Emlac is an AT&T Value Added Applications Provider (VAAP). "As a Vaap member,Emlac provides the best data network application product... money can buy."

Why AT&T ?

To help our clients achieve success, we need to clearly present the value that AT&T brings to the table. AT&T is committed to providing clients with value that performs, incorporating reliability and innovation into everything we do. The total value package that AT&T delivers to our clients is unmatched in the industry.

RELIABILITY

 AT&T manages the largest communications network in the world: 2.75 billion circuit miles of transmission facilities; 310 network control points; 136 4ESS switches with real-time network routing.

AT&T has 2.75 billion circuit miles of transmission facilities or three times as much as our nearest competitor. This means that AT&T is able to deliver high-quality services worldwide on AT&T facilities. Additionally, AT&T owns 25% of the world's cable capacity, far more than our competitors, with three international gateways to Mexico, seven to Canada, and six to the World. Our nearest competitor has three gateways to Canada, three to Mexico, and three to the World.

AT&T has 310 network control points in our network or almost 50% more than our nearest competitor. The network control points are dual processors that house client-specific information for advanced Software Defined Network Services. They are geographically separated for back-up purposes. They have mated pairs and work in conjunction with AT&T's #4 Electronic Switching System^TM or 4ESS^TM to provide the best redundancy and protection provided by any telecommunications carrier in the event of a disaster.

The heart of the AT&T network is the 4ESS. AT&T currently has 136 of them which are capable of handling up to one million calls per hour as part of their responsibilities in routing calls through the network. The way AT&T routes calls from one location to another, called real-time network routing, is unique to AT&T. It is a sophisticated call-routing system built into each 4ESS switch and allows each switch to check the trunk capacity of every other 4ESS in the network to find the most efficient path for any given call. That's one reason why AT&T completes 999 out of every 1000 calls on the first try.

 AT&T's digital network with 1,417,600 fiber miles has more route diversity, redundancy and digital back-up than any communications provider.

According to the annual FCC Fiber Report for year-end 1995, AT&T's digital network has more route diversity, redundancy and digital back-up because it has more fiber than any competitor - nearly three times as much as our nearest competitor.

AT&T has over 1,000 loops built into its network and owns over 99% of the facilities it uses in that network. Our nearest competitor's fiber map shows fewer loops and a dependence on other carriers.

The AT&T Worldwide Intelligent Network is the largest, most sophisticated and most reliable communications network in the world. AT&T has long been recognized by its clients as the leader in reliable networking and our network structure and configuration are uniquely designed for maximum reliability. Many reliability features are built into the network, automatically providing clients with fundamental protection. Also, clients have the option of purchasing additional enhancements to further customize the reliability of the network.

Fast Automatic Restoration (FASTAR) can move millions of signals from one cable to another in seconds and is the foundation of the AT&T Digital Access Cross Connect Systems (DACS) III. Using DACS III's deployed throughout the AT&T network and a software algorithm that calculates alternate paths throughout the network, FASTAR is able to reroute traffic around facilities in minutes. In the event of a fiber optic cable cut in the core network, FASTAR will automatically locate the exact site of the cut and transfer the affected circuits to spare capacity going around the cut. Approximately 93% of AT&T's fiber T3 miles are FASTAR restorable. Since FASTAR was introduced in 1992, it and other efforts have reduced circuit outage minutes by 92% and data networks have become significantly more reliable.

AT&T is currently deploying FASTAR II. The new network will consist of more than 50 double-interconnected SONET rings with ATM switching at crossover points to create a ring and mesh architecture for data services. Using overlapping, self-healing rings, FASTAR II will restore certain types of network failures, such as simple cable cuts, in milliseconds. With this type of outage, clients shouldn't even notice that traffic was interrupted.

Alternate Signaling Transport Network (ASTN) is also an AT&T exclusive. It is a powerful and reliable feature which supports all AT&T clients. It is a second, separate signaling network that serves as a back-up to the primary network. ASTN is the proprietary signaling network that backs up the signaling links between 4ESS switches and the Signaling Transfer Points (STP) or the brains of the network. It gives AT&T two distinct signaling networks. In the event of a problem, the ASTN can take over and continue processing calls.

The Enhanced Diversity Routing Option (EDRO) is an optional feature that provides electrically and physically diverse interoffice channels for AT&T's T1, T32, T3 and Fractional T3 Services, and takes advantage of AT&T's cable diversity and multiple serving offices within metropolitan areas. There is no warranty associated with EDRO. Formal management and documentation of diversity arrangements on a quarterly basis are available for ongoing support.

The T1-T3 Enhanced Reliability Option (ERO) is designed to deliver a guaranteed 99.99% availability for AT&T's T1 or T3 circuits - the highest availability guarantee in the industry on an end-to-end basis. It is backed by the best warranty in the industry. To accomplish this, all Enhanced Reliability Option circuits are 100% fiber routed and supported by AT&T's patented FASTAR restoration system. If one of these circuits suffers a cumulative outage of five minutes or more per month, clients receive a full months credit for that circuit.

