# Sastry NEST meeting notes
# Date: 2/6/2004
# Present: Phoebus, Tanya, Songhwai, Bruno, Cory, Terence, Luca, Bonnie

- Discussed the sensor network characteristics that we wish to monitor
  and the pieces of hardware and software we need to prepare to get things done


330 Cory Testbed Diagnostics (& implementation in software and hardware)
****************************
- TimeStamp mote behaviors
   -> SW 	TimeSync
   -> HW	possibly timesync using packets over ethernet for 
		"ground truth"
- Routing Analysis
  -> SW	  log data to EEPROM for retrieval later
  -> HW	  send over ethernet (eMotes) back to main computer
  * path
    -> SW	keep track of nodes the packet hopped through
    -> HW	report over ethernet for each hop?
  * packet loss rate
  * time of transmission (ties back to time sync)
  * Energy per packet per hop
    ~ ask Rob and Naveen how they monitor this
  * % control packets and % data packets
  * routing protocol interaction with MAC
    -> SW   B-MAC, S-MAC, other TDMA schemes?
- Sensor precision
  * to find if any correlation between number of sensor reports and the
    accuracy of the prediction
  -> SW	     keep track of number of sensing before aggregation


- Immediate Tasks
  * set up Song's old computer as "basestation" running database for 
    reprogramming the motes and gathering data.
  * everyone in the group should familiarize themself with using TinyOS
    at least at the tutorial level
  * set up 8 motes for programming with Cory on Tuesday
  * see POST MEETING INFORMATION below for more tasks not discussed at meeting


- VJ (DARPA NEST program manager) is coming on February 23rd... we're 
  expected to have a semi-working demonstration of our work by then.

- Problem: Cory has an Extreme Scaling Conference Call on Fridays at this meeting time (1 PM - 2 PM) may need an alternative time.




Post-Meeting Information
************************
- Immediate Tasks
  * purchase information for physical construction of testbed (pipes, netting)
    ~ we will not be moving the testbed up from Intel anytime soon 
      (Kamin is using it)
    ~ general list of questions:
      1) where to buy
      2) sizes and thickness available
      3) can harware store precut for us
      4) what types of drill tips are necessary to drill it?
    ~ plastic pipes
    ~ plexiglass
    ~ plastic netting
    ~ pulleys (preferably non-magnetic)
  * purchase hardware equipment (not discussed in meeting)
    ~ Power Over Ethernet Switches (ask Albert Goto)
    ~ ethernet cables (ask Albert Goto)
    ~ headers for eMote to mica2Dot connection
      http://www.newark.com/NewarkWebCommerce/newark/en_US/support/catalog/productDetail.jsp?id=50F3093
      http://www.newark.com/NewarkWebCommerce/newark/en_US/support/catalog/productDetail.jsp?id=96F2734
    ~ soldering iron (MBT-250SDPT from PACE)
      https://webvia.techni-tool.com/VIA5/validateLogin.jsp?req=https://webvia.techni-tool.com/VIA5/viaListProducts.jsp&searchText=SO3-61255&reqTitle=TITLE_VIAUSERDOCLIST
      MBT-250SDPT is on page 2
    ~ (check availablity in lab first) solder wick, sponge, vise for holding circuit boards, tweezers, flux, squirt bottle (for holding water for sponge)
    ~ automatic wire stripper
    ~ digital camcorder with still picture taking capabilities

- Problems
  * Terence said that the eMotes purchased by David Culler have issues that 
    prevent us from referring to the nodes via DNS (no problem with eMotes
    that were purchased separately by Intel).  We'll have to refer to them
    by their IP address.

- Related Work
  * Kamin Whitehouse and Rob (not Szewcyzk.  Another first year graduate 
    student) is constructing a testbed for localization at Intel
    which will use about 60 nodes
    ~ very similar setup to us: Power over Ethernet (PoE) and using mica2dots
    ~ we will harness some of their hardware and advice (ex. eMote to 
      pinOutBoard connectors, how to set up switch)
    ~ they purchased networking equipment through Albert Goto
      particularly the Foundry Power Over Ethernet Switches (48 port) and 
      ethernet cables
  * Terence Tong and Cory Sharp are building a network reprogramming testbed
    ~ we will be using Terence's scripts to remotely program and gather
      data from the motes through Ethernet.  See:
      http://www.ocf.berkeley.edu/~tmtong/tinyos/rsc.php