6

Communication

This chapter is concerned with computer to computer communication and contains information on protocols, communication options, and network interconnects.

6.1

General Introduction

There are two parts to EDI, the message standards and the communication options of transmitting those standards. This section provides an overview of the communication options available to a company planning to implement EDI. Its purpose is to highlight the areas in which key data communication design decisions must be made. There is no single or preferred solution in this decision making process. Each company must determine the proper approach based on current and projected transaction volume and level of investment.

When a company plans to implement EDI, consideration should be given to the form the data will take as they are communicated to the trading partners via physical or telecommunication means. Physical means include the use of magnetic tape or courier service. Data communication employs the use of a public or private telecommunications means to send a trading partner EDI data.

When data communication is the means selected, all parties need to identify, negotiate, and select the communication transfer mode. Issues to be addressed include the sending and receiving location, protocol and transmission parameters. Both long and short term EDI communications strategies should be analyzed to determine the current and future business and/or industry environment with respect to their EDI trading relationships. Criteria to be considered when determining the communication mode of data transfer include the following:

• Distance of transport
• Number of destinations
• Costs required delivery time frame
• Frequency of transport
• Security
• Volume of transactions
• Comparability of media
• Reliability

The computer to computer exchange of EDI data can be accomplished via the use of U.S. Mail, courier service, or via dedicated or switched networks. Each exchange method should be analyzed to determine if the approach meets the company and trading partners communication needs.

No matter which approach is selected, a contingency plan should be formulated to address the possible event of a communication failure.

Issues to be considered are backup procedures to address system failures, transmission error recovery including the establishment of a maximum number of retransmit attempts, security, network response time, and error reporting.

ANSI Publications:

• "An Introduction to Electronic Data Interchange" - ASC X12D/8702 dated July 1987
• "Transfer of ANSI Xl 2 Data Using Asynchronous or Binary Synchronous Communication Protocol" - ASC X12/88035 May 1987 Guideline
• EDI Asynchronous Guideline/ UNSM Purchase Order Message ASC X12S/88438 July 1989 Draft Publication

An organization's own specific EDI transmission requirements, including current and future systems, will determine the appropriate mix of network services needed. A communication survey of the proposed trading partners should be conducted in the early stages of planning to determine the software/hardware environment.

Questions that need to be answered:

• What type of computer will be used to process your EDI messages?
• What operating system software is your computer running?
• What type of communication equipment will be used to send and receive your EDI messages?

6.1.1 - Data Communication Function

The ASC Xl 2 publication ASC Xl 2/88035 May 1987, "Guideline: Transfer of ANSI X12 Data Using Asynchronous or Binary Synchronous Communication Protocol" provides guidance on the technical procedures needed to insure accurate data transfers between EDI trading partners. The guideline provides a definition of the data communication function and discusses the critical elements of data communication planning which include:

• Identification of intended communication partners
• Establishment of authority to communicate
• Agreement of privacy mechanisms
• Authentication of communication partners
• Determination of the adequacy of resources
• Determination of acceptable quality of service
• Synchronization of cooperating applications
• Selection of dialogue description
• Agreement of responsibility for error recovery
• Agreement of procedures for control of data integrity
• Identification of constraints on data syntax/structure

The guideline also references basic requirements concerning the trading partner's EDI communication system. These requirements are:

• Sufficient data storage for the retention of transmitted and received EDI data, trading partner authentication information, and status of data transfers
• Ability to provide the current status of data transfers
• Reliable storage devices providing data backup and restoration capability

6.2

Protocols

Protocol is the collective term for the rules that govern communication at a given layer in a network architecture. They are important because in each layer, in any communication system, a set of rules must be agreed to and followed in order for communication to be successful. If communications are to be standard for an industry, then the protocols should be specified in this section. One way to accomplish this is to identify specific communication software to be used: for example, IBM 3780 Data Communication Terminal Order No. GA2730633. Another method would be to specify a communication standard, such as the International Telegraph and Telephone Consultative Committee (CCITT) Recommendation X.25.

