Making the Packet Connection with Circuit Emulation-over-Packet: A Technology Primer
In the heady days of the telecom boom, hype led many to believe a full-scale network revolution was underway. Circuit switched infrastructure was going the way of the dinosaur, the forklifts were ready and running on the frontlines, and IP networks were on the rise.
A few years later, economic realities have driven the forklift back into the garage. The boom years’ buzz has become a more realistic hum, and revolution has turned into a manageable network evolution.
While the evolutionary pace has slowed from what was predicted post-telecom crash, carriers still face the same challenges when planning a migration from a circuit-based to packet switched network (PSN) infrastructure. Primarily, they need an efficient and cost-effective solution for transporting revenue-generating legacy services over the PSN.
One emerging solution meeting the economic and competitive issues pushing carriers towards network convergence is Circuit Emulation over Packet (CEoP) technology.
The Carrier Crunch – Driving Towards the Converged Network
The TDM-based circuit switched network has been the heart of the Public Switched Telephone Network (PSTN) for decades. Some think TDM’s days are numbered, while others believe this venerable technology it will co-exist alongside IP for years to come.
Regardless, much of the world’s voice traffic is still carried over the circuit switched network and accounts for a significant portion of carrier revenues. However, data traffic now outpaces voice traffic on the network.
Carriers are caught in an economic and competitive crunch. Recognizing the tremendous growth in data traffic, carriers have started to build out IP networks on top of their circuit infrastructure. Fiscally, this is only a short-term solution. The bottom line makes it impossible to maintain two disparate networks for voice and data traffic, each with separate support staff, unique hardware and different management systems. Cable providers and wireless operators are also aggressively moving off the sidelines, and are now circling the carriers’ once secure voice dollars.
Carriers are fighting back, and clearly the path forward is convergence towards one single PSN. They are installing new data networks with VoIP to transfer revenue-generating services to a less costly network, and announcing more IP or Ethernet services based on true Ethernet networks, not on circuit switched networks. This trend will continue as carriers endeavor to reduce their costs.
But the major issue remains – carriers need to transport legacy traffic originating from the home, enterprise and campuses over the new network.
The challenge of transporting circuit switched trunks over connectionless networks has been meet by both integrated circuit manufacturers and equipment vendors, with a solution that combines old and new approaches to transport TDM traffic over new IP, Ethernet or MPLS networks.
The agreed approach is patterned after Circuit Emulation Service over ATM. Some have taken ATM cells and packetized the payloads into IP packets. Others have put the TDM channels into the IP payload directly, thus eliminating the segmentation and reassembly step of ATM. CEoP has parallels to ATM AAL1 transmission. Both are constant bit rate, both transmit uncompressed voice, and both support structured and unstructured circuit transport.
One issue with convergence of circuit switched services on a packet-based infrastructure is synchronization. In the circuit switched network, timing and synchronization is inherent in the design of the technology. However, when this service is carried over a PSN, the timing element is lost and has to be transported by other means. With CEoP technology, all associated timing and signaling information is “tunneled” alongside TDM voice, video and data traffic across the packet network.
Standardization work in well underway on CEoP technology. The ITU’s recommendation ITU Y.1413 deals with TDM over MPLS networks. The MPLS and Frame Relay Alliance and the MEF Forum have recently released their implementation agreements for Circuit Emulation Services over MPLS and Metro Ethernet network. The IETF is also actively working on TDM-over-Packet standardization.
Comparing CE-over-Packet with VoIP
CEoP is not Voice over IP. Both are meant for different services and do not compete in the network. There are many similarities and differences between the two, as seen in Table 1.
|Network Transport||Over IP, MPLS or Ethernet||Over IP Only|
|Network characteristics to carry services||Runs over managed network||Runs over Internet|
|Traffic types supported||Voice, video and data||Voice only|
|Services supported||Trunking (T1/E1/DNs)||Typically one channel per connection|
|Call control and Signaling||No need for gateway signaling, as it is tunneled through the trunk||Complex signaling between PSTN and IP, eg. H.323 or SIP|
|Timing across network||Synchronization is critical||Less critical on synchronization|
|Processing Latency||Low latency||High latency due to packetization and compression algorithms|
|Voice processing requirements||No compression required, VEC depending on the service and packet delay variation||Compression or VEC used|
|Industry Standards||ITU recommendation released, with others to come from the IETF, MEF and MPLS Forum||Well defined by IETF, RTP, UDP, IP|
CEoP versus VoIP
As seen in the table, both services offer different strengths and weaknesses. VoIP is best suited to transporting one voice call over the PSN. CEoP can be used for different services, not just voice, and can transport multiple circuits. It’s clear that CEoP service is attractive for the access side of the network for transporting multiple services across a PSN.
