CSE/EE 461: Introduction to Computer-Communication Networks, Autumn 2007
  CSE Home   About Us   Search   Contact Info 
 
Course Home
 Home
Administation
 Overview
 Using course email
 Email archive
 Anonymous feedback
 View feedback
 Homework Turnin
 
Most Everything
 Schedule
 
Information
 UW/ACM Tutorials
    Homework 3
Out: Wednesday October 10
Due: Monday October 15
Turnin: Online; details TBA

  1. Suppose you were designing a protocol intended to be carry digital telephone calls. What approach to reliable transmission would you take? Briefly explain your reasons.

  2. Describe, in (just) enough detail to make clear it works, a sliding window protocol that uses selective ACKs. That is, instead of ACKing each arriving frame with the number of the first missing frame, another scheme (of your design) could be used in an attempt to reduce the number of frames that are retransmitted when an error occurs.

    For the purposes of this question, you can assume that transmitted frames arrive in the order sent, except for those that are lost. Also, you have no information about what the error rate might be. Your protocol should work correctly (reliably delivering all frames to the receiver-side application, in order, and reasonably promptly) using finite resources, no matter what the error rate is.

    You should think about frame formats when designing your protocol, and your description should include enough information that we understand how the sender and receiver will be able to decode any frames they receive.

    Also, in any reasonable design frame sequence numbers will be limited to a finite range. Your protocol needs to deal with that restriction.

  3. Just like Ethernet and 802.11 wireless, USB has a protocol specification. In contrast to those other protocols, USB addresses are handed out dynamically: when a device is connected, the host picks a currently unused 7-bit address and assigns it to the device. That 7-bit address is then used in all frames involving that device so long as it remains connected. (The next time it is connected it may get a different address.)

    Besides the fact that 7 bits is smaller than 48 bits, what advantage is there to this dynamic assignment scheme, relative to the 802.2 addresses used by Ethernet and 802.11?


Computer Science & Engineering
University of Washington
Box 352350
Seattle, WA  98195-2350
(206) 543-1695 voice, (206) 543-2969 FAX
[comments to zahorjan at cs.washington.edu]