Abstract

Tutorial Talk 37th European Conference and Exhibition on Optical Communication, Geneva, Switzerland September 21, 2011 Telecom Network Hierarchy An Access Network (PON) Basic PON Structure • Optical Line Terminal (OLT) • Optical Network Units (ONU) / Optical Network Terminal (ONT) Long-Haul, Backbone Networks TE vs. NE vs. NP • Traffic Engineering (TE) • Network Engineering (NE) • Network Planning (NP) TE vs. NE vs. NP • Traffic Engineering (TE) • Network Engineering (NE) • Network Planning (NP) TE vs. NE vs. NP • Traffic Engineering (TE) • Network Engineering (NE) • Network Planning (NP) TE vs. NE vs. NP • Traffic Engineering (TE) – “Put the traffic where the bandwidth is” • Network Engineering (NE) – “ Put the bandwidth where the traffic is” • Network Planning (NP) – “Put the bandwidth where the traffic is forecasted to be” TE vs. NE vs. NP • Traffic Engineering (TE) – “Put the traffic where the bandwidth is” • Network Engineering (NE) – “Put the bandwidth where the traffic is” • Network Planning (NP) – “Put the bandwidth where the traffic is forecasted to be” • TE – online, dynamic, provisioning problem, ms time scale • NE – intermediate problem, months time scale • NP – offline, static, dimensioning problem, 5-yr time scale TE vs. NE vs. NP • Traffic Engineering (TE) – “Put the traffic where the bandwidth is” • Network Engineering (NE) – “Put the bandwidth where the traffic is” • Network Planning (NP) – “Put the bandwidth where the traffic is forecasted to be” • TE – online, dynamic, provisioning problem, ms time scale • NE – intermediate problem, months time scale • NP – offline, static, dimensioning problem, 5-yr time scale TE vs. NE vs. NP • Traffic Engineering (TE) – “Put the traffic where the bandwidth is” • Network Engineering (NE) – “Put the bandwidth where the traffic is” • Network Planning (NP) – “Put the bandwidth where the traffic is forecasted to be” Blocking probability • TE – online, dynamic, provisioning problem, ms time scale • NE – intermediate problem, months time scale • NP – offline, static, dimensioning problem, 5-yr time scale Cost ( $$$ ) TE vs. NE vs. NP • Traffic Engineering (TE) – “Put the traffic where the bandwidth is” • Network Engineering (NE) – “Put the bandwidth where the traffic is” • Network Planning (NP) – “Put the bandwidth where the traffic is forecasted to be” Blocking probability time scale • TE – online, dynamic, provisioning problem, ms • NE – intermediate problem, months Exhaustion Probability time scale • NP – offline, static, dimensioning problem, 5-yr time scale Cost ( $$$ ) System/Network: Value Proposition (Convergence Across “Layers”) Optical Network Design: Different Paradigms • End-to-end ethernet • Mixed-line-rate network design • Dynamic optical circuit switching • Hybrid dynamic circuit/packet switched network • Robust network design • Excess capacity management • Energy-efficient network design • Data-center network design • Broadband access network design Ethernet Everywhere  Ethernet is a success story in Local Area Networks (LAN)  About 90% of LANs use Ethernet.  Extending its reach from LAN into Metro Area Networks (MAN) has already been established.  Focus now is to extend Ethernet into carrier core networks.  Future mode of operation: Ethernet over WDM  native Ethernet frames directly over WDM.  Elimination of several layers of other technologies.  CapEx and OpEX savings.  Connection-oriented Ethernet.  Forwarding: VLAN-XC, Provider Backbone Transport (PBT), T-MPLS.  Following requirements must be taken into account:  High resilience.  Long reach: 1500- 4000 km.  Rates of up to 100 Gbit/s Ethernet (GbE).  High degree of mesh. Telecom Nets: “End-to-End” Ethernet? Transmission Rates and Mixed Line Rates  Ethernet Rates:  100 Gbit/s Ethernet.  Max possible CapEx savings.  Constraint: Signal transmission range for a certain rate  Signal’s quality depends on the physical impairments.  Transmission Range = Signal traveled distance after which signal quality degrades to a level that it needs regeneration.  Transmission Ranges:  Range of 10 Gbit/s signal = 3000 km  Range of 100 Gbit/s signal = 500 km  Mixed Line Rates:  10 Gbit/s, 40 Gbit/s, 100 Gbit/s waves.  Need for (hierarchical) grooming.  