Search

Our paper focuses on the Interface Message Processor (IMP), an important device in the history of ARPANET. Designed as the interface between ARPANET nodes and the common carrier telephone system, t...

People from Bolt Beranek and Newman and others have extensively documented the Arpanet technology, including the Arpanet Interface Message Processor (IMP). This paper sketches the history (not the previously described technology) of the IMP program as originally written in 1969 for the modified Honeywell 516 computer. A sequence of other systems, evolving from the original software system and running on a variety of hardware platforms, are also enumerated. In 2013 a faded 1973 line printer listing of the IMP program was run through a special OCR program optimized to process such historical artifacts; an assembler was recreated to assemble the IMP code (looking like the modified PDP-1 Midas assembler used in 1973); and a software emulator of the original IMP hardware platform was created. This article also describes the methods used to recover a digital copy and assemble and run again the 1973 IMP code.

Today's network routers are choking on streaming video and peer-to- peer file sharing. The article looks at a new router design that keeps track of individual data flows and promises a fix.

We describe an optical interface message processor that explores the high-bandwidth, parallelism, and high information density offered by optics. The most time consuming operations in communication networks, such as switching and routing, are performed in optical domain in the proposed system. The design does not suffer from the optical/electrical conversion bottlenecks and can perform self-routing in the range of Gbits/s.

The history of the NFE (network front-end) processor, currently best known as a TOE (TCP offload engine), extends all the way back to the Arpanet IMP (interface message processor) and possibly before. The notion is beguilingly simple: partition the work of executing communications protocols from the work of executing the "applications" that require the services of those protocols. That way, the applications and the network machinery can achieve maximum performance and efficiency, possibly taking advantage of special hardware performance assistance. While this looks utterly compelling on the whiteboard, architectural and implementation realities intrude, often with considerable force.

Bu çalışmanın amacı, Dağıtık Gerçek Zamanlı ve Çok Görevli (Distributed Real-Time, Multl Tasking) İşletim sistemlerine sahip çok İşlemcili sistemlerde, mlkroişlemciler arasında yüksek hızda ve güvenilirlikte iletişim yapabilen bir Arabirim Mesaj îşlemclslnlnin (Interface Message Processor.IMP) gerçekleştirilmesidir. Gerçekleştirilen Arabirim Mesaj İşlemcisi anabilglsayardan (Host. Computer) bağımsız olmakla birlikte gerekli performans ölçümlerinin yapılabilmesi amacıyla IBM PC/XT/ AT veya bunlara uyumlu mikrobilgisayarlar arasında İletişim yapabilecek bir Yerel Ağ İletişim Kaxh (Expansion Local Network Card) niteliğindedir. Anabilgisayarlar arası iletişim; halka (ring) topolojisinde, paralel olarak ve paket anahtarlamalt veri iletimi yapacak şekilde tasarlanmıştır. ABSTRACT The purpose of this study is to develop an Interface Message Processor, (IMP) which provides reliable and high speed communications between microprocessors in multiprocessor systems which have Distributed Real-Time, Multi Tasking operating systems. Although the designed MP is independent of Host Computer in principle, the implementation is done as an Expansion Local Network Card which can provide communications between IBM PC/XT/ AT or compatible microcomputers. The communications between host computers uses a ring topology for parallel and packet switched data transfer. ABSTRACT Bu çalışmanın amacı, Dağıtık Gerçek Zamanlı ve Çok Görevli (Distributed Real-Time, Multl Tasking) İşletim sistemlerine sahip çok İşlemcili sistemlerde, mlkroişlemciler arasında yüksek hızda ve güvenilirlikte iletişim yapabilen bir Arabirim Mesaj îşlemclslnlnin (Interface Message Processor.IMP) gerçekleştirilmesidir. Gerçekleştirilen Arabirim Mesaj İşlemcisi anabilglsayardan (Host. Computer) bağımsız olmakla birlikte gerekli performans ölçümlerinin yapılabilmesi amacıyla IBM PC/XT/ AT veya bunlara uyumlu mikrobilgisayarlar arasında İletişim yapabilecek bir Yerel Ağ İletişim Kaxh (Expansion Local Network Card) niteliğindedir. Anabilgisayarlar arası iletişim; halka (ring) topolojisinde, paralel olarak ve paket anahtarlamalt veri iletimi yapacak şekilde tasarlanmıştır. ABSTRACT The purpose of this study is to develop an Interface Message Processor, (IMP) which provides reliable and high speed communications between microprocessors in multiprocessor systems which have Distributed Real-Time, Multi Tasking operating systems. Although the designed MP is independent of Host Computer in principle, the implementation is done as an Expansion Local Network Card which can provide communications between IBM PC/XT/ AT or compatible microcomputers. The communications between host computers uses a ring topology for parallel and packet switched data transfer. 74

