the PTC/OPS communication methods currently defined in Appendices 7 and 8 of the RIO, namely with regard to ordering circuits and other means of interconnection. Capacity/migration orders must be issued in writing to the contact point designated by PTC, which must keep records of all orders/refusals made for a minimum of three years.
| Question 6: Do you agree that procedures associated with purchasing interconnection capacity with PTC should be similar to the PTC/OSP communication procedures currently specified in the RIO? If not, specify which procedures you would change and why. |
3.7 Deadlines for poi creation/extension and circuit migration
To start, the maximum deadlines for creating and extending PoIs should not depend on the interconnection model (temporized or capacity-based), and are currently defined in Sections 13.4 and 13.5 of the RIO, corresponding respectively to:
a) Maximum deadlines for creating a new PoI:
1) Order analysis and PoI deployment: 22 working days
2) PoI deployment following order confirmation: 45 working days
b) Maximum deadlines for expanding an existing PoI:
1) Cases involving changes to the network structure or replacement/expansion of transmission equipment: 1 month
2) Other cases: 15 working days
The RIO should also establish a deadline for migrating temporized interconnection circuits to capacity-based interconnection and vice versa with no delays to the maximum deadline for expanding an existing PoI. On the other hand, given the need to deploy the model and guarantee user interests, migration order confirmation deadlines will have to be proactive, and thus ANACOM believes it is appropriate to establish the following deadlines:
a) Migration order confirmation deadline from temporized interconnection model to capacity-based interconnection model (and vice versa): 5 working days (note that this is identical to the deadline specified by the CMT)
b) Maximum deadline for migration (the same as the current deadline for extending an existing PoI, in the case of temporized interconnection):
1) Cases involving changes to the network structure or replacement/expansion of transmission equipment: 1 month
2) Other cases: 15 working days
| Question 7: Do you agree that deadlines should be established (deadlines for creating, extending and migrating PoIs from the temporized interconnection model to the capacity-based model and vice versa)? If so, explain which deadlines should be established, what their maximum time periods should be, and why. |
3.8 Minimum service agreement time and quality of service parameters
Alterations to be introduced under the new interconnection model do not imply direct changes to current interconnection service quality parameters and levels designated in Appendix 3 of the RIO, namely the quality of OSP networks and circuits and losses in interconnection trunks.
In order to deploy the new capacity-based interconnection model, PTC will be required to make changes to its network planning and network structure as well as its associated information systems. In this respect, a minimum service agreement time must be set for interconnection capacity to promote stability in interconnection and adequate planning of interconnection traffic.
In light of these issues, the Authority believes that the minimum service agreement time should be two years, the same as the time period practiced in Spain. At the end of this time period, the OSP may opt to continue, alter or terminate the capacity-based interconnection agreement with no applicable penalties to be imposed by PTC. In the event of non-fulfillment of this minimum time period, namely through premature cancellation of basic capacity units or premature migration, partially or in whole, of the capacity purchased for a given PoI, a reasonable compensatory fine to be defined in the RIO shall apply.
| Question 8: Do you believe it is necessary to establish service quality levels and indicators for capacity-based interconnection? If so, which service quality levels and indicators should be used to monitor the deployment of the capacity-based interconnection offer?
Question 9: Do you agree with a minimum service agreement time of two years to promote stability in interconnection and adequate planning of interconnection traffic? If not, specify what the minimum PTC/OSP service agreement time for capacity-based interconnection should be, and why. |
3.9 Price calculation methodology for capacity-based interconnection
The main principles to be upheld in calculating interconnection prices are that prices should be determined based on actual service costs and should foster the model?s economic continuity. In other words, prices should reflect the cost of long-term efficient service, including a reasonable return on investment, an average level of return for the operator providing the capacity and a simultaneous reduction in unit costs for operators requesting this same capacity through the most efficient anticipated use of the capacity.
Therefore, a relationship will be set between the monthly capacity price and the per-minute price by means of the criterion used for sizing: forecast monthly traffic.
Calculation of minutes switched during peak hours
Interconnection between operators is structured around a basic network unit of 2 Mbps. The number of 2-Mbps circuits to be purchased is determined by two parameters:
- Number of simultaneous conversations during the busiest hour (in terms of calls routed), i.e. during the peak hour (PH)
- Loss of calls in interconnection (service level B) = 1%10https://www.anacom.pt/render.jsp?contentId=55139.
