Title: OptimalComplexity Optical Router
1Optimal-Complexity Optical Router
- Hadas Kogan, Isaac Keslassy
- Technion (Israel)
2Router schematic representation
Router
Optic to electronic
Electronic to optic
Optic to electronic
Electronic to optic
- Problem - electronic routers do not scale to
optical speeds - Access to electronic memory is slow and power
consuming. - Data conversions are power consuming as well.
3Power consumption per chassis
Nick McKeown, Stanford
4How about an optical router?
- No electronic memory bottleneck
- No O/E/O conversions
- BUT
- An optical router is thought to be too complex.
- Is it?
5- Objective quantify the fundamental complexity of
an optical router
6Quantifying complexity
- Quantify the fundamental complexity of an
optical router ? reduce into most basic
building blocks - Switching 2x2 switches(and input/output lines)
7Basic optical buffering component
- Buffering 2x2 switches (and input/output/fiber
delay lines) - Mode of operation
- (a) Write
- (b) Circulate
- (c) Read
(a)
(b)
(c)
1
1
1
8- The complexity of a system is the minimal number
of 2x2 switches needed to implement it.
9Quantifying complexity
- Complexity lower-bound To get a state-space of
size K in time T, the minimal number C of 2x2
switches needed is - Examples
- NxN switch
- Time Slot Interchange with time frame N
10Quantifying complexity
- A construction algorithm is said to be optimal if
its number of 2x2 switches grows like the
construction complexity. - Examples
- NxN switch (Benes is optimal)
-
- Time Slot Interchange
Benes 65
Jordan et al. 94.
11Optimal buffer emulation
12Emulation definition
- Original buffer
- Buffer emulation (with delay D)
6
5
4
3
2
1
t
D
13Emulation idea
- Objectif emulate buffer of size B
- Universal buffer any policy
- Idea schedule using frames of size B
- During any frame of B slots, observe which
packets leave the original buffer and color them
in blue - After some pipeline delay, send these blue
packets in the same order
F - Frame of size B
Frame-based scheduling
Optical buffer
14Algorithm
departure
4
B
5
2
- Algorithm is optimal
- Complexity T(ln B)
- Complexity lower-bound T(ln B) (the Time Slot
Interchange is a special case)
15Optimal router emulation
16What we want an ideal router
- An output-queued push-in-first-out (OQ-PIFO)
switch. - OQ - Arriving packets are placed immediately in
the queue of size B at their destination output. - PIFO packets departure ordering is according to
their priority.
17What we want an ideal router
- Why it is ideal
- OQ Work-conserving ? best throughput and average
delay. - PIFO Enables FIFO, strict priority, WFQ
- But up to N packets could be destined to the
same output - Speed-up for switch
- Speed-up for queue
- PIFO is hard to implement.
18Finding the complexity
- Direct calculation of complexity seems impossible
what are the states? - Alternative way of finding the complexity
- Find a lower bound
- Find an upper bound via algorithm
- Algorithm complexity T (lower bound) ?
algorithm is optimal
19Lower bound - intuition
B
At least T(Nln(N))
At least T(ln(B))
- Intuition the complexity of an OQ-PIFO switch is
at least T(N ln(N) N ln(B)) T(N ln(NB))
20Lower bound
- A frames switch (time/space switch)
t1
t2
t3
t4
t5
t6
t3
t4
t5
t6
t7
t8
t9
Frames switch
1
2
3
7
8
9
N
6
2
5
11
9
10
4
5
6
10
11
12
3
4
1
12
7
8
B
- A frames switch is a special case of an OQ-PIFO
switch. - The practical complexity of a frames switch?
Complexity OQ-PIFO T(Nln(NB)) - Now, find an algorithm that reaches this
lower-bound
21Solving the speed-up problem
Leaves output A at time 1
- Example Emulating a non-idling OQ-FIFO switch
D1
Output A
Input 1
A1
Output B
C1
A2
Input N
C2
- Using parallel buffers to resolve conflicts
- At most one packet can enter a buffer at each
time slot (N-1 constraints). - A packet departing at time T should not enter a
buffer with a packet departing at T (N-1
constraints). - ? 2N-1 buffers are enough.
22The pigeonhole principle
- Proof intuition
- Pigeons ? Packets
- Holes ? Buffers
- For emulating PIFO behavior
- The departure process is slightly modified
- 4N-2 parallel buffers are required
Input 1
Output 1
Input N
Output N
23An optical emulation of an OQ-PIFO switch
The pigeonhole principle Our emulation of an
optical buffer An optical OQ-PIFO switch
Optical buffer
Output 1
Output N
Optical buffer
24An optical emulation of an OQ-PIFO switch
(4N-2)xN switch
Nx(4N-2) switch
B
pB
pB
B
B
pB
pB
B
...
B
pB
pB
B
- Number of 2x2 switches T(NlnNNlnB)
T(Nln(NB)) T(lower bound)? algorithm is
optimal
25Conclusion
- Buffer fundamental complexity T(lnB)
- OQ-PIFO router fundamental complexity T(NlnNB)
- Both can be reached using given algorithms
26Thank you!