6.2.1 - EDI Data Transfer Protocol

The ASC Xl 2 publication ASC Xl 2/88035 May 1987 "Guideline: Transfer of ANSI Xl 2 Data Using Asynchronous or Binary Synchronous Communication Protocol" defines the basic requirements for EDI data transfer procedures. The guideline advises that the data communication requirements to be addressed by EDI trading partners are: communication compatibility, security, and data integrity. A brief explanation follows:

Communication Compatibility refers to:

Electrical signaling

• signaling between communication hardware, modem communication, and channel modem
• modem types
• transmission speed compatibility

Line Control Protocol

• between communications software, such as asynchronous, binary synchronous, and other protocols
• call establishment
• data blocking and organization
• acknowledgement and signaling for handshaking and error control
• line turnaround procedures
• character synchronization
• escape interrupt disconnect

Data Transfer Protocol

• compatible EDI data transfer programs and techniques for managing data transfer

Security issues to be considered:

• EDI data going only to the intended trading partner
• control over access to communications and business systems
• identification and authorization of trading partners
• authentication

Data Transfer Process Integrity refers to the actions taken to prevent problems. An objective of the EDI communication system is to provide a high degree of data process integrity in order to:

• Minimize potential loss of data by providing intermediate safe storage, interchange authorization, retransmission approval, mutual results commitment
• Minimize the potential for data duplication by providing temporary data suspension
• Minimize the potential for situations that require human intervention by providing status retention and transfer restart capabilities

Error detecting protocols should be considered as the minimum communication requirement for EDI. Asynchronous and binary synchronous communication protocols provide error handling techniques based on the specific implementation and character set used.

Data link protocols fall into two major categories, character oriented and bit oriented. A character-oriented protocol uses a particular code set for transmission with some of the characters in the code set reserved for control functions. Asynchronous and binary synchronous protocols are examples of this protocol.

Asynchronous transmission is defined as a transmission that is synchronized by the sending and receiving Data Transmission Equipment (DTE) before each character is sent. Each character has a start and stop bit to indicate beginning and end of each/character. The start and stop bits are the mechanism by which synchronization is established. Typical asynchronous communications accommodated by microcomputers transmitted at a baud rate ranging between 300-4800 bits per second (bps). See ASC X12S/89438 July 1989 draft publication: "EDI Asychronous Guideline / UNSM Purchase Order Message" Asynchronous transmission is defined as a transmission that is synchronized by the sending and receiving Data Transmission Equipment (DTE) before each character is sent. Each character has a start and stop bit to indicate beginning and end of each/character. The start and stop bits are the mechanism by which synchronization is established. Typical asynchronous communications accommodated by microcomputers transmitted at a baud rate ranging between 300-4800 bits per second (bps). See ASC X12S/89438 July 1989 draft publication: "EDI Asychronous Guideline / UNSM Purchase Order Message"

Binary-synchronous transmission is different from asynchronous in that bit synchronization is established for a much longer duration, usually for the time it takes to transmit several thousand bits. This results in less transmission overhead but requires more complex circuitry. Typical binary synchronous communication is supported by a mainframe or communications front end processor transmitted at a baud rate ranging between 2400-9600 bps. See ASC Xl 2/88035 May 1987 guideline: "Transfer of ANSI Xl 2 Data Using Asynchronous or Binary Synchronous Communication Protocol". Binary-synchronous transmission is different from asynchronous in that bit synchronization is established for a much longer duration, usually for the time it takes to transmit several thousand bits. This results in less transmission overhead but requires more complex circuitry. Typical binary synchronous communication is supported by a mainframe or communications front end processor transmitted at a baud rate ranging between 2400-9600 bps. See ASC Xl 2/88035 May 1987 guideline: "Transfer of ANSI Xl 2 Data Using Asynchronous or Binary Synchronous Communication Protocol".