According to the Erlang B formula11https://www.anacom.pt/render.jsp?contentId=55140, during peak hours, for a basic capacity of 2 Mbps:
| No. of Circuits | Traffic Intensity (Erl) | Percentage of Occupancy | Minutes Routed (at PH) |
|---|---|---|---|
| 3112https://www.anacom.pt/render.jsp?contentId=55141 | 21.19 (for B = 1%) | 68.35% | 1.271 (31 x 60 x 68.35%) |
The number of minutes routed during PHs is obtained by multiplying the number of circuits (31) by the number of minutes in an hour (60) by the percentage of occupancy (68.35%). If telephone traffic were perfectly stable throughout the day and the month, one could simply multiply the above estimated amount by 24 hours (per day) and 30 days (per month) to obtain the amount of 915.00 minutes/month/basic unit. However, these assumptions would be incorrect given the normal levels of telephone traffic. Thus, the total estimated number of minutes routed per month in basic units is given by:
Minutes per month = Minutes at PH / WTr x WD x M
According to normal traffic profiles, it is generally assumed that:
1. During the busiest hour, an average of 10% to 15% of the total daily traffic is routed. Thus, WTr = weight of PH traffic on daily total.
2. To calculate average monthly traffic, the number of daily minutes is multiplied by the number of working days, normally between 20 to 25 per month. Thus, WD (working days per month) = 20 to 25.
3. One or two months should be subtracted to account for holiday periods where traffic is generally reduced (normally July and August). Thus, M = 10/12 or 11/12.
Examples:
| Minutes at PH | WTr | Daily Minutes | WD | M | Minutes Routed in One Month (2 Mbps) |
|---|---|---|---|---|---|
| 1,271 | 8% | 15,887 | 23 | 11-Dez | 334,961 (1,271/8% x 23 x 11/12) |
| 1,271 | 10% | 12,71 | 22 | 10-Dez | 233,017 (1,271/10% x 22 x 10/12) |
Estimation of Tariff for Capacity-Based Interconnection
Finally, the average price of the basic capacity unit can be determined by multiplying the minutes associated with the unit by the average per-minute interconnection price for the level in question (local, single transit or double transit), which is specified in the RIO for temporized interconnection.
Basic unit price = minutes per month x price per minute
| Question 10: Do you agree with the methodology and parameters used to calculate capacity-based interconnection tariffs, based on the per-minute price of temporized interconnection and scheduled monthly traffic? If not, specify the calculation methodology and parameters which you would propose, and why. |
-----
6 Generally supports 31 bi-directional interconnection channels at 64 Kbit/s. It is assumed that 1 of the 32 channels of an E1 (2 Mbps) is used for synchronizing. Normally only one channel needs to be allocated to signalling per every 10 E1 of interconnection capacity.
7 Basic unit used in the Spanish model for capacities purchased, per PoI, up to 4 E1s (120 channels). Multiples of 5x64 Kbps must be purchased for higher capacities.
8 /streaming/Grossistas-BE-GRUPOPT.pdf?categoryId=136402&contentId=246816&field=ATTACHED_FILE. (only available in portuguese)
9 http://www.cmt.es/cmt/document/decisiones/2003/RE-03-07-10-00.pdfhttp://www.cmt.es/cmt/document/decisiones/2003/RE-03-07-10-00.pdf.
10 According to the RIO: ?Circuits for interconnection should be sized so that the traffic lost at each interconnection trunk does not exceed 1%, with the amount of the loss calculated using the ADPH method over Erlang B in one week of observation each month.?
11 Erlang B is the most commonly used traffic model to determine how many channels are necessary for a given amount of traffic flow (measured in Erlang) during the busiest hour (peak hour).
12 31 64-Kbps channels may be used in each 2-Mbps circuit in the event that no signalling circuits exist. There is a reduced number of signalling circuits in the interconnection networks between PTC and OLOs: with up to 20 2-Mbps circuits, only one signalling circuit is needed (quasi-associated mode, up to 10 2-Mbps circuits).