 With FASTAR, AT&T has incurred 50% fewer network outages than our nearest competitor.

AT&T has had fewer outages than our competitors, thanks to FASTAR. Results of the ongoing, proprietary AT&T Labs testing program (RCIC-2) which monitors network outages that cause at least 50% call blockage for at least one hour, indicate AT&T had half as many network outages as our nearest competitor as of 3Q96.

To put FASTAR's performance of restoration capabilities in perspective, it took 10-12 hours using physical patching techniques and 2-4 hours using electronic cross-connect systems, to do what FASTAR does automatically in minutes. Operating at the T3 level, the system provides an integrated overlying layer of centralized and distributed intelligence to dynamically isolate failures and evaluate hundreds of potential alternate routes throughout the core network, not just one or two fixed alternatives per major span. On average, FASTAR restores 90+% of AT&T's T3 circuits within 10 minutes and approximately 93% of AT&T's fiber T3 miles are FASTAR restorable.

 AT&T is the market leader in frame relay. AT&T Frame Relay Service continues to win industry and user awards.

 Data Communications Magazine User Survey for the fourth year in a row gave top ratings to AT&T Frame Relay Services. (8/97)

 Internet at Work Magazine, in a poll of its editors and key industry analysts, selected AT&T Frame Relay Service as the best. (8/96)

 The 1996 Editor's Choice Award for Best Public Carrier for Frame Relay Service was awarded to AT&T by Internetwork Magazine. (9/96)

 AT&T is a recognized market leader in data communications, networking, and support as evidenced by the array of awards bestowed across multiple dimensions and offers.

 Our offers are backed by the most powerful guarantees in the industry.

SERVICE ASSURANCE WARRANTY

This warranty reinforces AT&T's commitment to service quality and reliability by crediting businesses if they experience an interruption on their domestic AT&T T1, Fractional T3, T32, T3 and OC-3 services. This includes interoffice channels purchased under Tariff 9 and T1, Fractional T3, and T3 local channels purchased from Tariff 11. AT&T will issue credit based on a credit schedule for any outages up to 100% of the monthly service cost. Discounts range from a 5% discount for outages of less than 1 hour to a 100% discount for outages greater than 9 hours. This warranty applies to the entire portion of the circuit that AT&T provides or bills and includes interoffice channels, local channels, central office connections, access coordination functions, T1-T3 Monitor, and AT&T Bandwidth Manager. Credits are applied to the next month's bill.

 

INNOVATION

 AT&T Labs developed ReliaBURST, the first ATM rate-based congestion management service, providing better application throughput.

As more and more companies move to frame relay and ATM, they are faced with growing throughput and congestion control challenges. AT&T is the first provider to offer ATM clients rate-based, closed-loop congestion management.

With ReliaBURST, client's applications have the capability to reliably burst traffic into the network at sustained rates higher than their subscribed rates on Class C service. While allowing extended bursts above the committed rate, ReliaBURST does not let such user applications affect other users' performance.

As a closed-loop bandwidth optimization system, ReliaBURST assures that no one PVC can "lock out" a client's traffic or degrade their application's performance such as in open-loop ATM systems when user applications can ignore congestion control. Spare network capacity is allocated in proportion to a client's CIR without jeopardy to any other client's traffic.

Clients can experience increased throughput, better response time, guaranteed minimum CIR, and potentially lower service costs through reduction in the size of PVCs.

AT&T is one of the very few closed-loop frame relay service providers in the world and the first closed-loop ATM provider. AT&T is well positioned to provide service interworking between frame and ATM that will enable clients to benefit from a seamless network.

 AT&T offers the most frame relay differentiators in the market.

 AT&T has over 520 access POPs for frame relay as compared to MCI and Sprint, our nearest competitors, who offer 375+ and 330 respectively.

  AT&T is the only major frame relay provider to provide both simplex and duplex PVCs.

 With ReliaBURST, AT&T is the first provider to deliver rate-based closed-loop congestion management; helping clients achieve higher average throughout without increased traffic loss risk and at no extra charge. Closed-loop congestion management uses spare PVC network capacity to relieve over-subscribed CIR demand. Both Sprint and MCI have an open-loop system where over-subscribed data designated as discard eligible is discarded to relieve CIR congestion.

 Pricing for AT&T Frame Relay Service is a simple flat rate based on port and PVC speeds. Many of the other providers, including Sprint and MCI, have a combination of fixed and usage rates.

 AT&T pioneered FASTAR which ensures network restoration in minutes.

AT&T Labs invented Fast Automatic Restoration or FASTAR and deployed it in 1992 as the primary system for automated restoration within the network. This technology automates the steps required for restoration from certain facility failures and significantly improves the reliability of network services.