A bit-oriented protocol is independent of any particular code set, and no character codes are reserved for control functions. High Level Data Link control (HDLC) and Advanced Data Communication Control Procedures (ADCCP) are examples of this protocol. The major advantage is in speed and standardization.

6.3

Point-to-Point (Dedicated Network)

Point-to-point circuits connect users in dedicated networks. These dedicated facilities are used when the need to transmit EDI transactions between the two ends is steady and continuous. Since the users are paying for the link regardless of the number of transmissions, its usage must be high for it to be economical. This approach is not recommended by the UIG.

6.4

Third Party Services

Selection of this alternative requires an in-house staff capable of managing the network. Items that must be addressed include:

• Network availability
• Service level
• Network response time
• Network communications
• Communication speed
• Transmission mode
• Modem capability
• Line protocol

Switched networks connect and disconnect circuits as required to transmit data. The three common methods are:

Circuit switching is used in the public telephone systems. A circuit is dedicated between the source and destination for the duration of the transmission. The sender and receiver must be available at the same time.

Message switch networks package the data in messages and pass the messages from switch to switch. The sender and receiver do not have to be available at the same time, since the message is stored at each intermediate step. For this reason, message switched networks are also referred to as store and forward networks.

Packet switching is similar to message switching, but it divides the data into smaller, equal-sized pieces called packets. The time it takes to move data through the network is less, since large messages don't have to be stored at each intermediate switch. The reduced delay, over message switching, allows the two users to carry on a dialogue, referred to as an interactive process. In addition, the reduced delay aids transaction processing by moving the transactions to their destination quickly.

Packet switching's advantage over circuit switching is in making efficient use of the data lines. Each packet carries a destination address, so packets from multiple sources heading to different destinations can be transmitted down the same dataline if it is desirable.

The above facilities and services may be obtained from commercial networks (Value Added Networks) rather than developed in-house. The commercial networks provide the network management and knowledgeable staff to support your communication requirements. Commercial networks now offer more than moving data from one site to another. Services provided include mailbox service, data storage, speed and format conversion, or translation.

Your communication survey will help you decide on the services you need. If a commercial network alternative is selected, the network will provide you with the technical instructions required to access the network.

Typically, not all companies have the communication facilities to accommodate the multiple communication protocols in use by their potential trading partners. Third party service providers eliminate the need for a company to invest heavily in communication hardware, software, and personnel. A third party service provider allows the convenience of a single data transfer link to multiple trading partners independent of operating schedules, protocol conversion, hardware interface, and conversion requirements.

When selecting a third party service provider, a company should evaluate the service capabilities and performance offered. Issues to consider:

• speed of deliver mode
• outdial capabilities (i.e. autodial, scheduled)
• data integrity
• reliability
• job queueing options
• interconnect capabilities
• tracking and control reporting (audit, historical, and exception reporting)

Before data transfer begins with a third party service, communications should be mutually defined and agreed upon. The use of third party communications should be transparent to trading partners.

Usually each trading partner pays for its own EDI services received. Costs associated with the use of a third party service include:

• startup charges
• monthly mailbox fee
• connect charge
• data storage
• network interconnect
• character charges
• reports

6.5

Network Interconnects

When multiple commercial networks are needed to satisfy EDI requirements, an additional level of complexity is added to your communication solution. It is important that each network representative be included in your planning process. Agreement must be reached on the items listed in Section 6.4. Special requirements, such as billing and overall network management, should be addressed.

Most EDI third party service providers offer a network interconnect capability. This capability allows one third party network to communicate with another. The connection is transparent to the EDI trading partners. Network interconnects should be defined by all parties involved, i.e., both trading partners or network providers.

The interconnect service eliminates the need for a company to have multiple mailboxes on multiple third party service networks. There may be an additional charge for the service, and this must be weighed against the cost of establishing additional mailboxes on other networks.