Analysis Tool for Passive Optical Access Network

— Fiber optic access networks have been considered a definitive solution to the constant need for increasingly higher bandwidths. For this reason, PON (Passive Optical Network) have grown enormously in the recent years, offering different solutions to the end customer, whether it is residential or business. However, the PON project is complex and a time-consuming process. This paper is focused on a tool design that, when combined with the schematic of the PON project, performs the survey of all equipment used in the projected cell. This tool provides also the attenuation calculations, as well as the techno-economic study of the network. Simulations were carried out to test the viability of the software, considering the network analysis in two cases: with one service provider using the network or with two service providers sharing it.

I. INTRODUCTION We currently live in an interconnected world, in which telecommunications shorten distances between continents and people. A world where the major challenge for service providers is to guarantee the greatest possible bandwidth for the largest number of users, with different terminals and services.
The installation of optical fiber in the access network, to provide adequate bandwidth and quality of service for the different services, is an expensive process. While the consumer demand for bandwidth continues to grow, solutions for the development of access networks have been studied [1]. The Fiber-To-The x (FTTx) solution becomes the focus of Service Providers (SP) in the development of the network, aiming to bring the fiber as close as possible to customers and make the network compatible with future fiber optic technologies. [2] discusses the future direction of optical access networks and proposes a new optical access architecture that will support the evolution of optical access systems from a Fiber-To-The-Home (FFTH) infrastructure to a common optical access platform that connects various devices and systems to anywhere they want to reach.
The project of an optical fiber network, with the goal to increase the bandwidth, the transmission distance and the number of users, needs to take into account all the parameters associated with its implementation. Thus, the realization of an access network project becomes a complex process, which Analysis Tool for Passive Optical Access Network This application was developed in MATLAB to complement the AutoCAD design tool. With this application it is possible to analyse the dimensioned networks, through the reading of the AutoCAD files. Furthermore, it is possible to calculate the loss budget for every client, identify the required components list and estimate the values involved in the implementation of the network. This latter feature provides the techno-economic analysis of the dimensioned network, in order to understand the impact of one or two SP in the global investment.
The remaining of this paper is organized as follows. Section II presents a new structure to read AutoCAD file projects of a PON and details the software's analysis and the algorithm used in different scenarios. Section III presents the input arguments for the simulation process and discusses the obtained results from the tool. Finally, Section IV enumerates the main conclusions of the paper.

II. ACCESS NETWORK SIMULATOR
In order to be able to associate a telecommunication network's project developed in AutoCAD, to the proposed tool, it will be necessary to introduce a new structure in the AutoCAD file. The presented solution consists in the introduction of names, attributes and labels in AutoCAD blocks.
With this approach it is possible to obtain all the information required to analyze the complete PON design and obtain the techno-economic study of the network.

A. Suggested Structure
For the extracted file to be readable in the software tool, it is necessary to add attributes to all the blocks in the AutoCAD project. Since in AutoCAD the representation of fiber cables is done using polylines, it will not be possible to assign attributes to them. Therefore, it was decided to add a label, thus making it possible to assign attributes to the fiber cables. Therefore, fiber cables will have nine attributes: the destination equipment, the source equipment, the quantity of cable fibers, the dimension in meters of the aerial cable, the dimension in meters of the duct cable, the dimension in meters of the building's front cable and the quantity of transitions between different optical cables locations, which can be performed by air, on the building's front and or in the underground.

B. Implemented Algorithm
As we can see from Figure 2, the results tab in the software shows the required components list, the loss budget, the location of the splitters and the financial forecast for the next 4 years.
In the required components list, all the material in CO (Central Office) is considered, but not the ones from the OLT (Optical Line Terminal). In the OSP (OutSide Plant) construction, it is considered the lengths of each type of fiber cables, the number of installed optical distributed points, the splitters in JSOs and their distribution points, the number of metallic tube risers, etc. In the ISP (Inside Plant) construction, it is considered the optical distribution point's modules (primary and secondary), the lengths of the cables installed in the vertical column, the number of floorboxes, etc.
To calculate the loss budget, Figure 1, can also be used as a support to understand the network's attenuations. According to [8], the total attenuation of the path is given by equation (1), where is the total attenuation, is the fiber attenuation coefficient, L is the fiber length, is the attenuation of the connectors, M is the attenuation of the multiplexer, F is the attenuation of fiber's fusions and LS is the splitters' attenuation. The attenuation values are given in Table I, where the splitters' attenuation values are given by the maximum insertion losses given in [10]. Regarding the number of splitters present in the splitting points, with the help of labels created in AutoCAD, it is possible to calculate the total number of splitters in the network's, splitting points and distribution points.
As for the economic part of the network, the total network price is calculated, according to the material used and the cost of licenses. The break even point is, by definition, the temporal point where the total income equals to the total invested, in this case, in months. In the scenario of two SP sharing the infrastructure, the total cost of the network will be divided by the two and the break even point is half the total cost of the network, assuming that the SP share the network's cost. The total net income is calculated for the first 4 years of the net.
The percentage of FTTH users is given as the ratio between FTTH users, with the ODP located inside the building's facilities, and FTTB (Fiber to the Building) users, with the distribution point located in the building's facilities [11]. In the techno-economic study of the network, CAPEX (Capital Expenditure) defines the cost of all equipment types and its installation for the network construction [12]. In our study, we consider the CAPEX cost, rather than the OPEX (Operational Expenditure) cost. The OPEX defines the cost of renting any type of network infrastructure or equipment, and the network's maintenance.
III. SIMULATION RESULTS Technical-economic analysis of a telecommunication's network is essential for the SP. In this concrete case, the main objective is to design a fiber optic network to provide broadband telecommunications services in the area of Lisbon city, in Portugal. The simulated cell is composed by 2113 users, being 1925 residential and 188 commercial [9]. Figure 2 illustrates the graphical interface of the software tool and the obtained results for one SP.