Potential opportunities for CE-over-Packet
Now let’s turn to the opportunities for this technology. There are many applications in different networks, as the technology can run over IP, MPLS or Ethernet networks.
One main benefit of the technology is its ability to transport many TDM trunks from one location to another over a packet network. The technology is initially being adopted in the access part of the network, primarily to transfer multiple trunks across a medium such as a gigabit fiber, or a fixed-radio access network.
The simplest application for CEoP technology is to provide packet interconnection for existing TDM equipment, enabling legacy traffic to move across an IP/MPLS/Ethernet network. By doing this, TDM vendors can extend the lifespan of their equipment, and generate years of income from older products. Evolving installed equipment to support new services, rather than a forklift upgrade, is something carriers are asking for now.
CEoP providing packet interconnection for existing TDM infrastructure
There are many new potential applications as well, focused on moving TDM circuits across various PSNs.
For example, DLCs (digital loop carriers) can benefit from one specific feature of CEoP. One specification DLCs must meet is a voice channel round trip delay of less than 1 ms. The technology does not need voice processing, which means that any delay comes down to three parameters: inherent delay of the semiconductor circuit; packetization delay; and network delay. DLCs are migrating to Ethernet over dedicated fiber, meaning network delay is kept very low. With a packetization scheme that provides for low delay and low inherent delay in the product, CES-over-Packet technology performs very well in providing legacy traffic support in a DLC application.
Similar applications are in the Multi Dwelling Units or Multi Tenant Units market, often referred to as the MxU market. The function of an MxU is to transport TDM traffic and Ethernet traffic over a common link back to a central location, such as a central office. By utilizing CEoP technology, the service provider can transport both voice and data across a packet switched link or network. The link or network can be Ethernet, IP, or MPLS.
Today’s wireless backhaul market relies on TDM service being carried across microwave radio links and equipment that is far more expensive than an IP network. There is a growing movement to use a packet network for backhaul, and CES-over-Packet technology to carry the TDM traffic. This would reduce the cost of base band processing and provide more cost-effective service.
Newer wireless technology, such as Wi-Fi and WIMAX, can also use CEoP technology to transport T1/E1 trunks across a campus. This marriage of technology can provide a low-cost solution for transporting TDM services in a limited distance situation, while WIMAX can be used for longer distances.
CEoP in a WIMAX and Wi-Fi network
Other applications are in fiber-to-the-premises applications. An EPON (Ethernet passive optical network) system is based on Ethernet as the backbone. It is a natural to cement Ethernet and CES-over-Packet technologies, and carry legacy traffic efficiently across the passive optical network.
Network applications for CEoP
CEoP technology is being used to carry SONET/SDH traffic across an IP or Ethernet network, such as a metro Ethernet network or a Resilient Packet Ring. The SONET/SDH sub rates are packetized and carried across the packet network. Thus a VT1.5 can be carried across the packet network and be reintegrated into a SONET/SDH ring.
We can see there are many applications for this technology in existing and new networks. Despite the best wishes of some, legacy TDM traffic will not disappear overnight, meaning carriers must find innovative and cost-effective ways to carry this traffic over the PSN.
CEoP is one viable approach. It simpler than VoIP, and can be complementary where VoIP technology is too complex or expensive to utilize. Connecting TDM equipment seamlessly across a packet network, CEoP delivers quality of service indistinguishable for the end-user.
Bruce Ernhofer is Product Manager, Packet Processors with proto Semiconductor.
|Copyright 2017, Proto Tech Inc. All Rights Reserved.|