Abstract

The paper describes some relevant solutions for Next Generation Access Network (NGAN, or NGN), considering evolution from copper legacy network to optical access network (PON, Passive Optical Network) and analysing pros and cons of several PON solutions (TDM-PON, WDM-PON, TWDM and OFMD-PON) as well as potential convergence/coexistence scenarios. Keywords—copper network; NGAN; PON; GPON; XG-PON; TDM-PON; WDM-POM; TWDM-PON; OFDM-PON; I. EVOLUTION OF COPPER NETWORKS In the last years several techniques (e.g., bonding, vectoring, phantom mode) have been developed to increase the performance of xDSL (Digital Subscriber Line, providing maximum 100 Mbps for single line), widely used in copper networks [1]. Currently the G.fast technology is under standardization at ITU, based on the mentioned technologies allowing copper networks to provide “up to” 1 Gbps symmetric (downstream and upstream) bitrate per user on short loops within 250 m), i.e. for access network in FTTB (Fiber-To-The-Building), FTTdp (Fiber-To-The-distribution-point) or FTTC (Fiber-To-TheCabinet) architectures (see Fig.1) [2, 3]. Fig.1. Main FTTx architectures. In line with “Horizon2020” objectives to provide 100 Mpbs to everyone, the G.fast technology should allow upgrading existing copper infrastructures in order to reduce “digital divide” in low economic return areas (hardly linkable through FTTH optical architectures) and to protect investment made on legacy copper network. 1 With VDSL2 (Very-High-bitrate DSL 2). 2 1 Gbps over short distances (50-100 m), some hundreds of Mbps for longer spans (within 200-250m). 3 Fiber reaches user’s building (few tens of meters far from final users). 4 Fiber is terminated on a street cabinet usually within 300 m from user home. 5 Fiber-to-the-home: fiber is terminated on the user’s home (usually in a box on the outer wall) – no copper wires used. The most appropriate access network should be carefully assessed (see Fig.2), since costs to realize new optical access are huge (civil works account for 60-80% of total CapEx) [4]. Fig.2. Evolution of Access Networks (Alcatel-Lucent) Although G.fast seems rather promising, it is hard to believe that 1 Gbps (or more) can be provided to every user in every country / area with copper networks, since there will always be areas with “local loops” longer than 100 m – the length of “last mile” greatly influences performances. Moreover, the vectoring technique is not compatible with LLU (Local Loop Unbundling) since different lines / signals of the same DSLAM (Digital Subscriber Line Access Multiplexer) can belong to different Operators: vectoring performs a pre-distortion (“precompensation”) of all signals / lines belonging to the same DSLAM in order to reduce noise and interferences (mainly crosstalk). Furthermore, new broadband services / applications are getting to the market (e.g., Digital TV HD and 3D, HD videocommunications, e-health, e-learning, online interactive games, cloud computing, immersive / virtual reality), and copper (even with technology evolutions) is likely not able to support these services or to enable the birth of others needing very broadband. Therefore, it is mandatory to plan an entirely optical access network (FTTH / FTTB) for the near future, providing at least 1 Gbps per user. In the following, FTTH configurations are considered – unless otherwise specified. 6 There are several EU study groups focussing on next generation PON solutions; in particular, the FSAN (Full Service Access Network) has outlined a NGN-roadmap to evolve the current PON systems onto NG-PON1 first (abandoned afterwards) and then onto the NG-PON2. II. PON (PASSIVE OPTICAL NETWORK) While for business customers it is reasonable to consider P2P (Point-to-Point) solutions, whereby each user has a dedicated optical link with Central Office (CO) like in copper networks, for residential customers it has been demonstrated that P2mP (Point-to-multi-Point) solutions – where a single fiber link is shared among multiple users – are more convenient and provide adequate performance. A study case performed in Lombardy region by the NGN Italy Committee has shown that a P2P solution would cost about 70% more than a P2mP solution (GPON) [5]. The most adopted P2mP solution is the PON (Passive Optical Network), where the ODN (Optical Distribution Network) is entirely passive. The most installed PON is the GPON (Gigabit PON), characterized by 2.5 Gbps downstream e 1.25 Gbps upstream aggregated bitrates for a single fiber link, shared usually among 32 users (thus providing 80/40 Mbps downstream / upstream to each user). The GPON solution allows providing triple-play services (i.e., phone, Internet, TV) using 3 wavelengths: 1490 nm (downstream), 1310 nm (upstream) and 1550 nm (video overlay). Since GPON is based on TDM (Time Division Multiplexing), it requires a MAC (Medium Access Control) to avoid upstream collisions. OLT (Optical Line Termination) transmits downstream signal broadcast to every ONT / ONU and they read only their respective packets. In the upstream direction, each ONT has a “grant” to transmit only in specific time windows. In the last years the XG-PON1 (or 10G-PON) solution has been developed, that upgrades (and can coexist with) legacy GPON systems, providing 10 Gbps downstream and 2.5 Gbps upstream [6]. One of the disadvantages of TDM-PON solutions concerns unbundling: since each optical link is not dedicated to a single user (like in copper networks or in P2P optical networks), it is not possible to provide LLU. Although there are other options to perform unbundling (e.g. bit-stream), the possibility of providing LLU is a key factor to conciliate different Operators’ interests and to unlock the realization of NGANs. There are several options, like WDM-PON (Wavelength-Division Multiplexing-PON) and OFDM-PON (Orthogonal FrequencyDivision Multiplexing-PON), allowing a “virtual” access to the physical medium using wavelength division: as a matter of fact, it is possible to assign a single wavelength / channel (or a group of channels) to a specific operator, sharing the same physical medium among multiple operators. A current and complete overview of LLU for European market is in [7]. Key factors for next generation access solutions are: ability to operate on existing ODN (ODN reuse), support of LLU 7 After OLT only fiber links and power splitters are used, without signal amplification or regeneration. 8 It is possible to use higher split ratios (theoretically till 1:128), but rarely used because of its huge impact on maximum reach. 9 The term ONT (Optical Network Termination) is used for FTTH configurations, and the term ONU (Optical Network Unit) for other architectures (e.g. FTTB). In the following, the term ONT is used for the sake of simplification. 10 The incumbent Operator, that built the access network, has to allow other operators (OLO, Other Licensed Operators) to provide Internet service to customers using a part of the existing network. 11 Theoretically is possible to use a single wavelength for both directions, but there are strong limitations due to backscattering phenomena (Rayleigh e Brillouin). For this reason, the most used systems are based on 2 different (possibly up to eight independent Operators) and compatibility with video overlay and XG-PON1 systems.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call