A. Input arguments for one SP
As it can be seen, the software's output displays data, such as the network required component list for its implementation, the network cost for each SP, the number of months until a SP pays the investment, the number of FTTH clients and the economic forecasts for the next four years. A simulation was carried out for this cell, for one SP, considering a 70% of penetration rate for the telecommunication's services. Table II represents an example of monthly fees for different service plans, delivered by a SP and its penetration rates for the covered customers. The penetration percentages of a plan are referring to the number of customers that are covered by the 70% of the network's coverage. It is concluded that, after four years, 98.88% of the residential customers covered by the network will subscribe to the services offered by the SP. It is important to note that other values could be used and the conclusion can change with different options.

B. Analysis of obtained simulated results for one SP
The required components list represents the survey of all the materials necessary for the developed network, that is, all the components that make up the network from the CO, ODN, and the ISP network. For the material price, the used table is available at [13]. Using the developed tool, the total cost of the network will be 266.574,27€. In this amount is not considered the cost of active equipment to be installed in the CO (OLT and RF overlay), nor the cost of renting telecommunications infrastructures and the network maintenance, since these items are considered in OPEX. Figure 3 shows the network's financial forecast over four years after the network was implemented. Also, in the same figure it is possible to see the break-even point, and in this case will be one year and nine months after the network is active.   To support the construction of the network, there is a feature in the software that provides information about the location of all the splitters calculated in the network, in each JSO and ODP.
Thus, Table IV represents the location of the splitters calculated in the network simulation.
The optical power budget is the amount of power needed to successfully transmit signals from a distance, via fiber connection. Table V represents a portion of the computed path attenuations.
Using this output will be possible to analyze if all connections are according to the standards. Terminal) and OLT are used in the network [14]. When implementing an access network, the SP has to choose which class of active equipment will be used. For example, for the ODP which ID is 11200121 (see Table V), the class B+ in ONT and OLT can be used, since class B+ can handle a Power Budget up to 28 dB [14], [15]. The formula for power budget calculation [14], [16] is given by  Figure 4 illustrates the graphical interface of the software tool and the obtained results for a network shared by two SP. As it can be seen, the software's output displays the same data as the previous simulation, but now for both SP. A simulation was carried out for this cell, for two SP, considering a 100% of penetration rate for the telecommunication's services, being able to cover the whole cell. Table VI and VII represents an example of monthly fees for different service plans and its penetration rates for the covered customers, delivered by a SP 1 and SP 2, respectively. D. Analysis of obtained simulated Results for two SP As mentioned previously, the components list represents the survey of all the materials necessary for the developed network, that is, all the components that make up the network from the CO, ODN, and the ISP network. Using the developed tool, the total cost of the network is 288.007,21€ which will be shared by both SP 1 and SP 2. For the material price, the used table is available at [13]. In this amount is not considered the cost of active equipment to be installed in the CO (OLT and RF overlay), nor the cost of renting telecommunications infrastructures and the network maintenance, since these items are considered in OPEX. Figure 5 shows the network's financial forecast over four years after the network was implemented for both SP. Also, in the same figure it is possible to see the break-even point for both SP. In this case, the break-even point will be reached one year and two months for SP 1 and 9 months for SP 2, after the network is active. Table VIII represents the network's performance over four years.  The cost of implementing a cell, being or not shared by two SP, ranges between 250.000 € and 300.000 €, using the prices from the table found at [13].

C. Input arguments for two SP
Considering that both SP use the same network, the number of splitters in JSO will be the same, although will be added splitters inside the buildings to compensate the higher number of clients.
Therefore, Table IX represents the splitters that were added in the builds and a portion of the computed path attenuations. As mentioned before, the introduction of new splitters in the network will cause higher loss. Using this output will be possible to analyze if all connections are according to the standards. For this example, and for the ODP0032, which ID is 11200106 (see Table IX), is the only ODP that is allowed to have services, if we consider the B+ class in both ONT and OLT [14], [15]. The projected cell, when simulated for one SP, had a cost of 266.574,27 €. In the case of monthly fees, for example plan R1 (see Table II), it is a price well below any competitor, thus aiming to attract any potential customer. However, this also results in a significant reduction of income. On the other hand, a break even point would be reached in about a year after the network is deployed.
At the end of four years, about 100% of potential customers are considered as customers of the SP and the network would yield 1.282.809,55€.
It can be concluded that, when dimensioning an access network, one of the most important points in a techno economic analysis is the network penetration rate. A SP, even with more expensive plans, but with a lower penetration rate, will take longer to reach a break even point. It is also important to take into account good practices in the access networks design. For example, a special care should be considered in the amount of fibers estimation for the network, in order to avoid the splitter usage, since they are the most important elements for the attenuation estimation.
ACKNOWLEDGMENT This work is funded by FCT/MCTES through national funds and when applicable co-funded EU funds under the project UIDB/50008/2020-UIDP/50008/2020