Etherpath = Lightpath carrying Ethernet frames. Single Line Rate Mixing up Line Rates Why Mixed Line Rate (MLR) Design? Mixed Line Rate (MLR) Networks • Each node can have transponders of different line rates • Same fiber can carry different line rates on different wavelengths DCS: Dynamic Circuit Switching DCS: Dynamic Circuit Switching • Emerging (video-enabled) applications: – Video downloads – Massively multiplayer games – Video collaborations – Telepresence – IPTV – Applications on a wire, etc. • If you are happy with the PMO of our networks: – slow downloads – jittery streaming – unreliable audio – Bursty (packet) traffic generated by users/applications – Aggregate traffic at the network edge – Establish high-bandwidth pipes between edge nodes through the network core – DCS offers bandwidth-on-demand capabilities to applications (users can “dial” for bandwidth) – Bursty (packet) traffic generated by users/applications – Aggregate traffic at the network edge – Establish high-bandwidth pipes between edge nodes through the network core – DCS offers bandwidth-on-demand capabilities to applications (users can “dial” for bandwidth) • Example Applications: – Real-time download (say within 5 sec) – Database/website backup (say between 1 am – 3 am, and not to exceed a 15-min duration) – Bursty (packet) traffic generated by users/applications – Aggregate traffic at the network edge – Establish high-bandwidth pipes between edge nodes through the network core – DCS offers bandwidth-on-demand capabilities to applications (users can “dial” for bandwidth) • Example Applications: – Real-time download (say within 5 sec) – Database/website backup (say between 1 am – 3 am, and not to exceed a 15-min duration) • Provision DDRs (Deadline-Driven Requests): : allows flexibility w.r.t. when to when to schedule the request. – Time dimension – Transmission bandwidth (rate) dimension: allows flexibility (adjustable to network state) w.r.t. data rate to be allocated to the request. ESNet: Hybrid Dynamic Circuit/Packet Nets Hybrid Dynamic Circuit/Packet Nets: Motivation Robust Network Design Robust Network Design • Multi-path routing – “Degraded service” vs. no service at all • Multi-path routing – “Degraded service” vs. no service at all • Multi-path routing – “Degraded service” vs. no service at all • Fast reprovisioning – (upon network-state change) – “Degraded service” vs. no service at all • Fast reprovisioning – (upon network-state change) • Multi-path routing – “Degraded service” vs. no service at all • Fast reprovisioning – (upon network-state change) • Data replication – across Disaster Protection Zones (DPZs) • Multi-path routing – “Degraded service” vs. no service at all • Fast reprovisioning – (upon network-state change) • Data replication – across Disaster Protection Zones (DPZs) • Multicast support? – for various services? Excess Capacity Management Exploiting Excess Capacity to Improve Robustness • Exploit excess capacity (EC) to improve: –Set up time –Protection switching time –Availability • Pre-provisioning to decreases the connection setup time by reserving capacity for primary and backup paths in advance. • Mixed-protection schemes that exploit EC to improve protection switching time and availability. Energy Consumption of Telecom Networks US Commercial Electricity Price Internet Traffic Profile Network Domain Utilization Network Domain Utilization Energy-Efficient MLR/OFDM Network Design Efficient allocation of flexible bandwidth network Tradeoffs • Capacity vs. all-optical reach • Capacity vs. energy consumption • Energy consumption vs. all-optical reach Data-Center Network (DCN) Network Failure due to Disasters Disaster Resilient Data-Center Network Design Energy-Efficient Content Distribution over Telecom Network Infrastructure Energy-Efficient Content Distribution over Telecom Network Infrastructure • Problem Statement –Given • Network with DC, Core, Metro, and Access • Set of content objects with known popularity • Storage availability for hosting nodes (can be infinite) • Requests of contents of the form (Client Node, Content, BW, [timestamp]) following Zipf distribution –Goal: energy-efficient content placement while satisfying the requests Long-Reach Broadband Access Wireless Optical Broadband Access Network (WOBAN)

© 2011 Optical Society of America