[net2-wg] information density in MultiHopLQI link metric

[Omprakash Gnawali]

I took a dump of all the adjustLQI(msg->lqi) received by all the nodes
in a network:

(aggregated log from all the nodes)
count, adjestLQI(msg->lqi)
      1 110
  20115 165
  28580 189
  20157 216
  10615 243
   6160 273
   4105 307
   3053 343
   2450 380
   2035 420
   1728 465
   1590 512
   1467 561
   1317 612
   1193 669
   1179 729
      1 740
   1084 790
   1046 855
   1008 926
   1053 1000
    977 1076
   1014 1155
    907 1241
    887 1331
    846 1423
    857 1518
    862 1621
    752 1728
    698 1837
    684 1950
    678 2071
    614 2197
    606 2325
    532 2457
    499 2598
    433 2744
    382 2892
    375 3045
    342 3208
    275 3375
    250 3545
    206 3720
    209 3906
    152 4096
    124 4290
     99 4488
     83 4698
     69 4913
     61 5131
     43 5355
     32 5591
     18 5832
     20 6077
     11 6327
     12 6590
      7 6859
      3 7132
      2 7410
      1 7702
      1 8000

When the count is 1, it seems that is a result of corruption. For
example, 110, 740.

Logs for individual nodes are at:
http://enl.usc.edu/~om_p/net2/lqilog-06152007/

- om_p