1. An Optoelectronic Neural Network Packet Switch Scheduler K. J. Symington, A. J. Waddie, T. Yasue, M. R. Taghizadeh and J. F. Snowdon. http://www.optical-computing.co.uk

2. Outline Packet switch scheduler. Previous demonstrator has proven system feasibility. Current demonstrator enhances functionality and performance. Motivation. Implementation and scalability. Conclusions.

3. The Assignment Problem Solution is computationally intensive. Neural networks are capable of solving the assignment problem. Their inherent parallelism allows them to outperform any other known method at higher orders. Can be found in situations such as: Network service management. Distributed computer systems. Work management systems. General scheduling, control or resource allocation.

4. Crossbar Switching

5. A size N crossbar switch has the same number of inputs as outputs: i.e. m=n=N. Crossbar Switching

6. Packets stored in buffer until output free. Packets can request any output line. Buffer depth very important. Real traffic tends to be ‘bursty’.

7. Crossbar Switching Channel operation exclusive. Maximum capacity of N packets per switch cycle.

8. Crossbar Switching Packet can only pass when crosspoint set. N 2 crosspoint switches required. Generic crossbar switch architecture.

9. Crossbar Switching Neural network chooses optimal set of packets. One neuron required for every crosspoint.

10. Crossbar Switching

11. Banyan Switching

12.    Routing input 2 to output 2 allows only 1 packet to pass. Solution is sub-optimal. Solution Optimality 24 2 Routing input 2 to output 2 allows only 1 packet to pass. Solution is sub-optimal.    Routing input 2 to output 4 and input 4 to output 2 allows 2 packets to pass. Solution is optimal.

13. The Neuron Inputs taken from the outputs of other neurons.

14. The Neuron Inputs taken from the outputs of other neurons. Synaptic weights multiply inputs.

15. The Neuron Inputs taken from the outputs of other neurons. Synaptic weights multiply inputs. Inputs are summed and bias added.

16. The Neuron Inputs taken from the outputs of other neurons. Synaptic weights multiply inputs. Inputs are summed and bias added. Transfer function f(x) performed before output.

17. Neural Algorithm x ij :Summation of all the inputs to the neuron referenced by ij: including the bias. y ij :Output of neuron ij. A, B and C: Optimisation parameters. ‘Iterations to Convergence’ is an important parameter. Iterations related to, but not necessarily equal to, time.  :Controls gain of neuron. Next state defined by: Neural transfer function:

18. Neural Interconnect

19. Convergence Example Start state – all requested neurons are on.

20. Convergence Example 1/3 Evolved: Neurons (2, 4) and (4, 2) are beginning to inhibiting neuron (2, 2).

21. Convergence Example 2/3 Evolved: Neuron (2, 2) is nearly off.

22. Convergence Example Fully Evolved. Optimal solution reached.

23. Neural network scalability limited in silicon. Optoelectronics allows scaleable networks. Free-space optics can be used to perform interconnection. Only transfer function f(x) need be performed in silicon. Input summation is done in an inherently analogue manner. Noise added naturally. Why Optoelectronics?

24. The VCSEL Array Optical output element. A laser that emits from the surface of the substrate. High optical output powers.

25. The VCSEL Array Each neuron has one VCSEL for optical output. Performance: Capable of >1GHz operation. Scalability: Currently N=16.

26. Detector Arrays Optical input element. Available in a wide range off the shelf. Performance: >1GHz. Caveat: faster detectors require more power.

27. Diffractive Optic Elements (DOEs) Large fan-out possible. Efficiency: ~50-60%. Non-uniformity: <3%. Period Size: 90µm. These elements are used as array generators and interconnection elements.

28. Crossbar switch interconnect. Banyan switch interconnect. Optical Interconnect DOE interconnect is space invariant.

29. Optical System

30. First Generation System Constructed using discrete components. Lacked ability to prioritise packets: can lead to channel saturation. Uses similar optical system (~330mm).

31. Current System System uses 4×40MHz Texas Instruments 320C5x DSPs. DSPs perform transfer function. Transfer function fully programmable. Reduction of hardware by digital thresholding.

32. System Scalability

33. Digital vs. Analogue Analogue: Optimal ~97%. Digital: Optimal ~91%.

34. Crossbar Switch Results Histogram of packets routed successfully in a crossbar switch.

35. Banyan Switch Results Histogram of packets routed successfully in a banyan switch.

36. Mean Packet Delay

37. ISLIP4 cannot be implemented larger than N=16.

38. Mean Packet Delay ISLIP4 cannot be implemented larger than N=16.

39. 3 Major effects to consider: Active effects: <1Hz thermal changes and component creep. Static effects: Tolerances in fabricated components could lead to misalignment in final system. Adaptive effects: Vibrational effects >1Hz - e.g. 10kHz. Solutions: Measurement and correction of focusing and positional error in real time (active optic alignment or adaptive optics). Commercially viable: e.g. personal CD player, ASDA £22:95. Pre-packaged, pre-aligned modules. Engineering Issues

40. Encapsulated System R. Stone, J. Kim and P. Guilfoyle, “High Performance Shock Hardened Optoelectronic Communications Module”, OC2001, Lake Tahoe, pp. 105-107.

41. Conclusions Performance of 100MHz feasible, 1GHz foreseeable. Scalability mainly limited by VCSEL array size (N=16). Scalability independent of number of inputs/outputs (N). A digital system running at 1GHz could supply 2.5 million switch configurations per second. Second generation builds on first in that it supports prioritisation. What good is a truck without a steering wheel? Further work: Smart pixel implementation and packaging. Examination of QoS provided by scheduler. FPGA or custom ASIC implementation using optical interconnects. Novel neural algorithms and learning.