AIMD fallacies and shortcomings

1
AIMD fallacies and shortcomings

• Prasad

2

• 1

3

• AIMD claims
• Guess What !?

Proposition 3. For both feasibility and optimal
convergence to fairness, the increase policy
should be additive and the decrease policy should
be multiplicative.
4
AIMD claim is untrue !

• Consider the following simple example
• No. of users 2
• Init loads of users X1 17 and X2 0
• Load goal, Xgoal 20
• Fairness goal, Fgoal 99

5
AIMD equations
Let aI 1,aD 0, bD 0.01 and as per AIMD
claim, bI should be 1 Fairness index is given by
6
After plugging in all the values

• Result is (after 3 iterations)
• Now, change bI to 1.1. In other words,
• introduce a multiplicative-component during
• increase. Result then is (after 3 iterations)

7

• 2

8

• With AIMD, there is a possibility of unlimited
  overload after convergence

9

AIMD equations

After summing the values for n users we get,
10

• Defining overload to be
• We get
• Overload
• The problem is, as n becomes large, overload
  becomes large as well !

11

12

• 3

13

• AIMD is rather slow w.r.t convergence of
  efficiency

14

• 4

15

• All issues mentioned till now have one thing in
  common they are all related to the synchronous
  communication system

16

17

• This model is too simple and unrealistic and
  hence, inferences made based on it may not hold
  at all in a real system
• And Guess what !?

18

• 5
• This is the best part !

19

• AIMD does not guarantee fairness !
• (in a more realistic asynchronous communication
  system like the Internet)

20
A better model

21