Paper Review #2: A Protocol for Packet Network Intercommunication

From: Yuhan Cai (yuhancai@cs.washington.edu)
Date: Tue Oct 05 2004 - 23:25:50 PDT

  • Next message: Tyler Robison: "Review of "A Protocol for Packet Network Intercommunication""

    Title: A Protocol for Packet Network Intercommunication
    Author: Vinton G. Cerf and Robert E. Kahn
    Reviewed by: Yuhan Cai
     

    Main results of the paper:

    , It introduces a generic protocol for information transfer and sharing among different types of packet-switching networks.

    , It provides a mechanism to resolve the issues of addressing, buffering, sequencing, formatting, flow control, and so on.

     

    Strengths of the paper:

    , It introduces a brand new notion, namely gateway, which connects one network with another. The gateway is advantageous in that it provides an effective interface among different types of networks, on which all of switching, routing, fault detection, fragmentation and formatting can be done.

    , It presents a rather complete overview of a new protocol that accommodates almost all of the problems faced by designers in interconnecting existing networks.

     

    Key limitations:

    , The details of the algorithms presented are relatively weak. For example, for retransmission and duplicate detection, the dealing is rather loose.

    , It is too much on the design level, as nearly no implementation or performance perspectives are addressed. Whether the protocol is efficient or even whether it is implementable at all is still questionable.

     

    Relevance of the paper:

    , As the pioneering work in internetworking, it is relevant in that, it not only introduces a powerful and flexible protocol for internetworking of different networks, but also provides fellow researchers a step stone for developing more sophisticated protocols.

     

    Future work:

    , A detailed specification of the protocol is in place.

    , The protocol presented only deals with networks that are packet switching. Other types of networks (such as circuit switching) might be worth further research.


  • Next message: Tyler Robison: "Review of "A Protocol for Packet Network Intercommunication""

    This archive was generated by hypermail 2.1.6 : Tue Oct 05 2004 - 23:25:55 PDT