The rapid growth in local area networks (LANs) over the past few years has generated interest in the interconnection of LANs. The need of interconnecting LANs arises because it is advantageous to allow users in a much wider geographic span to communicate and to share resources. Furthermore, when a communication system with a large number of users in it, it is expected that the performance of the system will be better than the case that all these large number of users are directly connected if the users are firstly grouped into clusters which are interconnected. This paper describes a packet switching architecture which interconnects sections of local area networks operated in ALOHA mode. The properties that this packet switching node possesses are similar to those of an interface message processor (IMP), also known as a node, in a computer communication network. This node has the functionality of selectively connecting Users in different sections of slotted ALOHA channels. Detailed throughput and...

In this paper, we discuss flow control in a resourcesharing computer network. The resources consist of a set of inhomogeneous computers called hosts that are geographically distributed and are interconnected by a store-and-forward communications subnet. In the communication process, messages pass between hosts via the subnet. A protocol is used to control the flow of messages in such a way as to efficiently utilize the subnet and the host resources. In this paper, we examine in some detail the nature of the flow control required in the subnet and its relation to the host flow control and subnet performance.

Computer networks are creating a serious problem: increasing numbers of users can access a variety of different information systems. Unlike a set of application programs residing in a single computer, these different information systems have very little in common. They were built for different communities of users, by differing groups of programmers, to run on different hardware configurations, and with very different protocols for man/machine interaction. In addition, the network itself may not be transparent to the user; it may require login procedures and access protocols. Networks also can cascade the number of serial components being used to accomplish a task; using the ARPANET as an example, one might interact with a remote system via a local host computer, the Interface Message Processor (IMP) which connects it to the network, possibly several intervening IMPs, the remote host computer's IMP, and the remote system's host computer. A failure in one of these components creates the need for rather complex strategies for gracefully 'backing out' of the other systems, preferably leaving some intermediate results intact so that a later 'restart' procedure can be initiated.

This U/A report describes the status of UH/ALOHA participation in the ATS-1 Computer Communications Experiment through January, 1974. This experment was initiated in January, 1972, by the Spacecraft Data Systems Branch of the NASA Ames Research Center, to demonstrate the feasibility of utilizing satellite communication links to rpovide computer-computer and terminal-computer communications between remotely located sites. In order that the experiment be conducted under realistic conditions, computing facilities at the University of Hawaii and the University of Alaska were connected ot the Advanced Research Projects Agency computer network, via an ATS-1 VHF link, through the Terminal Interface Message Processor located at the Ames Research Center. The ATS-1 VHF transponder was utilized as a broadcast repeater for the three nodes mentioned above, with the satellite network operating in the ALOHA random-access burst mode.

The ARPANET was originally conceived to support high-data-rate distant communication between large mainframe computers. Because of the hardware limitations of the Honeywell 316/516 processor, which was selected as the original ARPANET interface message processor (IMP), most ARPANET IMPs are resticted to supporting a maximum of four host attachments (computers or gateways to other networks). As the ARPANET community expanded many sites experienced the need for more than four host ports. The response was to install additional IMPS. With the H316/516 no longer available and the C/30 IMP, being developed by the BBN Computer Company not expected to arrive until 1981. no more ARPANET community, hosts attached to IMPs have increasingly tended toward less powerful types, such as the DEC PDP-11 minicomputer. Many of these microcomputer hosts require only moderate data-rate network communication or occasional access. Newer node-switching equipment utilizes the Bolt Beranek and Newman Pluribus, which eliminates hardware constraints upon the number of host attachments. Nevertheless, the large majority of deployed IMPs are still of the older Honeywell type. The desire to attach a large number of minicomputers or microcomputers to the ARPANET induced DARPA to assign SRI the task of developing the port expander concept into a working product. This manual describes the functions, installations, and operation of the port expander.

