For a variety of bandwidth-rich applications, the implementation of dynamic synchronous transfer mode (DTM) combines the best of IP and modern optical transport.
By Dr. Christer Bohm
Net Insight
In today's competitive marketplace, telecom operators are faced with decreasing revenues from telephony and Internet services, as well as competition from MSO/CATV operators. The key to survival is differentiation and finding new, attractive services that compensate for lost revenues. To meet the competition, many operators are looking at IPTV as a major new service.
Recent price reductions on servers and access equipment have made IPTV services a viable offering, but current networks are not adapted to handle large amounts of rich media. On top of that, support and maintenance costs become major burdens for operators, often exceeding capital investments. While triple-play services do have great potential, how can they be implemented without running telecom companies into the ground?
Background
With the evolution of optical networking and the Generalized Multi-Protocol Label Switching (GMPLS) framework, resources can be reserved and services can be provisioned through the use of bandwidth channelization and the optical layer.
Channels based on TDM guarantee the transport of data, thereby ensuring the quality levels demanded by TV and video services. Traffic belonging to one stream is completely separated from other traffic throughout the network, eliminating congestion or delay variation in the network. TDM allows the quality of service (QoS) to be guaranteed with a maintained high utilization in the network.
While packet networks are unable to give strict guarantees on a service, the TDM layer provides 100 percent QoS. TDM layers also facilitate functions like performance monitoring, protection switching, and framing. Overall, TDM technology is simple in structure, reliable, and easy to manage.
Traditional TDM technology, however, has a number of drawbacks in data and video networks. Establishing a connection can be time consuming and the hierarchical infrastructure often limits network utilization. In addition, traditional telecom infrastructures transport data at fixed rates, which can cause problems when trying to incorporate new services. Because data and video signals do not conform to telecom rates, configurable sized channels are essential.
DTM and hybrid networks
Combining standard network protocols with the dynamic synchronous transfer mode (DTM) standard overcomes the limitations of standalone GMPLS and TDM networks. DTM, a TDM technology that enables high-speed optical transport, merges data, video, audio and voice into a single, hybrid network, allowing the system to run directly on fiber or other existing standard transport connectivity, essentially combining the best of Ethernet data access and next-generation transport technologies.
In DTM, an optical control plane is used to provision all data, video, audio, and voice services, providing several benefits when it comes to provisioning channels and integrating a signaling control plane into all products. DTM assures 100 percent QoS for video traffic, even if the network load is over 95 percent. This technology enables a high utilization of infrastructure investments and reduces capital expenditures. Unlike traditional TDM, DTM's dynamic channel structures can set up channels of varying size without diminishing video quality, and all channels may also be seamlessly multicasted to any number of ports, efficiently supporting IPTV and other multicast applications.
For very high capacity, QoS-demanding applications, DTM can be used as a standalone technology. However, it is typically used in a multiservice environment combined with routing and Layer 2 switching to form a hybrid architecture that combines the best of IP and modern optical transport. Here, QoS-demanding traffic is short cut and best effort and less demanding traffic is aggregated and statistically multiplexed in routers and Layer 2 switches.
System advantages
With DTM, a channel can be provisioned up to the full link speed. By adapting to the service instead of existing network conditions, DTM is able to provide 300 percent more bandwidth than traditional network architectures. DTM channels have dedicated bandwidth and dynamic routes between network ingress and egress, capable of quickly passing data from end to end. This means address information no longer needs to be checked and packets don't have to be stored on buffers. As a result, there is no risk of overflow, loss, or congestion.
Several other advantages can be seen with hybrid IP and DTM networks.
- Media-ready IP and Ethernet QoS transport: A hybrid IP and DTM architecture enables low-jitter and lossless transport of IP, making TV and video services economically scalable to large deployments by offering "one IP hop" transport and avoiding shared resources to disturb the QoS.
- Supports any network topology: DTM uses variable-sized channels in conjunction with a dynamic optical control plane to simplify network engineering and decrease network expansion costs significantly. At the same time, it is able to automatically re-route channels and provide protection for critical services with minimal down time. The flexibility in channel size enables efficient mapping of different service traffic onto the infrastructure.
- Incremental scalability: The network automatically detects and connects several links at varying speeds, enabling networks to be built out step by step instead of upgrading all at once. To simplify network planning, links can be added between any nodes, regardless of size, to increase overall capacity. Furthermore, the switching delay is very short and always constant for a channel. Multi-hop channels display the same properties as a channel on a single link, except for a slightly longer delay. Strict resource allocation and the lack of shared buffers for switching data prevent any overflow or congestion inside the network. The high quality for the transported traffic is independent of the size of the network.
- Signaled end-to-end provisioning: To achieve high utilization and maintain a low-cost operation, it is critical that service provisioning be a simple task to perform. This is one of the main objectives of GMPLS. From the beginning, it has been central to the DTM standard. DTM technology uses a separate channel at all times for specific control information. This simplifies provisioning, avoids individual node management, and allows for third-party provisioning. The best approach is to keep it simple. When setting up a channel, it is ideal to specify end points and capacity then let the network find the shortest path. DTM technology allows channel capacity to be changed at any time, shortening and simplifying provisioning.
- QoS-capable multicasting: A substantial part of next-generation, media-rich broadband networks are based on distribution services, requiring efficient network usage while maintaining QoS. To meet these needs, DTM systems can provision all channels to have any number of receivers. For example, mapping IP/MPEG services on top of the quality-preserving TDM channels solves this issue for multicast.
DTM enables network operators to increase revenues by implementing new services and entering new markets. Operators also benefit from increased utilization of the existing network. High utilization means fewer fibers/wavelengths, fewer ports, and less equipment needed.