According to measurements performed by AT&T Labs, the base AT&T core network infrastructure, protected by FASTAR, has dramatically reduced outages for AT&T clients up to 75% since its inception.

 Our technical specialists access innovative, leading-edge network design tools for the utmost in network optimization and integration.

AT&T's approach to network design is a combination of the expertise of professional technical designers and our internal proprietary design tools developed and supported by AT&T Labs.

G-INOS Data, the primary tool used, creates network models based upon topology or traffic/performance networks. The G-INOS system integrates topological design with response time analysis and enables the designer to easily compare network performance vs. cost. The system supports a wide range of protocols and design options.

If the network is global, G-INOS can now be used to design international Private Line and WorldSource Frame Relay networks.

Design tools like G-INOS, combined with the knowledge and expertise of AT&T network designers, provide AT&T clients with the most economical, customized, performance-based designs.

 

Why AT&T Summary

RELIABILITY

 AT&T manages the largest communications network in the world: 2.75 billion circuit miles of transmission facilities; 310 network control points; 136 4ESS switches with real-time network routing.

 AT&T's digital network with 1,417,600 fiber miles has more route diversity, redundancy and digital back-up than any communications provider.

 With FASTAR, AT&T has incurred 50% fewer network outages than our nearest competitor.

 AT&T is the market leader in frame relay.

 AT&T is a recognized market leader in data communications, networking, and support as evidenced by the array of awards bestowed across multiple dimensions and offers.

 Our offers are backed by the most powerful guarantees in the industry.

INNOVATION

 AT&T Labs developed ReliaBURST, the first ATM rate-based congestion management service, providing better application throughput.

 AT&T offers the most frame relay differentiators in the market.

 AT&T pioneered FASTAR which ensures network restoration in minutes.

  Our technical specialists access innovative, leading-edge network design tools for the utmost in network optimization and integration.

Why Buy Frame Relay from AT&T?

by Neal Goodson
Data Network Consultant

It's a good question with an easy answer if you're familiar with frame relay technology and AT&T's advantages. In case you are not, however, this job aid is a quick reference that will help you articulate why AT&T's Frame Relay Service is unparalleled in the industry.

Step One: Understand the desirable characteristics in a public frame relay service.

  Low Delay

Delay is measured from edge to edge of the frame relay cloud (the edges of the cloud are where the customer's access circuit connects to the frame relay service). Low delay, which is usually measured in milliseconds, is desired. It means that user data traverses the network quickly. Consistent delay is also desired. It means that user data transmissions through the network are predictable. Low consistent delay is most desired. This provides predictable high speed performance and is important when you have applications that are delay sensitive such as SNA applications. Customers with networks that have variations in delay will not get predictable and dependable performance. In the AT&T network, the service objective for maximum delay from ingress point to egress point is 75 milliseconds - the lowest delay objective in the industry. The AT&T Frame Relay network has an ATM backbone and the ATM technology employed in this backbone takes the offered data and sends it across the network in standard 53 byte ATM cells. Since each of the ATM cells is uniform, delay is predictable.

 High Throughput

All Frame Relay networks transmit data that may be discarded at the ingress, in the network, or at the egress of the network for various reasons. Data is discarded primarily because it has been damaged in transit or because it has been marked discard eligible (DE) as it enters the network and encounters congestion. Open loop network architectures will allow data to enter the network even though data may not be able to be transported across the network. In frame relay networks, all of the data that is above the committed information rate (CIR) is marked DE (consider a zero CIR). When congestion occurs in the network, all data marked DE is subject to being discarded. When this data is discarded in the network for any reason, it must be retransmitted by the CPE, thus the same user data requires two times the amount of network resources (or more depending on the number of occurrences). Because the network resource is reused for the same transmission, there is lower overall network throughput on an application basis and the potential for continued congestion in the network. In a closed loop environment, such as the AT&T Frame Relay network with Reliaburst, any user data in excess of CIR that is permitted to enter the network will have a very high probability of being delivered to the egress end of the network (unless it is damaged). This permits higher overall throughput.

 Predictable ( No "Freeze Out")

Freeze out occurs when the architecture of the frame relay service does not manage or control the excess bursts of data coming into the network. Many open loop congestion control algorithms use a shared buffer within the switch. As data arrives at the switch from different customer sites, the data is put into a shared queue on a first come first served basis. Given the wide mix of access speeds and the varying traffic at a given time, the response time and delay across the network can vary dramatically. In addition, consideration must be given to the behavior of various protocols and the impact of delay on these protocols. As an example, TCP/IP may attempt to send huge amounts of data in an effort to complete a large file transfer while SNA (in a terminal to host mode) may be sending small amounts of data expecting a timely response. These two may be contending for the same shared buffer. In an extreme situation, the SNA user may actually be unable to send data across the network for periods of time because the TCP/IP user has filled the shared buffer. The SNA user then experiences "freeze out", the inability to send data even at CIR, because the open loop congestion control network has no way of controlling how other users burst above their CIR. This lack of management control causes unpredictable response times. All architectures based on open loop have this trait. In a network where there are management controls, as in the AT&T Frame Relay network with Reliaburst, response times are more predictable because the network assures equitable sharing of the available excess bandwidth.