This report describes the use of a terminal connected to a Terminal IMP (TIP) in the ARPA Network. The report assumes that the user knows how to operate a server Host system somewhere on the network once he becomes connected to that system, and the report defines the procedures and options the user has available to establish that connection. The ARPA Network, IMPs and TIPs, hardware maintenance, TIP operation, and formats and protocols are not described here. The bibliography (Appendix C) lists the relevant documents. Where possible, all types of these devices have been operated with direct connections to the TIP and also over a 103A dial-up modem. We have also briefly operated the TIP or heard of the TIP being operated with a variety of other types of terminals. These are listed in Appendix D. For your own safety, before you purchase any terminal listed in Appendix D or any other terminal for use with the TIP, you should check with BBN and try it with a TIP. One TIP is configured with a magnetic tape drive which is used as discussed in Section 8. It should be noted that the TIP is designed to support interactive, asynchronous terminals. It cannot normally support synchronous devices, or devices whose input (to the TIP) characteristics are significantly different from a human typist.

The ARPANET was originally conceived to support high-data-rate distant communication between large mainframe computers. Because of the hardware limitations of the Honeywell 316/516 processor, which was selected as the original ARPANET interface processor (IMP), most ARPANET IMPs are restricted to supporting a maximum of four host attachments (computers or gateways to other networks). As the ARPANET community expanded many sites experienced the need for more than four host ports. The response was to install additional IMPs. With the H316/516 no longer available and the C/30 IMP, being developed by the BBN computer Company not expected to arrive until 1981, no more ARPANET nodes/host ports were available. Simultaneously with this growth in the ARPANET community, hosts attached to IMPs have increasingly tended toward less powerful types, such as the DEC PDP-11 minicomputer. Many of these minicomputer hosts require only moderate data-rate network communication or occasional access. The desire to attach a large number of minicomputers or microcomputers to the ARPANET induced DARPA to assign SRI the task of developing the port expander concept into a working product. This manual describes the functions, installation, and operation of the port expander.

The ARPA Network will provide store-and-forward communication paths between a set of computer centers distributed across the continental United States. The message handling tasks at each node in the network are performed by a special purpose Interface Message Processor (IMP) located at each computer center. The centers will be interconnected through the IMPs by fully duplex telephone lines, of typically 50 kilobit/sec capacity.

The document contains information for connecting terminals and modems to the terminal interface message processor

The basic function of the IMP computer network is to allow large existing time-shared (Host) computers with different system configurations to communicate with each other. Each IMP (Interface Message Processor) computer accepts messages for its Host from other Host computers and transmits messages from its Host to other Hosts. Since there will not always be a direct link between two Hosts that wish to communicate, individual IMPs will, from time to time, perform the function of transferring a message between Hosts that are not directly connected. This then leads to the two basic IMP configurations -- interfacing between Host computers and acting as a message switcher in the IMP network. The message switching is performed as a store and forward operation. Each IMP adapts its message routine to the condition of those portions of the IMP network to which it is connected. IMPs regularly measure network performance and report in special messages to the network measurement center. Provision of a tracing capability permits the net operation to be studied comprehensively. An automatic trouble reporting capability detects a variety of network difficulties and reports them to an interested Host.

For many years, small groups of computers have been interconnected in various ways. Only recently, however, has the interaction of computers and communications become an important topic in its own right. In 1968, after considerable preliminary investigation and discussion, the Advanced Research Projects Agency of the Department of Defense (ARPA) embarked on the implementation of a new kind of nationwide computer interconnection known as the ARPA Network. This network will initially interconnect many dissimilar computers at ten ARPA-supported research centers with 50-kilobit common-carrier circuits. The network may be extended to include many other locations and circuits of higher bandwidth.