DTM approaches are especially suited to the following segments:
Broadcast and media networks: Optical communication has evolved to allow high-bandwidth transport and falling prices on fiber capacity, making optical communication a very competitive alternative to satellites and tape delivery. Media network services based on new optical transport technologies can radically change the economics of content distribution.
Most broadcasters have made the transition from analog to digital and many are heading towards HDTV, meaning major infrastructure upgrades. When it comes to building larger media networks, traditional communication approaches have suffered from scalability, cost, flexibility, and QoS issues. DTM supports transport of the media industries' own formats with maintained quality and optimal network utilization.
The broadcast industry is shifting towards IP-based production since most of the new studio equipment has IP/Ethernet interfaces. This will, however, not relax the requirement on the transport, meaning that low latency, low jitter, lossless transport is of importance to support interactive editing, live transmission, and other advanced applications. Combining IP and DTM offers a smooth, risk-free migration to IP, avoiding the QoS issues of normal router- or MPLS-based networks.
Connection to a media WAN simplifies footage trading and enables faster turnaround times in production. DTM enables uncompressed video to be sent between studios, allowing full quality throughout production. Video transport can be combined with other services, establishing the possibility of remotely controlled editing machines and cameras. This optical approach enables faster production with sustained quality and improved efficiency.
DTT and mobile TV distribution: Many countries are shutting down analog TV transmissions and migrating to digital terrestrial TV (DTT). IMS research anticipates a worldwide annual growth rate of more than 43 percent for DTT households. At the same time, competition for TV customers is growing, with TV services now provided over various platforms like mobile TV and IPTV. To justify building national DTT networks, the overall cost needs to come down. DTM platforms feature powerful multicasting and efficient resource management, making it the preferred technology for DTT networks. It enables contribution applications over the same infrastructure.
Service-rich broadband access networks: Broadband access over xDSL or FTTx is one of the fastest growing networking markets. While these networks were initially built for high-speed Internet access alone, low profit margins and competition have led operators to consider IPTV services. While other technologies have problems with low utilization and scalability to large deployments, DTM can maintain utilization of more than 95 percent with guaranteed QoS for the IP video services in large installations. Also, the service management is greatly simplified using DTM for IPTV transport.
Traditionally, incorporating IPTV services meant multiple uplinks and costly new equipment. DTM offers enhanced QoS and multicast functionality for efficient triple-play networks. Video traffic is separated from other traffic, meaning a direct route from end to end, ensuring a QoS even at very high network loads.
Cable-TV transport networks: Cable-TV networks are being upgraded for more effective video transport, HDTV, and more interactive services. DTM platforms offer effective video and data transport that is always guaranteed due to the inherent service separation. Compared to traditional alternatives, the capacity of the underlying network infrastructure is utilized three to four times more.
Most video servers use Gigabit Ethernet for sending interactive video and on-demand services. Since DTM offers full-quality transport for Ethernet traffic alongside the multicast, these services are easily added to current networks.
DTM and IPTV virtual headends
IPTV can sink or save rural telecom operators competing against intruding cable carriers. Cable operators already offer triple-play services encompassing video, data, and voice. Telecom operators must integrate video to retain and build subscriber levels.
Telecom operators have mastered voice and data, but video presents new hurdles. The cost of ingesting and transmitting video tops the list. Expensive IPTV headend costs start at $500,000 and can easily scale to $2 million over a five-year period. Initial hardware costs alone can cripple rural telcos.
Virtual headends offset the costs of IPTV services by amortizing the investment across multiple telecom cooperatives. In this scenario, multiple telecom operators share capital, integration, content aggregation, and distribution costs to create a single, networked headend. Content is then multicasted throughout the shared network to deliver the needed programming and services for each subscriber. The result is triple-play capabilities at a fraction of the cost.
Developing virtual headends require numerous technology considerations. Among the primary infrastructure concerns are:
- Scalability: The infrastructure must be scalable to share the cost with as many telecom operators as possible. The network must also be able to add and remove telcos as needed. The same goes for new programming and service options.
- QoS reliability: Telecom operators must guarantee a consistent high level of video quality delivered to the subscriber. It is important that the signal be delivered in a predictable and easily monitored manner.
- Highly efficient transport: Video transport must be highly efficient to preserve the business case of IPTV virtual headends. Distribution costs need to be minimized by optimizing bandwidth utilization.
- Highly available transport: The transport network must be highly available.
Multicasting the IPTV signal simplifies transport while further improving QoS across the network. DTM delivers scalable multicasting capabilities at Layer 2 at a fraction of the overhead costs of the complex Layer 3 multicasting used by other approaches.
Implementing and managing backbone networks for video transport and IPTV can be an operational nightmare. Considerable manual configuration and fine tuning is needed, not to mention the hefty changes required to add or remove telco customers and programming options. As the installation grows, this burden increases exponentially. DTM creates an automatic signaled control plan to offset the operational burden - provisioning and other operational tasks are handled automatically.
Summary
The DTM standard creates media networks that significantly reduce operating expenses, allowing operators to successfully launch new media oriented services and migrate legacy services towards IP. The technology cost-efficiently enables triple-play and media services over public network infrastructures, making it a great fit for broadcast, media networks, IPTV, as well as digital and mobile TV. Operators can upgrade existing networks to better utilize capacity and offer additional services, making triple play an attainable business endeavor.
Dr. Christer Bohm is CTO and co-founder of Net Insight (http://www.NetInsight.net). Prior to founding Net Insight, Bohm researched multimedia communication in the areas of optical IP-based communication and high-performance I/O systems at the Royal Institute of Technology in Stockholm. Bohm holds 28 patents and patent applications.