 Fairness

A feature of most frame relay networks is the ability of the user to "burst" above the committed information rate. Fairness issues are important when bursting is considered. Since frame relay networks utilize shared facilities and switches to transport data, there is a finite amount of excess bandwidth available for users to share when bursting occurs. AT&T guarantees the CIR of all PVCs and assures fair sharing of excess bandwidth with the Reliaburst feature in the network. The Reliaburst algorithm does several things. First, it is constantly aware of the available network resources with which to send data. Second, it knows the CIR of all PVCs in the network and their respective relationships on any given path in the network. Third, it knows (by constant sampling) how much excess bandwidth is available to share among PVCs. Fourth, it has the ability to apply the knowledge of available shareable bandwidth in conjunction with the relationships of the network users' CIRs. This means that everyone in the network gets to share available bandwidth (ability to burst above the CIR) in a fair and equitable manner and with no negative effect on any other user.

Step Two: Become familiar with the technology platforms utilized by various vendors of public frame relay services.

Frame Relay is an interface specification that connects users to a frame relay carrier's wide area network. As long as the customer's equipment and the network provider's equipment meet the same specification, they can interface with frame relay. The transport side of the frame relay interface affects the desirable characteristics of the frame relay service. Here we are concerned with the transport technology in the wide area network.

 Frame Relay with frame switched backbone.

Frame switching networks handle each entire frame (regardless of the frame size) from the user network one at time. This means the entire frame is read into the switch before any data is transmitted out of the switch. Once the frame is completely in the switch, it is "switched out" over the wide area network. This process is replicated at each switch the frame traverses in the provider frame relay network. The process of frame switching adds delay to the transmission of the end user data. This process also causes variable delay (non-uniform) from transmission event to transmission event because not all frames are the same size.

 Frame Relay with cell relay switched backbone.

Cell relay switching technologies divide user frames of data into uniform cells (53 byte ATM cells in the AT&T Frame Relay network) to be switched across the wide area network. Data can be sent from the ingress switch across the wide area network as soon as a cell is filled (no need to wait for the entire frame to be read into the switch). This means there is lower transmission delay for customer data. This technique is used in the AT&T Frame Relay network and is called "pipelining". Because the ATM cells are of uniform small size, there is low predictable delay across the network.

 Frame Relay with hybrid switching backbone.

Hybrid switching occurs in networks where multiple technologies (frame switching, cell switching, or other technologies) are implemented to provide an end-to-end frame relay service. Hybrid switching is no better than the worst link in the chain. Congestion control in these networks (either with one vendor or in a multivendor environment) is difficult to control.

Step Three: Understand the effects of different congestion control methodologies.

 Open Loop

Allows all users to transmit at full port speed regardless of network capacity. The sending switch has no knowledge of congestion in the network. All frames above CIR are marked as discard eligible and with zero CIR, all frames are marked DE. Vendors who use this method claim they will provision additional trunk capacity to overcome congestion. Can they provision additional capacity in time to solve your customer's problem? User retransmissions can exacerbate the congestion situation.

 Closed Loop

Network utilization is monitored and spare capacity is allocated fairly among currently active frame relay PVCs. This is done by components in the network passing messages to the source switch on available capacity. AT&T's Reliaburst provides constant feedback to the transmitting port. In this environment, customers always receive at least their CIR and, depending upon traffic flow and network utilization, they can burst above CIR up to the full port speed.

Step Four: Understand the AT&T Frame Relay advantages.

The AT&T Frame Relay network is managed by AT&T Labs which has many years of experience managing the largest networks in the world. AT&T's frame network is a seamless global ATM based network which provides:

 Low delay with predictable response time characteristics due to ATM technology and management controls. The domestic PVC network maximum delay objective is 75 ms port to port.

 Consistent contention/congestion management across the Frame Relay and ATM platforms. The closed loop flow control method (endorsed by the ATM forum) is utilized on both the ATM and Frame Relay networks.

 Industry leading Customer Network Management capabilities which are consistent across both Frame Relay and ATM services.

 Unparalleled reliability with AT&T's FASTAR protected core network and planned FASTAR II network.

 End-to-end T3 facilities with movement to SONET.

 Enhanced PVCs which assure users they will have their CIR and also a "fair share" of the available additional bandwidth in the network for bursting above the CIR.

 The most robust set of options for disaster recovery.